FINAL REPORT Prepared For European Commission Directorate General for by knowledgegod

VIEWS: 0 PAGES: 145

									FINAL REPORT



Prepared For:

European Commission

Directorate General for Internal Market
and Services

B-1049 Brussels

Belgium




                  Potential cost savings in a
                  fully integrated European
                  investment fund market


Prepared By:

Tim Wilsdon and Kyla Malcolm



1 Undershaft

London EC3A 8EE, United Kingdom



Avenue Louise 480, Box 18

B-1050 Brussels, Belgium


Date:   September 2006

CRA Project No. D0-8482
Potential cost savings in a fully integrated European investment fund market

September 2006                                                                 CRA International




        Acknowledgements

        During the course of this project we spoke to a wide variety of industry participants,
        regulators and trade associations across ten Member States of the European Union and
        in the United States. We are also grateful for the time and effort put into completing
        detailed cost information that has fed into the analysis undertaken. We would like to
        thank all of those involved for their extensive help in the preparation of this report.

        We would also like to thank members of the European Commission’s Internal Market
        Directorate for their input and direction in writing this report.




                                                                                          Page i
Potential cost savings in a fully integrated European investment fund market

September 2006                                                                                                   CRA International



                                          TABLE OF CONTENTS


EXECUTIVE SUMMARY ....................................................................................................... 7


1.   INTRODUCTION........................................................................................................... 15

     1.1.      OUR APPROACH ................................................................................................................ 15
     1.2.      THE STRUCTURE OF THE REPORT ....................................................................................... 17


2.   METHODOLOGY FOR ASSESSING COST SAVINGS ............................................... 19

     2.1.      STYLISED FUND CHOICES ................................................................................................... 19
     2.2.      THE MEMBER STATES INVESTIGATED .................................................................................. 20
     2.3.      THE INVESTMENT FUND VALUE CHAIN ................................................................................. 21
     2.4.      METHODOLOGY FOR ESTIMATING COSTS............................................................................. 24
     2.5.      SCENARIOS USED TO ESTIMATE POTENTIAL COST SAVINGS .................................................. 28


3.   COSTS OF PRODUCTION COMPONENTS OF THE VALUE CHAIN......................... 31

     3.1.      REGULATION RELATING TO FUND CREATION AND MAINTENANCE ........................................... 31
     3.2.      FUND ADMINISTRATION ...................................................................................................... 43
     3.3.      CLIENT ADMINISTRATION .................................................................................................... 54
     3.4.      ASSET MANAGEMENT ......................................................................................................... 58
     3.5.      MANAGEMENT OVERHEADS AND SYSTEMS COSTS ............................................................... 63
     3.6.      SUMMARY OF PRODUCTION COSTS ..................................................................................... 65


4.   COSTS OF DISTRIBUTION COMPONENTS OF THE VALUE CHAIN ....................... 69

     4.1.      DISTRIBUTION COSTS BY TYPE OF FUND .............................................................................. 71
     4.2.      DISTRIBUTION COSTS BY MEMBER STATE ............................................................................ 73
     4.3.      DISTRIBUTION COSTS BY CHANNEL ..................................................................................... 78
     4.4.      DISTRIBUTION COSTS BY DOMICILE OF THE FUND ................................................................. 79
     4.5.      SUMMARY ON THE TOTAL COST OF THE VALUE CHAIN........................................................... 81


5.   QUANTIFICATION OF POTENTIAL COST SAVINGS................................................. 84

     5.1.      COMPARISONS WITH TODAY’S FUND SIZES .......................................................................... 84
     5.2.      COMPARISONS WITH EU AVERAGE FUND SIZES ................................................................... 86
     5.3.      COMPARISONS BASED ON TODAY’S COSTS BUT DIFFERENT FUND SIZES ................................ 88
     5.4.      COMPARISONS OF SAVINGS RESULTING FROM EFFICIENCY VERSUS SCALE ............................ 89
     5.5.      COMPARISONS INCLUDING DISTRIBUTION COSTS ................................................................. 90
     5.6.      TRANSLATING COST SAVINGS INTO POTENTIAL CONSUMER BENEFITS.................................... 92


                                                                                                                                 Page ii
Potential cost savings in a fully integrated European investment fund market

September 2006                                                                                                CRA International


6.   COMPETITION, INTEGRATION AND COSTS............................................................. 94

     6.1.       INTEGRATION AND COSTS – AT THE FUND LEVEL .................................................................. 95
     6.2.       INTEGRATION AND COSTS – ALONG THE VALUE CHAIN ........................................................ 102
     6.3.       COMPETITION AND COSTS – AT THE FUND LEVEL ............................................................... 106
     6.4.       COMPETITION AND COSTS – ALONG THE VALUE CHAIN ....................................................... 110
     6.5.       PASSING ON COST SAVINGS TO CONSUMERS ..................................................................... 118


ANNEX A: TECHNICAL DETAILS ON THE MODEL........................................................ 128

     STRUCTURE OF THE MODEL .......................................................................................................... 128
     ADDITIONAL SCENARIOS USED TO TEST THE IMPORTANCE OF SCALE ................................................ 135


ANNEX B: BIBLIOGRAPHY.............................................................................................. 140




                                               LIST OF FIGURES

         Figure 1: Comparison of total production costs for equity funds by member
             state based on current average fund sizes ...............................................8

         Figure 2: Comparison of total production and distribution costs for equity
             funds by member state based on current average fund sizes ...............9

         Figure 3: Countries in our sample ...................................................................21

         Figure 4: Comparison of gains from efficiency compared to gains from
             economies of scale .....................................................................................29

         Figure 5: Regulatory compliance costs for different asset types in basis
             points.............................................................................................................34

         Figure 6: Regulatory compliance costs for different countries (€ per fund
             per year)........................................................................................................36

         Figure 7: Fund accounting costs for different fund types..............................46

         Figure 8: Custody costs for different fund types.............................................48

         Figure 9: Audit costs by type of fund................................................................49

         Figure 10: Fund accounting costs for different member states....................50

         Figure 11: Comparison of the cost of custody by country of domicile ........52

         Figure 12: Comparison of audit costs by country (€ per fund per year) .....53



                                                                                                                            Page iii
Potential cost savings in a fully integrated European investment fund market

September 2006                                                                                            CRA International


        Figure 13: Client administration costs for different member states.............56

        Figure 14: Asset management costs versus size of funds (€m) by types of
            fund................................................................................................................61

        Figure 15: Asset management costs by member state.................................63

        Figure 16: Comparison of total production costs by member state based
            on current average fund sizes...................................................................65

        Figure 17: Comparison of total production costs for equity funds by
            member state based on current average fund sizes .............................66

        Figure 18: Comparison of total production costs for equity funds by
            member state based on average fund size for all member states ......67

        Figure 19: Ongoing cost of distribution by type of fund.................................72

        Figure 20: Ongoing cost of distribution as a percentage of the annual
            management charge...................................................................................73

        Figure 21: Comparison of the costs of distribution of an equity fund by
            member state ...............................................................................................75

        Figure 22: Comparison of the average distribution costs by member state
            for an equity fund.........................................................................................76

        Figure 23: On-going distribution as a percentage of the annual
            management charge for equity funds ......................................................77

        Figure 24: Distribution of investment funds by channel for all funds ..........78

        Figure 25: Comparison of total costs by fund type based on current
            average fund sizes ......................................................................................81

        Figure 26: Comparison of total costs for equity funds by member state
            based on current average fund sizes.......................................................82

        Figure 27: Comparison of total production costs for equity funds by
            member state based on average fund size for all member states ......83

        Figure 28: Comparison of total production costs for equity funds by country
            based on current average fund sizes.......................................................85

        Figure 29: Comparison of total production costs for equity funds by country
            based on average EU fund sizes..............................................................87

        Figure 30: Comparison of total costs for equity funds by country based on
            current average fund sizes ........................................................................91

        Figure 31: Cross-border integration versus costs ..........................................96


                                                                                                                         Page iv
Potential cost savings in a fully integrated European investment fund market

September 2006                                                                                               CRA International


        Figure 32: Mergers/closures as a percentage of funds on the market .....100

        Figure 33: The relationship between concentration and production costs
            ......................................................................................................................109

        Figure 34: Index tracking weighted average management fee..................123

        Figure 35: Index tracking standard deviation of management fee ............124

        Figure 36: The proportion of new business sold via direct and fund of
            funds channels...........................................................................................125

        Figure 37: Comparison of total costs by member state based on current
            average fund sizes for equity funds........................................................132

        Figure 38: Comparison of total production costs for bond funds by member
            state based on current average fund sizes ...........................................133

        Figure 39: Comparison of total production costs for bond funds by member
            state based on current average fund size .............................................134

        Figure 40: Comparison of total costs by member state based on current
            average fund sizes for bond funds .........................................................134

        Figure 41: Comparison of total production costs for equity funds by country
            based on maximum EU fund sizes.........................................................136

        Figure 42: Comparison of total production costs for equity funds by country
            based on average US fund sizes............................................................138



                                              LIST OF TABLES

        Table 1: Cost savings in production based on current fund size for equity
            funds..............................................................................................................10

        Table 2: Cost savings for equity funds.............................................................11

        Table 3: Cost savings in production and distribution based on current fund
            size for equity funds ....................................................................................11

        Table 4: Participation in interviews and cost survey......................................16

        Table 5: Components of the value chain.........................................................22

        Table 6: Cost of authorisation ...........................................................................37

        Table 7: Cost of notification ...............................................................................39

        Table 8: Time taken to get regulatory approval..............................................41

                                                                                                                            Page v
Potential cost savings in a fully integrated European investment fund market

September 2006                                                                                            CRA International


        Table 9: Cost savings in production based on current fund size for equity
            funds..............................................................................................................86

        Table 10: Cost savings in production based on average EU fund size for
            equity funds ..................................................................................................88

        Table 11: Cost savings due to increasing fund size ......................................89

        Table 12: Cost savings in production based on current fund size for equity
            funds..............................................................................................................90

        Table 13: Cost savings in production and distribution based on current
            fund size for equity funds ...........................................................................92

        Table 14: Impact of cost savings on investment returns...............................93

        Table 15: Description of the cost components in cost model.....................130

        Table 16: Linear regression results for custody fees of equity funds .......131

        Table 17: Cost savings in production based on at least average EU fund
            size for equity funds ..................................................................................136

        Table 18: Cost savings in production based on maximum EU fund size for
            equity funds ................................................................................................137

        Table 19: Cost savings in production based on average US fund size for
            equity funds ................................................................................................138




                                                                                                                        Page vi
Executive summary

September 2006                                                                     CRA International




       EXECUTIVE SUMMARY

       CRA International was commissioned by DG Internal Market of the European
       Commission to assess the potential for cost savings resulting from full integration of the
       European investment funds market. There are three main parts to the assignment:

       •   The identification and description of the cost function for the fund industry;

       •   The study of the relationship between cross-border integration, competition and
           costs; and

       •   The quantification of the potential cost savings linked to greater integration or
           competition.

       To investigate this issue we reviewed the existing literature, undertook 61 interviews with
       regulators, trade associations and industry participants in the ten member states under
       investigation (Belgium, France, Germany, Ireland, Italy, Luxembourg, Poland, Spain,
       Sweden and the UK) and analysed the results from 28 detailed quantitative survey
       responses (focused on the cost of a representative money market, bond, equity, capital
       guaranteed and index tracking fund).

       The identification and description of the cost function for the fund industry

       The first objective was to characterise the cost of funds sold in each of these member
       states and to compare the differences in cost. We found that:

       •   Regulatory costs associated to setting up and maintaining the fund are predominantly
           fixed and largely invariant to type of fund. There are some significant differences
           between member states with the biggest differences being due to tax compliance
           required in some member states.

       •   Fund administration has been broken into three main components: fund accounting;
           auditing; and custody. Each of these shows evidence of economies of scale.
           Auditing appears entirely fixed and invariant to fund type, with only some small
           variation between member states. Fund accounting does vary significantly by type of
           fund but we find very small differences between countries. Custody appears to vary
           both by country and type of fund (after allowing for the differences in scale).

       •   Client administration costs vary by member state, with a small number of countries
           appearing to have significantly higher transfer agency costs compared to the other
           member states primarily due to differences in the distribution model used in each of
           the countries. Transfer agency appears to be largely a variable cost and we could
           not identify significant differences in cost by type of fund (after accounting for size).

       •   The cost of asset management varies between types of fund. There is also strong
           evidence of economies of scale in this activity. However, we did not identify any
           differences in the cost of asset management by member state after allowing for scale.



                                                                                              Page 7
Executive summary

September 2006                                                                                                                                                     CRA International




                                              This is likely to reflect the fact that asset management is not always located in either
                                              the location of domicile of the fund or the location of sale of the fund since many fund
                                              managers arrange their asset management activities on a global basis or in regional
                                              centres rather than where the fund is domiciled or sold.

       Using the data from the quantitative survey and the interviews we have compared the
       cost of funds sold in different place. Figure 1 shows the current cost of production for an
       equity fund in each of the member states.

       Figure 1: Comparison of total production costs for equity funds by member state based on
       current average fund sizes


                                                     Regulator compliance of the fund    Audit         Fund Accounting   Custody         Transfer Agency   Asset management     Overheads

                                              90

                                              80
         Cost of production in basis points




                                              70

                                              60

                                              50

                                              40

                                              30

                                              20

                                              10

                                               0
                                                                               Germany




                                                                                                                Italy




                                                                                                                                                                                     UK
                                                     Belgium




                                                                                             Ireland




                                                                                                                            Luxembourg



                                                                                                                                               Poland



                                                                                                                                                           Spain
                                                                  France




                                                                                                                                                                       Sweden




       Source: CRA analysis


       Part of the reason for the difference in the current cost of production seen in Figure 1
       reflects the extent to which different member states are more or less able to exploit any
       gains from economies of scale. We have also compared the cost of funds after taking into
       account the differences in funds under management. After accounting for this, France and
       Sweden remain the cheapest two countries and Poland remains the most expensive
       country although the cost of production falls significantly for Poland reflecting the fact that
       it currently has the smallest average fund size and therefore would make considerable
       gains from economies of scale. The difference in cost between member states is much
       reduced when funds of the same size are compared suggesting that much of the current
       differences in the cost of production in investment funds is mainly due to scale.

       The next stage of the analysis is to consider distribution costs. In Figure 2 we compare
       the total production and distribution costs for an equity fund based on the current fund
       size in each member state.




                                                                                                                                                                                      Page 8
Executive summary

September 2006                                                                                                                                    CRA International




       Figure 2: Comparison of total production and distribution costs for equity funds by member
       state based on current average fund sizes


                                                            Regulatory compliance of the fund and overheads   Administration   Asset management   Distribution

                                           250
           Cost of funds in basis points




                                           200



                                           150



                                           100



                                            50



                                             0
                                                                           Germany




                                                                                                    Italy




                                                                                                                                                                 UK
                                                  Belgium




                                                                                       Ireland




                                                                                                                 Luxembourg



                                                                                                                               Poland



                                                                                                                                          Spain
                                                              France




                                                                                                                                                       Sweden
       Source: CRA analysis


       There are clearly very large differences between member states when distribution costs
       are included. Poland is found to be the most expensive country at 235 basis points with
       France the cheapest by some margin at 80 basis points. Examining the differences in
       total cost between countries, it is clear that the variation in distribution cost is the main
       component that drives differences between countries.

       The quantification of the potential cost savings linked to greater integration or
       competition

       To consider the potential cost saving from integration of the investment fund markets we
       consider four different scenarios relating to moving high cost countries to the cost position
       of lower cost countries:

       •                                   Average cost – For those countries where the cost of production is greater than the
                                           weighted average across countries, we reduce the cost of production to the weighted
                                           average;

       •                                   Cross border cost – Where the cost of production is greater than the average cost for
                                           Luxembourg and Ireland, we reduce the cost of production in other countries to this
                                           cost;

       •                                   Lowest cost – For all countries we reduce the cost of production to the cost of
                                           production in the lowest cost country (Sweden); and




                                                                                                                                                                 Page 9
Executive summary

September 2006                                                                     CRA International




       •     US cost – For all countries where the cost of production is greater than the cost of
             production in the US (in this case all ten member states), we reduce the cost of
             production that observed in the US.

       Table 1 below highlights the cost savings of these four scenarios compared to today.
       Looking at today’s equity funds in the ten member states, the total cost of production is
       estimated as around €6.4 billion.

       Table 1: Cost savings in production based on current fund size for equity funds

                                 Cost of production       Cost savings in €      Cost savings in bp

           Today                     €6,420 million                -                       -

           Average cost              €6,166 million          €254 million                2 bp

           Cross-border cost         €6,369 million           €50 million                <1 bp

           Lowest cost               €5,390 million          €1,030 million              8 bp

           US cost                   €4,635 million          €1,785 million              14 bp

       Source: CRA analysis


       This shows that the impact of reducing the cost of production for those countries with
       above average costs would be a saving of approximately €254 million per year, which is
       equivalent to a reduction in the costs of the fund by 2 basis points. Putting this in context,
       the average TER of equity funds across the ten countries is around 180 basis points, so
       this represents less than 1% cost saving. The small cost saving reflects that where
       countries have costs substantially above the average (Poland and Belgium), these tend to
       be the smaller markets for equity funds.

       Considering the second scenario, the benefits from moving to the production cost of
       cross-border funds is found to be very small indeed. This result is not surprising when we
       consider that Ireland and Luxembourg are not found to be among the cheapest of
       countries in terms of total production costs as seen above.

       However, we see more significant cost savings resulting from a move to the lowest cost
       member state or to US costs. Moving to the cost of production observed in Sweden would
       lead to a cost saving of some €1,030 million or 8 basis points. However, this is dwarfed
       by the potential savings that would arise if all member states could replicate the current
       costs of the US where nearly €1.8 billion of cost savings or 14 basis points could be
       achieved. However, even assuming a wholesale move to US costs, this represents less
       than a 10% reduction in the total expenses currently charged to the funds.

       Examining the scenarios provided in Table 1 above combines the effect of efficiency and
       scale since some countries will have relatively low costs of production because they are
       producing relatively large funds. We have therefore separated out these effects.

       In order to measure the potential arising from efficiency alone we first estimated the cost
       of production for the average fund size across Europe. We then modelled the savings
       that would occur if the more expensive funds in Europe moved to the average cost in


                                                                                                 Page 10
Executive summary

September 2006                                                                                 CRA International




       Europe or to the minimum cost in Europe. In order to measure the impact of scale, we
       have calculated the cost savings that would accrue if funds all increased in size to either
       the largest average fund size observed in any of the member states or to the average
       fund size observed in the US.

       Table 2: Cost savings for equity funds

                                                    Scenario                             Cost savings in bp

                                                   Average cost                                  2 bp
        Efficiency effects1
                                                    Lowest cost                                  8 bp

                                         Maximum member state fund size                          6 bp
        Scale effects2
                                               Average US fund size                              17 bp

       Source: CRA analysis


       Table 2 shows that by moving to the most efficient cost structure in Europe a saving of 8
       basis points could be made, whereas moving to the average US fund size would bring a
       saving of some 17 basis points. This clearly shows that the cost savings arising due to
       scale are even bigger than those arising from increased efficiency.

       As well as considering the potential cost savings in the production component of the value
       chain, we also examine savings from the total value chain including distribution. The
       results of this are found in Table 3 below.

       Table 3: Cost savings in production and distribution based on current fund size for equity
       funds

                                          Total cost              Cost savings in €          Cost savings in bp

        Today                            €16,979 million                    -                            -

        Average cost                     €15,451 million              €1,528 million                 12 bp

        Lowest cost                      €11,627 million              €5,352 million                 41 bp

        US cost                          €9,270 million               €7,710 million                 60 bp

       Source: CRA analysis


       It is clear from Table 3 that there are very substantial potential cost savings once
       distribution costs are included in the analysis. If those countries that have total costs
       above the current weighted average reduced to the current average, this would lead to a




1      Efficiency effects are based on moving to the average and lowest cost of production for a fund size at the
       member state weighted average.

2      Scale effects are based on moving to different fund sizes but assuming that the current cost function in each
       member state is retained.



                                                                                                             Page 11
Executive summary

September 2006                                                                       CRA International




       potential cost saving of €1.5 billion or 12 basis points. If costs fell as far as the lowest
       cost country, the figures would be a huge €5.4 billion or 41 basis points and if they fell as
       far as the current US cost figures then there would be potential for cost savings of some
       €7.7 billion or 60 basis points. Comparing this to the average TER, this represents a 33%
       reduction in costs of total costs of the fund.

       Given that the variation in the cost of distribution was seen to be much greater than the
       variation in other parts of the value chain, and since distribution is the largest component
       of costs in all member states, it is not surprising that when distribution is included in the
       scenario analysis the potential cost savings are substantial.

       From the overall analysis we can conclude that:

       •   While it is the case that in Europe some of the very largest funds may have already
           exploited economies of scale, it is clear that for the majority of funds there is
           someway to go before this is the case. Hence integration and other measures that
           allow for investment funds to grow and to consolidate would be expected to lead to
           lower costs of provision;

       •   If greater integration of the investment fund market means more funds are domiciled
           in Ireland and Luxembourg this alone would not result in a significant cost reduction;

       •   The potential cost saving resulting from distribution easily outweighs the potential for
           cost savings resulting from producing of funds; and

       •   Even though there are clear gains to be made from increasing the size of funds in
           Europe, best practice European funds are highly efficient, indeed, this analysis
           suggests that allowing for scale, they are at least as efficient as US comparators.

       The relationship between cross-border integration, competition and costs

       To understand whether potential cost savings can actually be achieved it is important to
       understand whether existing cost differences reflect differences in the intensity of
       competition in each of these markets, the level of cross-border integration or other
       factors, and whether ongoing changes in the market will result in these cost differences
       being exploited in the future (without any change in the regulatory environment).

       Generally we expect more competitive markets to be those where providers exploit cost
       efficiencies in order to offer their customers the best value products. Equally we expect
       markets with greater cross-border integration, because it exposes providers to a wider
       range of competitors than only those in the domestic market and it allows providers to
       take advantage of scale and specialisation, would have lower costs.

       Considering the current position, we find that although there is a general agreement that
       investment fund markets are broadly competitive, there is little evidence to support the
       hypothesis that more competitive markets, as measured in terms of concentration or
       looking at market structure, result in lower cost funds. It is likely that this reflects a trade-
       off between the benefits of having many rival providers in the market and the necessary
       concentration required to exploit economies of scale.


                                                                                                Page 12
Executive summary

September 2006                                                                    CRA International




       Examining the level of cross-border integration, there is evidence that markets with a
       higher level of cross-border integration (as measured by the proportion of funds that were
       notified from abroad) have lower costs. However, as noted above countries that are
       generally associated with cross-border exporting (Ireland and Luxembourg) are not
       actually found to be the locations where costs are lowest.

       At the overall level of the fund, one of the key concerns regarding competition and
       integration raised throughout the study was that of fund consolidation. Although there
       have been many significant domestic mergers of funds, cross-border mergers are much
       more difficult due to problems of shareholder approval, whether such mergers are seen
       as a taxable event and restrictions preventing the transfer out of funds from some
       jurisdictions. There is no evidence that market trends alone are able to change this.
       Hence, it will not be possible to merge existing legacy funds, and substantial increases in
       fund sizes which would yield significant cost savings are prevented from arising.

       Looking within the value chain, we find different implications regarding competition and
       integration for the different activities in the chain. For example, in the regulatory
       component, improvements in authorisation and notification could improve integration.
       However, given the relatively small element of costs that these represent this is unlikely to
       realise big cost savings.

       In fund and client administration, regulatory restrictions have resulted in activities having
       to occur in the country of domicile with the result of some duplicative costs preventing
       gains from integration being made. Within countries, there is convincing evidence that we
       will continue to see further outsourcing for particular activities among smaller fund
       managers, with the result that contestability of fund and client administration are
       progressively leading to efficiency improvements. This is likely to lead to lower costs over
       time, although not necessarily to convergence in costs due to country specific factors.

       In the asset management activity, where integration is well progressed, there is general
       agreement that this represents a global market and any cost savings arising from
       efficiency improvements appear to have already been largely achieved. However, it is in
       the asset management activity where many of the gains from fund consolidation would
       occur.

       The most significant activity identified as lacking competitive discipline is distribution.
       Access to distribution was highlighted by fund promoters as a significant barrier; however,
       they themselves recognised that it was no more difficult for a cross-border fund than a
       domestically domiciled funds. In other words, access to proprietary channels is a problem
       from a competition and an integration perspective. Nonetheless, there is some evidence
       that distribution is opening up to external providers through open architecture and new
       channels offering consumers choice but for the moment these changes are mainly
       targeting high-value customers. Any significant changes are therefore likely to only arise
       slowly over time.

       Therefore, although there are many areas where costs will converge over time, many of
       these are slow moving. We do not therefore believe that ongoing trends identified above
       are likely to lead to significant convergence in costs over the short-term.



                                                                                            Page 13
Executive summary

September 2006                                                                    CRA International



       Would any cost reductions be passed onto consumers?

       The final question addressed in this report is whether cost savings, if they arose, would
       be passed onto European consumers. There is concern that due to the perceived
       complexity of products and the reliance of consumers on their existing distribution
       channels, any cost saving would not be passed on. Indeed, this is supported by evidence
       that fund mergers have not resulted in better terms being offered to consumers.

       However, European index trackers show evidence both of falling prices and price
       convergence, consistent with competition resulting in cost reductions being passed to
       consumers. In addition, evidence from US markets demonstrates price sensitivity, and
       trends in new distribution channels are arising which should encourage more active
       consumers over time. Finally, there is a regulatory trend to improve the transparency of
       investment funds through the provision of better information.

       However, even after taking these observations into account, we conclude that without
       further measures to encourage financial capability it is likely that any cost savings arising
       from integration will only be passed onto consumers very slowly.




                                                                                            Page 14
Introduction

September 2006                                                                         CRA International




1.      INTRODUCTION

        CRA International was commissioned by DG Internal Market of the European
        Commission to assess the potential for cost savings resulting from full integration of the
        European investment funds market. The analysis undertaken relates only to those
        investment funds that comply with the UCITS Directives.3

        There are three main parts to the assignment:

        •      The identification and description of the cost function for the fund industry;

        •      The study of the relationship between cross-border integration, competition and
               costs; and

        •      The quantification of the potential cost savings linked to greater integration or
               competition.


        1.1.       OUR APPROACH

        In order to undertake this project detailed cost data and a good understanding of how the
        investment fund market work in each of the member states was required. Therefore we
        adopted an interview-based approach supported by a detailed quantitative survey
        capturing data on the costs of activities along the value chain.

        Before embarking on the main interview programme we:

        •      undertook a literature review of previous attempts to characterise the investment fund
               value chain (although it should be recognised that these were often created with other
               purposes in mind). The value chain set out in this report builds extensively on these;
               and

        •      undertook preliminary interviews with individual industry associations (the IMA in the
               UK and the BVI in Germany) as well as with the European Fund Management
               Association (EFAMA) in order to test this value chain to ensure it was appropriate to
               all member states.

        An overview of the resulting value chain is set out in Chapter 2.

        To understand the relationship between competition, cross-border integration and costs,
        to collect existing studies on the marketplace and to collect cost data on different activities
        along the value chain we have undertaken 61 detailed interviews with market participants,
        regulators and investment fund trade associations in 10 member states as well as with
        pan-European organisations as detailed in Table 4 below.




3       UCITS III Directive 2001/107/EC.



                                                                                                 Page 15
Introduction

September 2006                                                                                   CRA International




        The interviews also allowed us to identify and test on-going trends in the market.
        However, our approach to this has been largely qualitative.       A second study
        commissioned by the European Commission is looking in more detail at whether the
        quantitative evidence supports each of the trends.

        Interviewees were each sent copies of the interview notes in order to verify the accuracy
        of understanding and information. We quote directly from the interviews only where the
        interviewee has given their consent, although the majority of information quoted is
        attributed anonymously to fund managers. The country appendices setting out the
        qualitative description of each of these markets were also provided to domestic trade
        associations and comments received have been taken into account.

        We have supplemented the data gathered from interviews with a quantitative survey.
        This survey was sent to a wider group of asset managers, fund administrators and
        distributors than those interviewed. This detailed survey was sent to around 80 fund
        managers and external administrators. The results in this report are based on 28 detailed
        quantitative surveys.

        Table 4: Participation in interviews and cost survey

                  Country                        Participation in interviews and/or cost survey

          Belgium                                             CBFA, Fortis, ING, KBC

          France                              AFG, AMF, BNP Paribas, Diiffusion, Generali, Grouprama

          Germany                                    Allianz GI, BaFin, BVI, DWS, Indexchange

          Ireland               Bank of Ireland Securities, Dublin Funds Industry Association, IFSRA, Invesco, KBC,
                                                             Pioneer Investments, Russel

          Italy                 Assogestioni, Banca d’Italia, CAAM Sgr (formerly Nextra), Consob, Fineco, Pioneer
                                                               Investments, Sanpaolo

          Luxembourg                          ALFI, BNP Paribas, CSSF, EFA, Fortis, KBL, State Street

          Poland               Citigroup, Izba Zarzadzajacych Funduszami I Aktywami (IZFIA) Chamber of Investment
                                Funds Managers, KBC Asset Management, Komisja Papierow Wartosciowych I Gield
                                         (KPWIG) Securities and Exchange Commission, PKO/Credit Suisse

          Spain                                  BBVA, CNMV, Inverco, Morgan Stanley, Santander

          Sweden                    Didner Gerge, Finansinspektionen, Fondbolagens Förening, Handelsbanken, HQ
                                                            Fonder, Nordea, Robur, SEB

          UK                   AXA, Fidelity, Financial Services Authority, Gartmore, IMA, JP Morgan, Jupiter, Merrill
                                                                  Lynch, Schroders

          Europe-wide                                            EFAMA, Euroclear

        Source: CRA International


        The data collected through the interviews undertaken for this project and the detailed
        quantitative survey has been used in conjunction with external data sources. Our
        assessment of the available data sources through interviewing data providers,
        discussions with industry associations and the European Commission, led us to the



                                                                                                              Page 16
Introduction

September 2006                                                                         CRA International




        conclusion that no single dataset covers the entire value chain or is sufficiently
        disaggregated. However, we have extensively used Fitzrovia, FERI and EFAMA data.

        The Fitzrovia data was an important component of our background research prior to the
        interviews. There are clearly some areas where the Fitzrovia data does not, and does not
        claim to, provide a full picture of cost to the fund e.g. it is weak on splitting out the costs of
        distribution. In other areas, it provides useful top-line data but is not as disaggregated as
        we would have liked. However, in a number of areas, such as audit and custody, our
        interviews supported the fact that the Fitzrovia data accurately captures the true cost of
        these activities. There we have used the survey undertaken for this survey only to
        corroborate the Fitzrovia data. In agreement with the European Commission we
        structured the questionnaires to obtain the maximum complementary information to that
        already available.

        An important issue with regard to data collection is the issue of confidentiality. The data
        collected contains very commercially sensitive information. We therefore only present
        information in this report (and to the European Commission) at an aggregate level for the
        industry in a particular member state or for a particular activity. For this reason, it is not
        possible to show the underlying observations feeding into particular charts and we report
        averages or only show trend lines fitted to the data. We will not release any quantitative
        information that could be identified to a particular firm.

        Since financial services markets are dynamic markets which can rapidly change and the
        regulatory debate in each member state is constantly under review, it must be noted that
        information contained within this report is accurate at the time of writing and may become
        out of date as the relevant markets develop. Due to the length of this project, the date of
        writing will vary from country to country between March 2006 and June 2006.


        1.2.     THE STRUCTURE OF THE REPORT

        In the next chapter we address some of the methodological issues. For example, how we
        have chosen to focus on particular types of fund, the countries we have investigated, the
        decomposition of the value chain and how we try to differentiate between cost differences
        that arise due to inefficiency and those that arise because different sizes of funds.

        In Chapter 3, we review the evidence in each stage of the value chain in producing a fund
        (regulation, fund and client administration, asset management, and management
        overheads and systems) and consider whether costs vary by type of fund, by member
        state they are sold in, and set out the key drivers of these costs.

        In Chapter 4, we look at the evidence regarding distribution costs and consider whether
        these costs vary by type of fund and by member state they are sold in, and set out the
        key drivers of these costs.

        In Chapter 5, we examine the size of the potential cost savings that could result from
        further cross-border integration by looking at how different scenarios would impact on the
        cost of funds in the ten member states under investigation.




                                                                                                  Page 17
Introduction

September 2006                                                                 CRA International




        Finally, Chapter 6 reviews the evidence regarding whether the link between costs, cross-
        border integration and competition, how this will change due to on-going trends and the
        implications for whether any potential cost savings would be passed onto consumers.




                                                                                         Page 18
Methodology for assessing cost savings

September 2006                                                                      CRA International




2.      METHODOLOGY FOR ASSESSING COST SAVINGS

        In order to make meaningful comparisons of the different funds around Europe it was
        important to make a number of methodological decisions:

        •   Firstly, there are clear differences in the cost structure of different types of fund and
            the importance of different fund types varies significantly from country to country. We
            therefore chose a number of stylised fund choices to focus our study on;

        •   Secondly, given the level of detail required to make meaningful comparisons it was
            not possible to look at all 25 member states. We therefore focused the analysis on
            ten member states with the aim of making the results as representative as possible;

        •   Thirdly, as discussed in the introduction, to make meaningful comparisons we needed
            to have a common value chain that would help us to make like for like comparisons;

        •   Fourthly, it was important to allow for the structure of costs i.e. we wanted to
            understand how costs vary with the size of the fund and the degree to which
            elements of costs are fixed or variable; and

        •   Fifthly and finally, we set out a set of stylised scenarios that could be tested using the
            cost data to estimate the potential savings from a fully integrated investment fund
            market.


        2.1.     STYLISED FUND CHOICES

        To consider which funds to examine, we looked at the classification used by industry; the
        activities undertaken by different funds and implications for what we can learn from
        comparing across funds; and reviewed the data on the types of fund chosen in different
        member states. This resulted in us choosing to focus on five fund categories:

        •   Money market funds;

        •   Bond-based funds which primarily focused on government based bonds;

        •   Capital protected funds including all funds that offer a capital guarantee;

        •   Equity based index tracker focusing on the domestic stock market index; and

        •   Equity based, widely diversified, actively managed funds excluding sector specific
            funds (technology, pharmaceutical, emerging markets, small caps).

        We have not focused on balanced funds. Although, these are a popular fund choice
        throughout Europe, the results from equity, bond and money market can be used to draw
        conclusions about this category. Particular issues relating to this category were discussed
        where appropriate during the interviews.




                                                                                              Page 19
Methodology for assessing cost savings

September 2006                                                                    CRA International




        2.2.      THE MEMBER STATES INVESTIGATED

        We did not set out to collect data from all 25 Member States but rather segmented the
        European investment fund market into different types. The criteria, agreed with the
        European Commission, were based on:

        •      Size of market and market growth - we wanted to include the biggest consumer
               markets such as France, Germany, Italy and the UK. However, it is possible that the
               biggest potential benefits will be in markets that are currently reasonably small but
               where there is the possibility of considerable growth in the investment funds market.
               This is particularly the case as we are looking for benefits over a 20 year time
               horizon and hence we included Poland as a market demonstrating considerable
               growth;

        •      Degree of openness (both as a fund importer, e.g. Italy and as fund exporter, e.g.
               Luxembourg and Ireland, and countries with low levels of cross-border trade, such
               as bonds funds in France);

        •      Form of distribution (comparing countries that predominantly use bank distribution
               such as Spain to those with a bigger focus on specialist intermediaries such as UK);

        •      Types of fund (including countries mainly focused on money market funds such as
               France and those with an equity focus such as Sweden);

        •      Existing nature of investment fund market, i.e. average size of fund and average
               fund charges (where we might contrast Italy which is often seen as having the high
               charges to a country such as Germany which is seen as having low charges); and

        •      Income of the country and the development of the financial services market (for
               example including historically relatively poorer member states).

        The result of these selection criteria resulted in the following member states: Belgium,
        France, Germany, Ireland, Italy, Luxembourg, Poland, Spain, Sweden, and the UK. In
        2004, these member states had over €3.7 trillion assets under management in UCITS
        funds, representing around 90% of assets under management in UCITS funds across
        Europe.4 In addition to these member states, we have also briefly examined the US in
        order to have a further comparator for examination.




4       EFAMA Factbook 2005.



                                                                                             Page 20
Methodology for assessing cost savings

September 2006                                                                      CRA International




        Figure 3: Countries in our sample




                                                           Sweden




                                                   UK


                                         Ireland
                                                        Belgium



                  US                                                                        Poland
                                                                                          Germany
                                                                                         Luxembourg
                                              France




                                                        Spain
                                                                    Italy


        Source: CRA International



        2.3.     THE INVESTMENT FUND VALUE CHAIN

        One of the first tasks of the analysis was to set out the activities undertaken in the
        provision of investment funds. There are three main reasons why this was important:

        Firstly, it is important to make sure we are comparing like-for-like funds. If this is not the
        case then differences in the overall level of costs could be explained because different
        funds undertake different activities. For example, guaranteed funds have an additional
        cost because of the provision of the guarantee that is not incurred by other funds;
        similarly index funds do not need to undertake research regarding stock selection where
        other equity funds would.

        Secondly, we want to decompose the value chain so we can compare the cost of different
        functions across countries to see how this depends on competition and integration of
        markets. This is important because if different participants in the value chain undertake
        different activities, this could also explain observed differences in costs between these
        participants. For example, if custodians were to undertake fund accounting for a fund, this
        could lead to misleadingly high custody figures because they are also performing another
        function in the value chain.

        Thirdly, we want to investigate the degree to which these costs vary with size of the fund
        (as measured by the number of customers or assets) or the extent to which costs are
        fixed. Identifying the nature of cost for different parts of the value chain allows us to




                                                                                              Page 21
Methodology for assessing cost savings

September 2006                                                                                  CRA International




        determine the impact of increasing fund size which is an important issue when
        investigating further integration of markets.

        The first task was to determine what should be included in the value chain for this project
        and what should not. In agreement with the European Commission steering group we
        therefore set out:

        •      Activities included in the value chain: The value chain set out includes some activities
               that are not directly involved in the production of investment funds but which are
               complementary activities required to reach the end consumer. Given the importance
               of distribution costs we believe it is important to account for these costs and hence
               we have included advice costs in the distribution component of our value chain. It is
               also likely that the costs of advice substitute for other forms of marketing and
               distribution costs (which would be charged to the fund) where advice does not arise.

        •      Activities excluded from the value chain: At the other end of the value chain we have
               excluded trading costs. Although trading costs vary by country and affect the overall
               returns of the investment fund, they are a characteristic of where the trading is
               undertaken and the functioning of different exchanges rather than the service of the
               investment fund provider. We see this as an input into the investment fund value
               chain.

        Given the objectives above a value chain was created with sufficient detail, so we could
        be confident we were comparing like for like without becoming so detailed we would not
        be able to collect the corresponding data. The resulting value chain is set out in Table 5.

        Table 5: Components of the value chain

             Category               Function                                Brief description

                             Authorisation and          Authorisation and notification of the fund with the
                             notification               competent authorities

                             Fund compliance            Regulatory reporting and monitoring activities relating to
                                                        the fund

                             Tax compliance             Ensuring fund meets necessary tax rules in the member
            Regulatory
                                                        state
            compliance of
            the fund
                             Documentation production   Designing and producing any necessary documents about
                                                        the fund that will be sent to investors

                             Investor protection        Most funds will have some element of consumer
                                                        protection enabling investors to receive compensation in
                                                        the event of a deficiency with either the fund or the fund
                                                        management company

            Fund and         Fund accounting            Provisions of valuations, tax reclaims and management
            client                                      information, calculation of the NAV
            administration
                             Auditing                   External audit of the fund

                             Fund order processing      Automated processing from the deal to the administrator




                                                                                                              Page 22
Methodology for assessing cost savings

September 2006                                                                                    CRA International




                             Performance                    Provision of investment performance reports, attribution
                             measurement                    analysis of returns

                             Safe custody                   Security safe-keeping and control

                             Depositary / trustee           Oversight of the fund by the depositary
                             oversight

                             Stock lending                  Arranging and processing loans of stocks and bonds

                             Administration of shares /     Client dealing and associated administration including
                             units (including transfer      contract notes, distribution and trustee liaison, opening
                             agency)                        accounts for clients

                             Shareholder services           Payment of income, reinvestment of dividends, valuation
                                                            reports to customers

                             Cash management                Placing deposits, foreign exchanges

                             Research                       Fundamental & technical economic and company analysis

                             Strategic and tactical         Long-term asset allocation, currency and risk
                             asset management / asset       management
                             allocation
          Asset
          management
                             Operational asset              Stock selection, decision making and implementation,
                             management and dealing         decisions to buy and sell investments, netting of trades,
                             decisions                      pre-trade broker liaison, deal administration and control,
                                                            post-trade liaison with brokers and custodian

                             Guarantee provision            Hedging portfolio in order to provide a guarantee on to the
                                                            capital value of the fund or on the returns made

                             Management company             Overhead allocation for premises, senior management,
                             overheads                      cost of capital requirements at the management company
                                                            level
          Management
          overheads
                             Systems development            Planning and implementation of new IT and major
          and systems
                                                            enhancements to existing systems

                             Systems maintenance            Operational and technical maintenance of existing IT

                             New product development        Development of characteristics associated with design of a
                                                            new fund

                             Promotional activity           Advertising to gather assets (including internal sales and
                                                            marketing costs where you have an estimate for these)

                             Compensation to                Sales activities including commission or other payments to
                             distributor for reaching end   distributors
          Distribution       customer
          and marketing
                             Documentation provision        Provision of marketing and product documentation

                             Distribution compliance        Regulatory requirements regarding the conduct of
                                                            business or sale of investment funds

                             Advice                         Many funds bundle the provision of advice into the cost of
                                                            distribution; others will be provided on an execution only
                                                            basis

        Source: CRA International, IFSL, interviews.




                                                                                                                Page 23
Methodology for assessing cost savings

September 2006                                                                                       CRA International




        As can be seen from Table 5 above, we distinguish between categories of activities in the
        value chain and different functions which make up these categories.5 In grouping these
        activities we have attempted to aggregate similar aspects of the value chain. It should be
        noted that this does not necessarily suggest that moving down the table would represent
        the chronological order in which a new investment fund would be put together.
        Depending on whether the value chain under consideration is that relating to a new fund
        being developed, services to a consumer, or whether the activities are undertaken by the
        same firms, a different order might be determined for the activities. The ordering above
        therefore represents a compromise. A more detail description of each of the production
        components is set out in the next chapter and the distribution elements in Chapter 4.


        2.4.      METHODOLOGY FOR ESTIMATING COSTS

        Our aim is to understand the current cost of undertaking these activities for an “average”
        fund but as important is to understand the nature of these costs. Costs may be fixed at
        the management company level, fixed for a given fund or may vary with the activities of
        the fund, and therefore linked to the number of transactions, the funds under
        management, and the number of customers.

        In devising the surveys for this project we have focused on the most significant costs and
        have tried to capture the underlying cost structure.


        2.4.1. The level of granularity

        The objective in has been to collect data for each of the activities described in Table 5
        above. However, it has been important to aggregate the data in some areas. The degree
        to which we can estimate the cost of every component of the value chain depended on:

        •    The data that is already collected by the industry because it is published by
             distributors, fund managers or because of regulatory restrictions that require data to
             be collected by providers for use in accounts or the simplified prospectus;

        •    The need to aggregate data to focus on the areas of most interest in terms of
             consumer benefits;

        •    The degree to which it represents useful management information – not all costs are
             separated for accounting purposes, a good example is where shared costs are
             spread across many funds. Some firms will choose to allocate these shared costs
             others will not;




5       Although there are a large number of different stages in the value chain, it is important to note that the same firm
        may undertake many of these different functions.        For example, the fund manager would be expected to
        undertake both strategic and operational fund management, although research could be done by investment
        banks.



                                                                                                                   Page 24
Methodology for assessing cost savings

September 2006                                                                   CRA International




        •   The willingness to disclose commercially sensitive data – there is a particular
            sensitivity regarding distribution data with providers unlikely to disclose commission
            arrangements on particular funds or with particular distributors and hence we asked
            for average data across a funds category; and

        •   A de minimus rule – there are some costs that are not separately accounted as the
            cost is too small to warrant the cost of collection. For example, we have not focused
            on some parts of the value chain such as cash management costs, since these costs
            are considered likely to be negligible for our purposes.

        A significant methodological issue is the degree to which we are able to measure the cost
        of undertaking the activity or only the price charged to the customer. There is a desire to
        understand the cost drivers at different stages of the value chain but where activities are
        outsourced we will often only be able to observe the price charged for the activity. As
        agreed with the Commission, in this report we have focused on the costs charged to, or
        incurred by, the fund where this is possible. We have used the interviews to test what
        drives these costs and the degree to which cost differences are likely to be due to
        differences in the degree of competition in a particular part of the value chain or due to
        differences in the underlying cost conditions.


        2.4.2. Characterising the cost structure

        In order to understand the differences between the cost structures in different countries
        an important factor to be taken into account is the importance of scale. Economies of
        scale result when the cost of undertaking an activity falls as the quantity increases. This
        is commonly observed where there are high fixed costs since as quantity increases these
        fixed costs can then be shared across a greater number of units of production. That there
        are economies of scale in the fund management industry is well documented both in
        academic analysis and in analysis undertaken by regulators and industry participants.




                                                                                           Page 25
Methodology for assessing cost savings

September 2006                                                                                        CRA International




        Evidence on economies of scale6

        There have been many studies on the existence of economies of scale in asset management. The most
        significant studies include:


        In the US, the SEC found that mutual fund expense ratios declined as the amount of fund assets increased.
        Other things held equal, a fund with assets of $10 million had an operating expense ratio that was 22 basis
        points lower than a similar fund with assets of $1 million. A fund with assets of $1 billion had an operating
        expense ratio that was 66 basis points lower than a similar fund with assets of $1 million.7 Other research finds
        that while there are economies of scale and average costs diminish over the full range of fund assets, the rapid
        decrease in average costs is exhausted by about $3.5 billion in fund assets.8 This is the analysis used in the
        Invesco study of the benefits resulting from cross-border mergers.9


        Similarly, a comparison of the Canadian market finds that the average expense ratio paid by Canadian mutual
        fund investors is 50% higher than that paid in the United States. This research found that 24% of the
        discrepancy is accounted for by Canadian funds not taking advantage of economies of scale and because they
        face less competition along with other measurable fund attributes.10


        In Europe, there have been a number of studies investigating economies of scale.               These have found
        economies of scale and scope in the French mutual funds (SICAV) industry. The results suggested economies
        of scale and scope for small institutions but diseconomies for larger firms. An appropriate size for a diversified
        company is in the range of FF 2.9 billion.11


        Fitzrovia analysis show there is a clear relationship between the size of the fund and the total expense ratio after
        allowing for domicile (by comparing funds domiciled in Luxembourg).12




6       This literature review represents all of the available academic research on the relationship between investment
        funds and economies of scale. The evidence identified relates to data covering a range of different periods and
        therefore caution needs to be taken when interpreting evidence. It is particularly important to note that in a
        dynamic market such as investment funds, some of the older papers may now be less relevant due to structural
        changes in the market. In addition, most of the academic literature is based on the US market experience.

7       Report on Mutual Fund Fees and Expenses, SEC - Division of Investment Management (2000), available at
        http://www.sec.gov/news/studies/feestudy.htm.

8       Economies of Scale in Mutual Fund Administration, Latzko, 1999, Journal of Financial Research, 22.

9       Benefits of an integrated European Fund Management: Cross border merger of funds a quick win? Invesco
        January 2005.

10      Expense ratios of North American mutual funds, Ruckman, 2003, Canadian Journal of Economics, 36.

11      Economies of Scale and Scope in French Mutual Funds, Jean Dermine and Lars-Hendrik Roller, Journal of
        Financial Intermediation, 1992, 2, pp. 83-93. FF 2.9 billion is approximately €440 million.

12      Are European Funds Achieving Economies of Scale? Fitzrovia Press Release 22, Many 2002, London.



                                                                                                                   Page 26
Methodology for assessing cost savings

September 2006                                                                                            CRA International




        There have also been recent studies by European regulators. Recently, the AMF studied the relationship
        between size and the cost (in terms of the performance of funds i.e. the returns to investors) of funds in
        France.13 The study did not show a clear connection between the size and cost of funds. One reason for this
        was that collecting cost data on French funds was found to be difficult. Before 2006, French asset management
        companies were not obliged to publish TERs but an indicative fee. Another reason for the lack of relationship
        between the size and the performance of funds was thought to be that large funds are generally not high
        performing because it is more difficult for large funds to be innovative.


        An analysis undertaken by Consob in Italy used data of 43 asset management companies for the financial year
        2003, this found the ratio of operating costs (i.e. staff, advisory, data-processing, and other administrative costs)
        to assets under management decreased more than proportionately as assets under management increase,
        confirming the presence of economies of scale. Consob also found that the ratio of the costs incurred by
        investors (which include front-end and back-end load charges, management and incentive fees) to assets under
                                                                                           14
        management does not vary significantly with total assets under management.


        There are two significant caveats with all of these analyses. Firstly, they do not attempt to account for the
        problem of causality. Even if there is a clear correlation between the costs of the fund and the size of the fund,
        this does not determine causality. It is perfectly possible that lower cost funds result in funds attracting net
        inflows, with the outcome that these funds become large. To our knowledge none of these papers account for
        this issue.


        Secondly, it is important to note that all of these aspects relate only to the cost side of the fund, it is possible that
        the performance of the fund could increase or decrease with the size of the fund. This is supported by some
        empirical research which finds that fund returns, both before and after fees and expenses, declines with lagged
        fund size, even after accounting for various performance benchmarks. This association is most pronounced
        among funds that have to invest in small and illiquid stocks, and the research concludes that these adverse
        scale effects are therefore related to liquidity.15


        Hence although there may be considerable cost savings to be made through consolidation, it is not necessarily
        the case that this will lead either to lower prices for consumers or to higher returns for consumers. Nonetheless,
        this issue seems mainly to be limited to specialist areas where there are concerns about liquidity suggesting that
        cost savings would be expected to dominate other concerns given today’s position.




13      Interview with the AMF (Autorité des marchés financiers), March 2006.

14      CONSOB Annual Report 2004, p. 45.

15      Does Fund Size Erode Mutual Fund Performance? The Role of Liquidity and Organization, Chen, Hong, Huang,
        and Kubik, 2004, American Economic Review, 94.



                                                                                                                       Page 27
Methodology for assessing cost savings

September 2006                                                                        CRA International




        We investigated the importance of scale directly through our interviews with fund
        providers, and also indirectly through examining the rationale behind decisions to out-
        source activities or merge funds and by capturing the size of the “representative” fund in
        our quantitative survey for each fund manager.

        The advantage of focusing on the cost structure of particular funds, i.e. the representative
        fund for a particular provider, means that we are able to disentangle efficiency from scale
        effects by comparing the cost of average funds and the average cost of funds of a given
        size. The disadvantage of this approach is that it makes the results fund specific and
        hence generalising the results must be undertaken with care.

        A further complication of understanding the cost structure of investment funds is whether
        we are interested in where the fund is domiciled, where the activities involved in the
        creation of the fund are undertaken or where the fund is sold. All of these issues may be
        interesting for different questions and can vary significantly from fund to fund of the same
        fund manager and between fund managers. For example, in the same market we found
        fund managers that domiciled, created and sold the fund all domestically. We also found
        fund managers, in the same country, who domiciled their funds in Luxembourg, undertook
        the asset management in the UK, and sold the funds in another country.

        To try to capture the most relevant data we focused our quantitative survey on where the
        funds were sold. We also captured data on where the funds were domiciled and whether
        the fund was therefore a cross-border fund, a round-trip fund (only sold in that country but
        domiciled in another member state) or a domestic fund. For each of the activities under
        investigation we asked whether the costs would have varied if the fund had been cross-
        border or not. This strategy was used for all member states other than Ireland and
        Luxembourg, where we focused on funds that were produced in these countries but
        expected to be sold in other markets.


        2.5.      SCENARIOS USED TO ESTIMATE POTENTIAL COST SAVINGS

        In order to estimate the potential cost savings from a fully integrated investment fund
        market we wish to run a series of scenarios. To some extent the scenarios that are most
        useful to consider depends on the question at hand. There are two choices in
        undertaking scenario analysis:

        •      First, it is possible to examine a series of policy alternatives and use the stylised cost
               model to test the benefits that might arise from these policies. For example, we could
               consider the potential benefits that would arise from removing barriers to the cross-
               border consolidation of funds or the potential benefits from having a depositary
               passport. This choice would involve collecting data on the costs of changing
               regulation; or

        •      Second, a series of scenarios can be set out that test the potential benefits that
               could arise from moving from the current position of the market to different
               comparators. Further analysis is then needed to consider the potential for policy
               proposals to lead to these cost savings being realised.




                                                                                                 Page 28
Methodology for assessing cost savings

September 2006                                                                      CRA International




        The European Commission set out that we should focus on quantifying the cost savings
        associated to particular scenarios rather than quantifying the potential impact of policy
        options such as those set out in the Green Paper.

        Secondly, in order for the scenarios to be useful it is important to understand what is
        driving any potential cost savings. In particular, in designing the methodology for
        calculating cost differences, it was important to distinguish between cost savings that
        reflect differences in efficiency, compared to cost savings that reflect differences in scale.

        Figure 4: Comparison of gains from efficiency compared to gains from economies of scale


                   Costs




                                  Efficiency



                                                    Scale




                                                                  Assets under management

        Source: CRA


        Figure 4 illustrates the two types of cost savings. It is possible to imagine that different
        funds have different costs due to inefficiency, i.e. they could be produced cheaper at the
        same size – we describe this as an “efficiency effect”. Alternatively, it is possible that if a
        fund was bigger it would be able to be produced more cheaply, as fixed costs are shared
        over a larger pool of assets, we describe this as a “scale effect”.

        Equally, it was agreed that it was important to differentiate between potential cost saving
        arising on the production side of investment funds from those on the distribution side.

        The scenarios to be tested therefore focus on:

        •    Different comparators: comparing today’s fund to the average European fund, the
             cheapest European fund, a cross-border fund and the US average for the production
             of funds;

        •    Efficiency versus scale: the cost saving of using the most efficient cost function
             versus the scale effect resulting from larger funds.

        The output of the model is a cost saving in basis points and a euro value of the potential
        cost savings based on the member states that have been included in our study. This


                                                                                               Page 29
Methodology for assessing cost savings

September 2006                                                                 CRA International




        figure has been calculated through scaling the basis point estimates of potential cost
        savings according to the market size of the different fund types in each of these member
        states. This will give us the potential cost saving for the 10 member states examined in
        this study. It is also important to note that this headline calculation is based on the
        assumption that there is an immediate and total improvement in costs.




                                                                                         Page 30
Costs of production components of the value chain

September 2006                                                                                CRA International




3.      COSTS OF PRODUCTION COMPONENTS OF THE VALUE
        CHAIN

        In this section, we use the data collected in our quantitative survey, publicly available data
        sets and data from commercial sources in order to examine whether there are significant
        differences in cost at each stage of the value chain for funds of different types and for
        funds sold in different countries.


        3.1.     REGULATION RELATING TO FUND CREATION AND MAINTENANCE

        Each of the activities in the investment fund value chain – fund and client administration,
        asset management, and distribution – is affected by the regulatory environment.
        However, in this section, we focus only on the regulatory costs directly associated with
        fund creation and maintenance (this therefore only represents a lower bound of regulatory
        costs). Regulation that directly affects each of the other activities is discussed in the
        other chapters as relevant. With that caveat in mind, the regulatory costs included in this
        section relate to the following activities:

        •   Authorisation and notification - Authorisation involves those wishing to market UCITS
            funds gaining permission from the competent authority in their home Member State.
            While the UCITS Directives allow for UCITS funds to be marketed in a different
            Member State without requiring authorisation in the host state, they do nonetheless
            require that funds are notified in each Member State where the funds are to be
            marketed. The notification procedure requires that the UCITS fund informs the
            competent authorities of the home and host country of their intention to market the
            fund in another member state and also that the fund provides the host country's
            authority with additional documentation (e.g. simplified prospectus, fund rules, annual
            reports, marketing arrangements etc).16

        •   Fund compliance - There are a number of regulatory requirements that must be met
            for investment funds, including those relating to the UCITS Directives. Hence there
            will be compliance costs incurred. These will include preparation of regulatory reports
            and the performance of in-house monitoring activities (e.g. that investments are made
            according to the investment objectives of the fund and that the corresponding
            thresholds are not violated).

        •   Tax compliance - UCITS funds must meet the requirements of local tax laws which
            will vary in a number of ways including in respect of the rate of withholding tax,
            corporation tax, VAT, stamp duty and capital gains tax.

        •   Documentation production - Most regulators will impose requirements on funds to
            provide documents to their customers regarding the characteristics of these funds.
            These documents must first be designed in order to meet regulatory standards. In




16      A Harmonised, Simplified Approach to UCITS Registration, EFAMA and IMA, April 2005.



                                                                                                        Page 31
Costs of production components of the value chain

September 2006                                                                       CRA International




            particular, UCITS funds will need to design the simplified prospectus which must be
            provided alongside the fund.

        •   Investor protection - Most funds will have some form of consumer protection enabling
            investors to receive some element of compensation in the event of there being a
            deficiency with either the fund or the fund management company. The provision of
            this protection will impose a cost to the fund.

        In order to get funds authorised and notified all fund managers agree that some form of
        local presence is required. However, firms use different models in order to undertake this
        activity, which could include:

        •   A centralised function with a local regulatory presence often provided by external
            lawyers. For example, one fund manager with over €50 billion in assets under
            management has a single person responsible for authorisation and notification
            globally; and / or

        •   A small team locally to facilitate fund notification as well as undertaking many other
            activities.

        Some fund managers interviewed also stated that while domestic authorisation is usually
        handled in-house by the asset management company, fund managers employ law firms
        to assist them with notification procedures abroad, indicating that notification
        requirements in foreign jurisdictions differ sufficiently from domestic jurisdictions to justify
        such expenditure. In reality, firms often use a combination of these models depending on
        the particular country in question and the importance of that country to the firm. The cost
        of the regulatory functions will be recovered through the management charges by the
        fund manager.

        Authorisation and notification

        In terms of measuring the cost of regulation, it is important to look at the different types of
        cost that are incurred. For example, the cost of getting a fund authorised and notified can
        be measured in terms of:

        •   The cost to the fund manager/promoter of seeking approval by the regulator, these
            might be direct costs paid to the regulator or compliance costs from staff time or the
            use of external lawyers and translators to assist with the authorisation or notification
            process;

        •   The cost falling on the regulator, which may not be reflected in the costs to the fund
            manager and charged to the fund; and

        •   The opportunity cost of the time taken to get the investment fund through the
            regulatory authorities.

        Our survey captures the first of these elements. It does not capture the second element
        directly although the costs falling on the regulator will be partly captured through the cost
        of authorisation and notification. However, it does not capture the third element.


                                                                                                Page 32
Costs of production components of the value chain

September 2006                                                                                        CRA International




        Although the degree to which these time costs can vary by types of fund and by member
        state is already meant to be limited to some extent by the UCITS rules, it was
        nonetheless raised as an issue of considerable concern by numerous fund managers in
        many member states.17 Information on the cost of authorisation and notification along
        with further discussion regarding the time taken for these activities is found in section
        3.1.2.

        In the rest of this section we provide information on the total of the cost of regulatory
        compliance of the fund and more detailed information on the cost of authorisation and
        notification.


        3.1.1. Regulatory costs by fund type

        Given the objective of regulating investment funds is to ensure that these represent a
        “safe” investment for retail consumers, it may be reasonable to believe that the cost of
        regulation will vary by fund type. For authorisation and notification, the three areas where
        costs might vary by type of fund were identified as being:

        •    If the cost of authorising or notifying the fund is based on the size of the fund. To the
             extent that money market funds are, on average, significantly bigger than bond funds,
             this would result in them facing a higher total regulatory cost, although the same
             regulatory cost as a proportion of assets under management.

        •    If it was seen as being more complex, for example a derivative based product, this
             might lead to a greater time being required for authorisation. However it was clear
             that this was the case irrespective of the other assets in the fund and that such a
             product would only impose additional costs if it was thought to be particularly complex
             which the fund types under consideration in our study were not.

        •    Whether the fund was part of an umbrella. In some markets, notification is done at the
             umbrella level. In this case all sub-funds will be required to notify, independent of
             whether they will necessarily all be marketed in that country. In other countries,
             notification is only required for sub-funds that will be marketed in the host country. By
             contrast, most member states require notification of all share classes whether or not
             these will marketed to all investors.18

        Based on the evidence from the interviews, there was no explicit difference expected in
        the cost of authorisation or notification across the asset types under consideration. The



17      Article 46 as amended by UCITS III, Directive 2001/107/EC of the European Parliament and of the council
        states that, ”if a UCITS proposes to market its units in a Member State other than in which it is situated, it must
        inform the competent authorities of that other Member State accordingly. It must simultaneously send the latter
        authorities […]. An investment company or a management company may begin to market its units in that other
        Member State two months after such communication, unless the authorities of the Member States concerned
        establish, in a reasoned decision taken before the expiry of that period of two months, that the arrangements
        made for the marketing of units do not comply with the provisions referred to in Article 44(1) and Article 45.”

18      A Harmonised, Simplified Approach to UCITS Registration, EFAMA and IMA, April 2005.



                                                                                                                   Page 33
Costs of production components of the value chain

September 2006                                                                                CRA International




        exception to this was found to be in Belgium, where authorisation of money market funds
        is cheaper than that of other funds.19 However, since authorisation and notification relate
        to only one component of the regulatory costs associated to fund creation and
        maintenance, it is possible that differences might result from some of these other areas.
        The result of examining all of the regulatory costs covered in this section by the different
        types of funds under consideration is shown in Figure 5 below.

        Figure 5: Regulatory compliance costs for different asset types in basis points



                                     2.50




                                     2.00
          Cost of regulation (bps)




                                     1.50




                                     1.00




                                     0.50




                                      -
                                            Money market   Bond   Equity     Index trackers   Capital protected


        Source: CRA analysis of survey results


        The first issue to note when examining these figures is that they are all very low since the
        average cost of regulatory compliance is below 2.5 basis points. Capital protected funds,
        are found to be the most expensive funds from a regulatory compliance perspective and
        have a cost of around 2.3 basis points. Equity funds were found to have the lowest
        regulatory costs out of all of the fund types at 1.3 basis points. In some ways the results
        for equity funds are somewhat unexpected since interview evidence suggested that
        money market and bond funds would be expected to have the lowest regulatory costs,
        although we note that the difference between the cheapest and the most expensive funds
        is only 1 basis point.

        It is possible that differences in the cost of regulation by types of fund could be related to
        the average assets under management for each type of fund. Indeed interview evidence
        indicated that regulatory costs were expected to be fixed costs at the fund level.
        Investigation of the data provided by fund managers provides limited support for this view.




19      Document received from the CBFA March 2006 setting out the regulatory expenses in Belgium.



                                                                                                         Page 34
Costs of production components of the value chain

September 2006                                                                      CRA International




        Given the strong support in the interviews for the position that they are in fact fixed costs,
        we have chosen to model them in this way.


        3.1.2. Regulatory costs by member state

        As well as considering the regulatory compliance costs by fund type, it is also possible to
        examine whether the regulatory costs vary by member state. Evidence based on firm
        behaviour would suggest that the monitoring and management of the tax compliance and
        efficiency of the fund is a particularly significant activity in terms of determining the
        location in which fund managers domicile their fund. This could include reasons such as:

        •   Withholding tax - where the rate that is imposed will vary according to member states
            as will the extent to which it can be offset when incurred in other countries which will
            depend on bilateral tax agreements;

        •   Corporation tax - where no tax is imposed on funds in Dublin or Luxembourg but
            would be in other countries such as the UK (although it may be possible to offset
            withholding tax against this);

        •   VAT - which is not typically imposed on financial services companies. However,
            inputs into the fund management process may incur VAT which then can not be offset
            against VAT charged (since none is charged). However, if fund management
            services are being provided for funds that are domiciled in Luxembourg or Dublin,
            then UK tax authorities provide a tax credit allowing for the offsetting of VAT.

        •   Stamp duty - which is imposed on transactions in the UK, but as a percentage of the
            net asset value of a fund in Luxembourg, but not at all for funds in Dublin.

        •   Capital gains tax - which is imposed on hedge funds in the UK, but not in Dublin.

        Given that many fund managers suggested that their decision to domicile funds in Dublin
        and Luxembourg was at least in part due to tax, this suggests that the costs associated
        with tax and regulatory compliance are not insignificant since they are determining a
        major strategic issue such as the choice of location. However, it seems likely that these
        issues are not fully captured through an examination of pure costs since tax benefits may
        accrue to the fund management company or though performance rather that through a
        reduction in costs.

        In addition, costs may vary across member states because of issues such as the cost of
        labour where both internal and external staff are required to ensure that the fund complies
        with certain rules or to prepare the documents necessary for authorisation or notification.
        Hence, the regulatory cost will be a function both of the amount of time that it takes for
        staff to undertake regulatory activities and also of the cost of those staff.




                                                                                              Page 35
Costs of production components of the value chain

September 2006                                                                                                  CRA International




        Figure 6: Regulatory compliance costs for different countries (€ per fund per year)



                                   80,000

                                   70,000

                                   60,000
          Cost of regulation (€)




                                   50,000

                                   40,000

                                   30,000

                                   20,000

                                   10,000

                                      -
                                                               Germany




                                                                         Italy




                                                                                                           UK




                                                                                                                   Ireland
                                            Belgium




                                                      France




                                                                                 Poland




                                                                                          Spain




                                                                                                                             Luxembourg
                                                                                                  Sweden
        Source: CRA analysis of survey results. Note that there was insufficient data available for Poland to be able to
        accurately quantify the regulatory cost.


        As can be seen from Figure 6 above, in the majority of countries, the fixed cost of
        regulatory compliance is less than €30,000. Germany, however, is a clear outlier with the
        cost of regulatory compliance reaching nearly €70,000. Interview evidence supports the
        view that Germany is a relatively expensive location in which to sell investment funds with
        aspects such as tax compliance highlighted as an area of particular concern by
        interviewees from a number of different member states.

        Funds which are manufactured in Luxembourg and Ireland have relatively low regulatory
        compliance costs (along with France). Again this is supported by discussions with fund
        managers who noted the relative flexibility and speed of the regulatory authorities in both
        of these locations with this one of the factors that feeds into the overall cost of regulatory
        compliance – this is discussed further below.

        Authorisation by member state

        As discussed above, one of the important steps in developing an investment fund is to
        obtain authorisation in the home member state. As is clear from Figure 5 and Figure 6
        above, the overall cost of regulatory compliance of the fund is small. Since authorisation
        and notification is a component of these total costs, it follows that the cost of authorisation
        and notification is also expected to be small. Examining the direct monetary cost of
        obtaining authorisation this is indeed found to be the case, although it does vary
        significantly between member states as highlighted below in Table 6.




                                                                                                                             Page 36
Costs of production components of the value chain

September 2006                                                                                      CRA International




        Table 6: Cost of authorisation

        Member State       Regulator                           Direct cost imposed by regulators

                                                  Single fund                   Umbrella fund             Frequency of
                                                                                                            payment

          Belgium              CBFA           €0.075 per thousand            €0.075 per thousand              Annual
                                                  assets under                   assets under
                                            management plus €0.50          management plus €0.50
                                               per thousand new               per thousand new
                                            subscriptions; for money       subscriptions; for money
                                            market funds the cost is       market funds the cost is
                                             €0.05 per thousand net         €0.05 per thousand net
                                               asset value with a             asset value with a
                                               minimum of €314.               minimum of €314.

          France               AMF           €8 per million of assets       €8 per million of assets          Annual
                                              under management               under management

          Germany              BaFin                 €1,500                  €1,500 per sub-fund           Annual but
                                                                                                          only €500 per
                                                                                                            sub-fund

          Ireland             IFSRA                  €2,050                  €2,050 plus €550 per             Annual
                                                                               sub-fund up to a
                                                                              maximum of €4,800

          Italy               Consob                 €1,400                         €1,400                    Annual

                                BOI

          Luxembourg           CSSF                  €2,650                         €5,000                    Annual

          Poland              KPWiG                  €4,000                 €4,500 for umbrella and          One-off
                                                                              €300 for subsequent
                                                                             additions of sub-funds

          Spain               CNMV             0.014% of estimated            0.014% of estimated          Applied one-
                                            sales with range of €975-      sales with range of €975-       off or as long
                                             39,033 for fixed income        39,033 for fixed income            as the
                                              and €1,626-65,055 for          and €1,626-65,055 for           estimated
                                                 variable income                variable income           sales increase

          Sweden                 FI             SEK 5,000 (€ 540)             SEK 5,000 (€ 540)               Annual

          UK                   FSA               £1,200 (€1,735)                £2,400 (€3,470)             Annual fee
                                                                                                              varies
                                                                                                           according to
                                                                                                          the number of
                                                                                                          sub-funds and
                                                                                                            varies from
                                                                                                           £620-13,640
                                                                                                          (€900-19,720)

        Source: Interviews with, and data from, regulators. Where fees are noted as annual then annual payments are
        the same as the initial authorisation costs unless otherwise specified. In Ireland, an annual payment rather than
        an authorisation fee per se is payable. Note that interview evidence with the Spanish regulator suggested that
        the basis of fees may change to become a fixed amount in the future although this is not currently the position.
        In the UK, the annual fee depends on the number of funds/sub-funds in each category: 1-2 £620; 3-6 £1,550; 7-
        15 £3,100; 16-50 £6,820; >50 £13,640.




                                                                                                                Page 37
Costs of production components of the value chain

September 2006                                                                      CRA International




        The structure of these costs means that they will have a differential impact on different
        types of funds because the average assets under management varies by fund type. The
        impact of this will also vary between member states depending on the cost structures that
        are used in the different countries. For example:

        •   Where there is a fixed authorisation fee (i.e. Italy and Luxembourg) this will favour
            large funds in those member states and hence this will tend to favour money market
            funds compared to equity funds;

        •   Where fees are applied as a percentage of assets under management (such as
            Belgium and France) this would favour fund types with smaller assets under
            management (although Belgium has a lower rate for money market funds than other
            fund types);

        •   Where fees as a percentage of estimated sales (Spain) this is more favourable to
            those funds with high growth rates than the other approaches.

        Furthermore, although the direct costs associated with fund authorisation and notification
        are small, they may also be indicative of the different objectives of different regulators. In
        some countries there is a clear objective to share regulatory costs with the industry. For
        example, in Ireland, 50% of regulatory costs are paid for by the industry and 50% from
        central government. In Italy, the Bank of Italy does not make any particular charge to the
        industry but this is covered from central financing. In the UK and France the regulator is
        entirely paid for by the industry.

        The different objectives by regulators may also be reflected in the basis for the fee
        structure. For example, whether regulatory fees are based on assets under management
        or based on sales will have different implications for newly launched funds or less
        successful funds. To the extent that these regulatory fees are significant (which they do
        not appear to be), this could contribute to a higher number of funds being launched in
        some countries than others.

        Notification by member state

        The management company may begin to market its units in a Member State two months
        after information has been communicated to the host regulator unless the regulator
        objects. The ability of the regulator to refuse notification should be limited to compliance
        with domestic marketing rules. Notification of this under UCITS is from the fund to the
        host state regulator which contrasts with requirements in other Directives which operate
        from home state regulator to host state regulator.

        As well as the cost of authorisation, if funds are to be sold in other member states then
        they must notify these to the competent authority in the host member state. These fees
        are provided in Table 7 below and it is clear that the cost of notification is frequently the
        same as the cost of authorisation seen in Table 6. This is surprising since the host
        regulator would be expected to be undertaking less activity than the home regulator who
        has already authorised the fund. The significant exceptions to this are Ireland, where no
        fees are charged on those funds that are notified, and France, where the cost of



                                                                                              Page 38
Costs of production components of the value chain

September 2006                                                                                        CRA International




        authorisation is based on the value of the fund whereas the cost of notification is a fixed
        fee.

        Table 7: Cost of notification

        Member State         Regulator                           Direct cost imposed by regulators

                                                   Single fund                  Umbrella fund              Frequency of
                                                                                                             payment

          Belgium                CBFA          €0.075 per thousand           €0.075 per thousand              Annual
                                                   assets under                  assets under
                                             management plus €0.50         management plus €0.50
                                                per thousand new              per thousand new
                                             subscriptions; for money      subscriptions; for money
                                             market funds the cost is      market funds the cost is
                                              €0.05 per thousand net        €0.05 per thousand net
                                                asset value with a            asset value with a
                                                minimum of €314.              minimum of €314.

          France                 AMF                 €1,000                  €1,000 per sub-fund              Annual

          Germany                BaFin               €1,500                  €1,500 per sub-fund          Annual but only
                                                                                                           €500 per sub-
                                                                                                               fund

          Ireland               IFSRA         No fees are charged           No fees are charged            No fees are
                                                                                                            charged

          Italy                 Consob               €1,400                        €1,400                     Annual

                                  BOI

          Luxembourg             CSSF                €2,650                        €5,000                 Annual – part 1
                                                                                                            fund incurs
                                                                                                           annual cost of
                                                                                                           €2,650; part 2
                                                                                                          fund of €3,950;
                                                                                                          and umbrella of
                                                                                                              €5,000

          Poland                KPWiG                €4,000                €4,500 for umbrella no            One-off
                                                                             additional costs for
                                                                           additions of sub-funds

          Spain                 CNMV          0.014% of estimated           0.014% of estimated           Applied one-off
                                                sales with range of           sales with range of           or as long as
                                              €975-39,033 for fixed         €975-39,033 for fixed          the estimated
                                               income and €1,626-            income and €1,626-           sales increase
                                                65,055 for variable           65,055 for variable
                                                      income                        income

          Sweden                   FI          SEK 5,000 (€ 540)             SEK 5,000 (€ 540)                Annual

          UK                      FSA              £600 (€870)                 £1,200 (€1,735)               Annual fee
                                                                                                         varies according
                                                                                                         to the number of
                                                                                                          sub-funds and
                                                                                                            varies from
                                                                                                           £620-13,640
                                                                                                          (€900-19,720)

        Source: Interviews with, and data from, regulators and fund managers. Note that interview evidence with the
        Spanish regulator suggested that the basis of fees may change to become a fixed amount in the future although
        this is not currently the position



                                                                                                                 Page 39
Costs of production components of the value chain

September 2006                                                                                CRA International




        As well as the direct regulatory costs, the process of notification also imposes costs
        regarding the provision of information to the host authorities. Given the restrictions in
        UCITS regarding the notification process, it would be expected that these additional costs
        would be similar across member states. However, according interview evidence and a
        recent report by EFAMA / IMA, many countries impose additional information obligations
        to those required by the UCITS Directive.20

        This results not only in delays to obtain the notification but also in additional costs such as
        the use of local legal support and translation of marketing documents. In some cases
        there is a need for local representation such as in Germany which requires a local paying
        and information agent and a letter from the German paying and information agent
        confirming their role. Further, many countries also require a special addendum including:
        Austria; Belgium; France; Greece; Luxembourg; the Netherlands; and Sweden.

        In addition, three of the largest markets also impose requirements for new sub-fund
        notifications:

        •   France requires the certification of accuracy of information contained in the
            prospectus to be signed by two directors;

        •   Germany requires letters from paying agents and information agents confirming roles
            as well as proof of payment of initial registration fee; and

        •   Spain requires a letter to the Spanish regulator signed by a director confirming
            documentation complies with the Spanish regulators requirements.

        In fact, according to the EFAMA / IMA report, only the UK and Norway, of the EEA
        countries surveyed, imposed no additional requirements to those under the UCITS
        Directives.

        As well as these additional requirements, and of more consequence to the fund
        managers, are the delays in terms of the additional time to market their funds in particular
        countries. In as far as this creates a delay in the time to market of new investment funds
        this reduces competitive pressures, this is likely to have a negative effect on costs and
        the prices that consumers face.

        In the table below we include the estimated cost in terms of time of going through the
        regulatory authorities. There is a limit on the time taken for notification under the UCITS
        requirements, however, in some member states, interviewees argued that this results in a
        minimum time for the regulator to respond with a request for more information but did not
        in fact guarantee the length of the notification process.




20      A Harmonised, Simplified Approach to UCITS Registration, EFAMA and IMA, April 2005.



                                                                                                        Page 40
Costs of production components of the value chain

September 2006                                                                                 CRA International




        Table 8: Time taken to get regulatory approval

                   Member state                     Fund authorisation                   Fund notification

         Belgium                                          1-2 weeks                            2 months

         France                                      4 weeks on average             8 weeks but occasionally up to 6
                                                                                                months

         Germany                               3 weeks (the time has improved                  8 weeks
                                               considerably in the past years)

         Ireland                                    6-8 weeks on average                       2 weeks

         Italy                                 Consob reviews prospectus in 20       2 months but up to 16 weeks
                                               days, Bank of Italy review of fund
                                                        rules 40 days

         Luxembourg                            2-8 weeks depending on the file                 8 weeks

         Poland                                           5-6 months                           2 months

         Spain                                 Typically one month, but up to 2     Typically one month, but up to 2
                                                           months                               months

         Sweden                                3 months, 6 months for complex       4 weeks sometimes but usually 2
                                                            fund                               months

         UK                                                6 weeks                             2 months

        Source: Interviews with fund managers and regulators.


        As well as the general concern about regulatory processes delaying the time to market,
        for some products, the timeliness of launch is vital to the success of the product. For
        example guaranteed products would have a window of opportunity during which the
        assets can be gathered before the guaranteed period begins. Other products may aim to
        rapidly take advantage of trends that are being observed in the marketplace.

        In addition, interviewees noted cases where the regulatory cost in terms of the delays to
        notification has resulted in them being unable to launch their products and where
        competitors have launched successful variants of the same product in the interim.
        Assuming that this is not a form of protectionism in which domestic competitors are able
        to sell products while foreign competitors are delayed in the notification process, the cost
        of this delay for consumers could be relatively small. That is, consumers may still have
        the opportunity to invest in a particular type of fund, but individual competitors may be
        frustrated by being unable to access the market. Nonetheless, there is a cost in terms of
        dynamic competition with the impact of competitive forces being reduced.

        It should be noted that although some countries have a one-off authorisation and
        notification, regulatory costs are not one-off. Over time changes made to the fund will
        need to be authorised by the regulator and require further expense (Irish interviewees
        suggested that this could happen up to three times a year) and most member states apply
        fees on an ongoing basis. Even taking this into account, the combined cost of both
        authorisation and notification in terms of internal resources and use of external lawyers,




                                                                                                           Page 41
Costs of production components of the value chain

September 2006                                                                                     CRA International




        often accounts for less than 0.25 bps for the largest providers (although it is usually more
        than this for smaller providers).21

        Overall, regulatory costs have been found to be a small component of the total costs of
        investment funds with an average cost of between 1.3 and 2.3 basis points depending on
        the type of fund under consideration. Despite this small figure, regulatory processes were
        nonetheless one of the key areas where fund managers have expressed most concern.
        In particular, fund managers were concerned about the differences between member
        states in respect of notification and saw this as one of the areas where potential exists for
        cost savings to be easily and rapidly made. Many fund managers have argued that the
        notification process is a purely duplicative cost that brings no benefit and that these costs
        could be stripped out of the value chain with no detrimental effect.

        It is estimated that the authorisation and notification processes represent around half of
        the regulatory cost and therefore less than 1.2 basis points. Given this small figure, it
        may be surprising that fund managers are so concerned about this. One reason for their
        concern is that regulatory processes are one of the areas of fund management costs that
        companies are unable to control themselves and hence an area where internal cost
        efficiencies can make only limited difference. Furthermore, companies may observe this
        element purely as a drain on resources where other cost elements such as distribution
        might be viewed as an investment to attract assets. A final frustration was that it a formed
        a element of protectionism. Some interviewees argued that member states could use a
        lax regulatory assessment on authorisation to attract the location of funds, while others
        could impose a long process on notification to benefit the home market.

        It was clear from discussions that all fund management companies were very keen to see
        additional harmonisation of the interpretation of the UCITS directives by the different
        regulatory authorities in each of the member states. They believed that such an action
        would lead to a reduction of costs incurred in the notification process when operating
        cross-border funds. Indeed, some fund managers noted that given that the notification
        procedure constitutes neither a re-approval nor a re-authorisation process, it is surprising
        that the time take for notification is often longer than the time taken to authorise a fund.
        At the time of interviewing (March 2006), there was nearly universal support for the CESR
        discussions aimed at improving notification through a simplified procedure. However, the
        resulting second consultation paper was not seen by market participants as necessarily
        resolving the issues.22

        However, although the differences in regulatory cost and the time taken to obtain
        regulatory approval to market funds were seen by fund managers as a nuisance, a
        reduction in costs here was not, in fact, seen by fund managers as likely to cause a




21      Interview with Irish fund manager, March 2006. This is broadly consistent with the statistic presented in the
        EFAMA and IMA’s “A Harmonised Simplified Approach to UCITS Registration” April 2005, which estimated that
        pan European providers have cost of approximately €0.5 million per year to maintain registration.

22      CESR’s guidelines to simplify the notification procedure of UCITS, 2nd Consultation Paper, May 2006.



                                                                                                               Page 42
Costs of production components of the value chain

September 2006                                                                    CRA International




        significant change in terms of the cost of bringing funds to market and hence the price
        charged to retail investors.


        3.2.     FUND ADMINISTRATION

        In the next two sections we focus on the administrative activities associated with the
        provision of investment funds. It is useful to differentiate between administration costs
        associated with the fund itself (which we refer to as fund administration), examples of
        these would include fund accounting or auditing and administration costs associated with
        the clients of the fund (referred to as client administration), which we look at in the next
        section. Fund administration is made up of several significant components:

        •   Fund accounting covers a number of activities but the most significant is the
            calculation of the net asset value (NAV) of the fund. This is typically done on a daily
            basis but this varies by the type of fund and there are some funds for which it might
            be done less frequently;

        •   Auditing, much the same as companies, funds must have their accounts checked. All
            funds must have an audit conducted by external auditors; and

        •   Safe custody involves taking physical custody of investment assets and the
            maintenance of accounts with central securities depositories (CSDs). Depositary
            oversight is the fiduciary function of supervising the actions of the management
            company. It includes ensuring that transactions (sale, issue, repurchase, cancellation
            of units, changes in assets, payments to and from the fund, the application of income)
            and the calculation of the value of units are carried out in accordance with the
            regulations and fund documentation. In addition the custodian must scrutinise the
            conduct of the management company in each accounting period and inspect the
            procedures and controls employed by the transfer agent. Although safe custody and
            depositary oversight are different activities, we have not been able to robustly
            differentiate between these two elements and hence cost information is presented for
            the two activities combined.

        As set out in the detailed description of the value chain, fund administration also includes
        performance measurement, stocklending and fund order processing (which relates to
        execution of fund orders and the settlement process and which typically occurs in
        different ways in the different member states in the EU). The information collected for this
        project has not allowed us to separately investigate these areas. However, fund order
        processing is discussed in section 6.2 when we consider existing barriers to integration.

        There is significant variation regarding whether these administrative activities are
        undertaken by the fund manager or outsourced and how these are paid for. Whether
        these activities are undertaken in house or externally, it is common to charge most of the
        components of administration to the fund. However, in some countries, elements must be
        paid directly by the fund managers, for example, custody costs in Sweden.

        The location of administrative tasks and the types of company who can provide the
        services is often in part determined by regulation. The UCITS directive requires that the



                                                                                            Page 43
Costs of production components of the value chain

September 2006                                                                                    CRA International




        management company and the depositary are located in the same member state (to be
        more specific it must either have its registered office in the member state or be
        established there), although it is possible to use delegation and sub-custodians. In the
        case of Irish or UK Unit Trusts a trustee effectively performs the role of a depositary.
        Member states are free, however, to determine the types of organisation that can be
        depositaries.

        Allocating costs between different fund and client administrative activities is complicated
        as the organisation undertaking these roles varies significantly from country to country. In
        addition to variation in the degree to which the administration is outsourced, there are
        also considerable differences in what each of these activities entails. Hence care must be
        taken when interpreting the different elements of fund and client administration since
        apparent differences in cost may reflect a different way of structuring the cost
        components. Although the value chain was sent to those participating in the cost survey
        such that they could identify the different functions for which cost information was being
        gathered, making like for like comparisons in complicated by:

        •    Variation in the activities undertaken by different organisations. For example, the
             depositary in Germany is responsible primarily for the accounting function but not the
             administrative functions. In some countries the depositary would calculate the NAV,
             whereas in other, Italy for example, the depositary is responsible for the safekeeping
             of the fund’s assets and is also required to control the operation of the fund in a
             thorough and extensive manner but has not been responsible for the NAV.23

        •    That in the same market some fund managers will undertake fund accounting and
             transfer agency and use a depositary of their parent bank, whilst other fund managers
             outsource the transfer agency to one organisation, the fund administration to another
             organisation and the custody to a third and hence cost data will be a mixture of in-
             house and external provision.

        •    That even in the same function the costs may vary significantly depending on how the
             activity is organised. For example, if omnibus accounts are used when outsourcing
             the transfer agency function the cost of transfer agency services is significantly
             reduced. However, the cost of keeping records for the ultimate client is simply
             accounted for elsewhere with these costs then likely to be incurred by the distributor.

        Of all of the elements in this part of the value chain, those for which detailed quantitative
        information was available included: fund accounting; auditing; and custody.24 No cost



23      See Banca Intesa’s comments on the Commission’s green paper on the enforcement of the EU framework for
        investment funds. According to Banca Intesa, a corollary of the legal framework would be that only safekeeping
        and custody of assets can be outsourced; in contrast, the control function could never be outsourced.

24      Looking at the survey undertaken for this project, very few fund managers are able to quantify the costs of
        undertaking stock lending or performance measurement.         The only estimate provided suggests that stock
        lending represents 5% of total administration costs. Assuming total administration costs in the order of 30bps,
        this would equate to 1.5bps. Only two fund managers identified significant performance measurement costs.
        These were thought to be less that 3% of admin costs, so are smaller still.



                                                                                                                Page 44
Costs of production components of the value chain

September 2006                                                                     CRA International




        information is available on the other components of the value chain (performance
        measurement, stocklending), however, discussions with fund managers suggest that the
        elements covered represent the largest and most important aspects of the fund
        administration component of the value chain.


        3.2.1. Fund administration by type of fund

        In this section we look at the evidence for whether each of these administrative activities
        varies by the type of fund and then by member state.

        Fund accounting

        The costs of fund accounting depend on the asset classes that the fund is invested in:25

        •    Equities are relatively easy to administer, although there might be a dividend to deal
             with (which would often have to be manually processed) and some corporate actions
             or risks as well as stock splits that need to be taken into account.

        •    Bonds accrue interest every day and this might have to be discounted or amortised.
             There was disagreement among fund service providers as to whether this led them
             to be more or less complicated than equity funds as many of these elements can be
             programmed into a standard calculation. Defaults are also a possibility, which may
             be a problem because bonds imply a legal obligation.

        •    Money market funds have interest every day but it is unlikely that there would be a
             default problem as they are short term, low risk instruments. The same calculation
             has to be done every day but it is a very standard calculation.

        •    Index trackers are easier to account for than equities unless derivatives are used in
             which case it becomes more complex as some of these derivatives are over-the-
             counter (OTC) products. This means that they need to use special pricing sources
             and have additional technology needs in order to identify the appropriate price for the
             NAV calculation.

        •    Capital guaranteed products need derivative components to put a floor on the asset
             management. There is a heavier compliance duty but the administration is not more
             expensive per se.

        Looking specifically at the calculation of the net asset value, the frequency of the NAV
        calculations is one of the most significant drivers of the cost of fund accounting.26 This is
        assumed to be daily for all retail funds, although some funds exist where the NAV is
        calculated less frequently (e.g. monthly or weekly). Guaranteed funds were also seen to
        be more complex in terms of the NAV calculation.




25      Interviews with various fund managers and service providers, March 2006.

26      Interviews with fund service providers, March 2006.



                                                                                             Page 45
Costs of production components of the value chain

September 2006                                                                                                    CRA International




        The cost of calculating the NAV also varies depending on the complexity of the fund.
        There can be multiple share classes in a fund each with different NAVs. This requires
        different accounting for each register for each share class and hence would also be more
        expensive.

        In general, although accounting services are priced according to asset classes, for large
        funds the main price determinant is the overall size of the portfolio. Furthermore, as the
        assets of the overall portfolio increases, the client receives the benefits of scale
        economies and the difference between asset class decreases since they are likely to
        have a mix of asset types in the portfolio.

        There is common agreement that this is an area where there are significant economies of
        scale. Some fund service providers suggested that the total fund accounting costs for a
        large fund (around €250 million) would be around 25% of that for a medium fund of €70
        million. While a small fund of around €10 million could have fund accounting costs 20
        times larger.27

        Figure 7: Fund accounting costs for different fund types


                                                       Equity     Bond      Capital protected   Money market   Index tracker
                                          60
          Cost fo fund accounting (bps)




                                          50


                                          40


                                          30


                                          20


                                          10


                                           0
                                               0-5 5-10 10-     15-   20-   30-   40- 50- 100- 150- 200- 300- 400- 500- 600- 700
                                                        15      20    30    40    50 100 150 200 300 400 500 600 700 +
                                                                                  Size of fund (€m)


        Source: CRA analysis of Fitzovia data.28




27      Interviews with fund service providers, March 2006.

28      In the data we only observe the cost costs exhibiting economies of scale. In reality, we would not expect this to
        continue and economies of scale are likely to be exhausted at some point, with diseconomies of scale
        potentially being introduced. It is has not been possible to specify the point at which this arises however and this
        is a weakness of the modelling approach.



                                                                                                                               Page 46
Costs of production components of the value chain

September 2006                                                                                       CRA International




        The analysis presented in Figure 7 shows clearly that fund accounting, as expected,
        varies significantly by fund type and demonstrates clear economies of scale (although the
        data suggests these are may be exhausted earlier than suggested in the interviews). The
        figures on index trackers should be interpreted with some caution since the sample size
        for some of the observations is reasonably small which explains the large variation
        observed in the figure.29

        Custody

        Depositaries have a statutory monitoring and control function which refers to checking
        asset allocation, settlement of transactions, stock-lending, borrowing by the management
        company. They are also in charge of collecting fund income and handling corporate
        actions. To understand the cost structure of custody it is useful to look at how it is
        remunerated. Custody services are provided with the payment schedule set out in a
        ratecard negotiated with the fund manager. Each country and each fund manager will
        have their own ratecard, this will include a holding fee based on holding asset value and
        transaction fee, mostly in terms of basis points charged directly to the fund for particular
        activities:

        •    Custody of equity transactions will depend on the location of the underlying
             investments where the sub-custodian costs might be something around:

                    0.25 bps in the US;

                    1 bps for assets held by Euroclear / Clearstream;

                    2-3 bps for Southern European securities;

                    5 bps in Greece; and

                    10 bps for assets in Brazil.

        •    There is also thought to be considerable variation in Europe based on the use of
             technology and the degree to which faxes is still used as the primary method of
             communication. To authorise a transaction you need a fax with a signed document
             because of money laundering and this is a costly manual labour. This explains why
             transactions in the EU are cheaper if the custodian has direct relationship with
             Euroclear or Clearstream.

        The cost of custody will therefore vary depending on:

        •    The degree to which the fund is trading - actively managed funds will have higher
             custody costs than passive managed funds;

        •    The types of assets that are traded; and



29      In the later modelling, econometric equations are calculated from this data to estimate the cost at different sizes
        of assets and under different scenarios, this also has the effect of smoothing the data.



                                                                                                                  Page 47
Costs of production components of the value chain

September 2006                                                                                             CRA International




        •                       The size of a fund. Comparing the custody costs of a single fund compared to two
                                funds half the size, there will be a clear saving as instead of two trades, there is only
                                one with a single settlement process.

        Figure 8: Custody costs for different fund types


                                                Equity     Bond      Capital protected   Money market   Index tracker
                                    70

                                    60
            Cost of custody (bps)




                                    50

                                    40

                                    30

                                    20

                                    10

                                     0
                                         0-5 5-10 10-    15-   20-   30-   40- 50- 100- 150- 200- 300- 400- 500- 600- 700
                                                  15     20    30    40    50 100 150 200 300 400 500 600 700 +
                                                                           Size of fund (€m)


        Source: CRA analysis of Fitzovia data.


        As expected, we find that the cost of custody is lowest for money market funds at an
        average of 6bps. The next cheapest are passive equity funds and capital protected.
        Bond and equity funds are the most expensive on average.

        External audit

        Turning to external audit, all member states require an external audit. Based on our
        interviews the large accounting firms compete in all member states and fund managers
        also have the choice of going to smaller domestic competitors.

        To look at the cost of external audit we can rely on the amount charged directly to the
        fund. Comparing these costs across fund types it is clear that there are strong economies
        of scale but that the costs do not vary significantly by type of fund.




                                                                                                                        Page 48
Costs of production components of the value chain

September 2006                                                                                              CRA International




        Figure 9: Audit costs by type of fund


                                              Equity        Bond         Money Market         Protection      Index tracker

                                18

                                16

                                14

                                12
            Audit costs (bps)




                                10

                                8

                                6

                                4

                                2

                                0
                                     0-5   5-10 10-15 15-20 20-30 30-40 40-50 50-   100- 150- 200- 300- 400- 500- 600- 700 +
                                                                              100   150 200 300 400 500 600 700



        Source: CRA analysis based on Fitzrovia data


        The only fund type that appears to have systematically lower audit costs are index
        tracking equity funds. Therefore, when comparing different fund types we do not allow for
        differences in audit cost.


        3.2.2. Fund administration by member state

        Fund accounting

        In addition to the cost of fund accounting varying by the type of fund, the cost of
        undertaking the calculation of the NAV varies from state to state. One reason put forward
        in the interviews for these differences is the responsibility for errors in the calculation of
        the NAV varies between jurisdictions. For example in:

        •                       Italy: an error of 0.1% in the NAV requires the errors to be communicated to
                                shareholders and for shareholders of the fund to be recompensed; and

        •                       Luxembourg: an error less than 1% of the NAV of equity funds does not need to be
                                reported to shareholders, otherwise the threshold is 0.25% of the NAV.30

        The larger error band permissible from domiciling the fund in countries such as
        Luxembourg is seen as one of the justifications for locating there.




30      Interview with Assogestioni, March 2006.



                                                                                                                       Page 49
Costs of production components of the value chain

September 2006                                                                                                           CRA International




        Figure 10: Fund accounting costs for different member states


                                            Belgium     Ireland     France      Germany       Italy     Luxembourg       Sweden       UK
                                            120
            Cost fo fund accounting (bps)




                                            100


                                            80


                                            60


                                            40


                                            20


                                              0
                                                  0-5 5-10 10-    15-   20-   30-   40- 50- 100- 150- 200- 300- 400- 500- 600- 700
                                                           15     20    30    40    50 100 150 200 300 400 500 600 700 +
                                                                                    Size of fund (€m)


        Source: CRA analysis of Fitzovia data.


        After taking into account the types of fund and the size of the fund, we do not find
        significant differences in the cost of fund accounting by country.

        Custody

        The market for custodians has been developed over recent years through the role of
        global custodians (including large investment banks, securities houses and trust
        companies). There is a general consensus that over time the market will concentrate on
        the top five global players.

                                              “Global custodian is a service whereby a single custodian assumes responsibility for the
                                              custody of a client’s portfolio of international securities and cast – i.e. the custody service
                                              extends beyond the custodian’s and client’s home base.” 31

        This would suggest only small differences between member states, however, custodians
        for pan-European asset managers will usually have small teams working locally in
        multiple jurisdictions. While the processing will often be centralised in a single member
        state (or even outside the EU), the client relations team is placed locally (complying with
        the UCITS rules). In fact, we find considerable differences in the importance of global
        custodians and, more generally the role of the custodian across member states. The
        differences include:

        •                               The types of organisation who can be a depositary;



31      OXERA “The Role of Custody in European Asset Management” A report for EAMA (November 2002).



                                                                                                                                    Page 50
Costs of production components of the value chain

September 2006                                                                                CRA International




        •   The regulatory requirements imposed on depositaries also vary considerably
            regarding for example, whether there are additional capital requirements. Here only
            France appears to imposes a specific capital requirement for the provision of custody
            services.32 In other countries, they are only subject to general requirements; and33

        •   The cost imposed on custodians by the regulatory authorities. For example, in some
            countries the custodians will also have responsibility for the oversight of Net Asset
            Value (NAV) calculation and for any inaccuracies in the way that it is calculated.

        The type of organisation undertaking these services varies significantly. Custodians have
        traditionally been banks. In a number of member states depositaries have to be banks
        licensed to conduct deposit taking business, such as in Germany, Italy and Luxembourg.
        In many countries, such as France, Spain and Sweden, however, the chosen depositary
        often belongs to the same management group (such as being the parent bank). Although
        these must also be a separate legal entity and act independently they may be related to
        the investment company through ownership linkages. Only in Ireland and the UK (where
        securities brokers and investment managers can act as custodians) do depositaries
        /trustees have to be outside the group of the management company.

        To complicate matters, even though the custodian has to be in the member state where
        the fund is domiciled, sub-custodians can be used. For example, in Luxembourg,
        regulation requires a UCITS underlying portfolio of securities to be held directly by its
        appointed custodian in Luxembourg although it can then be delegated to a sub-custodian
        domiciled in the local market. Luxembourg requires sub-custodians to be appointed via a
        direct contractual relationship between the appointed custodian and the local agent.

        Figure 11 below shows the cost of custody across the ten countries under examination:




32      OXERA “The Role of Custody in European Asset Management” A report for EAMA (November 2002).

33      In Germany depositary banks must have minimum capital of €5m. In the UK, the own-funds requirement is £4m
        or about €6.5m. In Luxembourg, it is €8.5m. France this is €3.6m. In Italy they are subject to €100m. OXERA
        “The Role of Custody in European Asset Management” A report for EAMA (November 2002).



                                                                                                          Page 51
Costs of production components of the value chain

September 2006                                                                                                        CRA International




        Figure 11: Comparison of the cost of custody by country of domicile


                                    Belgium   France   Germany    Ireland   Italy   Luxembourg    Poland    Spain    Sweden   UK
                                    30


                                    25
            Cost of custody (bps)




                                    20



                                    15


                                    10


                                     5


                                     0
                                              Equity             Bond          Money market      Capital Protected   Index Tracker


        Source: CRA analysis of Fitzrovia data and survey data


        We find that there is considerable variation between countries, with the cheapest member
        states by type of fund:

        •                           Equity - Sweden and the UK;

        •                           Bond – Sweden and France;

        •                           Money market – Belgium and Sweden;

        •                           Capital protected – Belgium and France; and

        •                           Index tracking equities – UK.

        Although there is general agreement that the cost of custody services varies around
        10bps for most funds, there is also clear evidence of economies of scale. Therefore to
        make meaningful comparisons across countries it is necessary to account for the size of
        the average fund.

        In order to assess the extent of economies of scale and cost differences in the custody of
        European-focused equity funds in different member states, we have carried out a simple
        econometric analysis where custody fees paid by each fund (expressed in basis points –
        bps) are regressed against fund size and several country dummies. The results of this
        are presented in Annex A.

        This shows that there are sizeable differences in custody fees across countries, even
        after controlling for fund size. For example, compared to Belgium, custody fees are lower
        in France, the UK, and Sweden (although the coefficient for Sweden is not significant)


                                                                                                                                   Page 52
Costs of production components of the value chain

September 2006                                                                                             CRA International




        and higher in Germany, Spain and Italy. This is useful to identify differences between
        countries at the overall level of the costs of custody.34

        External audit

        Fund managers agreed that audit fees represent a fixed cost for the fund and would be
        very similar between different fund types. (Since the cost of audit is seen as a fixed cost
        per fund irrespective of the size of the fund, it is presented in euros in Figure 12 below.)
        When comparing countries as indicated in the figure below, Luxembourg and Germany
        are found to be more expensive than other countries.

        Figure 12: Comparison of audit costs by country (€ per fund per year)


                                      14,000



                                      12,000



                                      10,000
          Average cost of audit (€)




                                       8,000



                                       6,000



                                       4,000



                                       2,000



                                           -
                                               France   Germany   Ireland   Italy   Luxembourg   Poland   Spain     UK



        Source: CRA analysis based on Fitzrovia data and survey data. Figures for Belgium and Sweden were not
        available and hence in later analysis the average of the other countries (€8,162) is used as the cost of audit in
        these countries.35


        In Germany, it was noted by BaFin that they require audit firms to act as a control on the
        reports of the German funds which may partly explain the higher cost observed in
        Germany than some of the other countries.36 In addition, German taxation arrangements




34      However, given the differences in the sizes of funds that are prevalent in different countries, the coefficients on
        the country dummies should not be used as the estimate of the actual difference in custody costs between the
        member states.

35      For the purpose of comparison with other figures, the basis point equivalent is as follows: Belgium 1.7, France
        0.7, Germany 0.5, Ireland 0.9, Italy 0.6, Luxembourg 1, Poland 1.4, Spain 0.8, Sweden 0.5, UK 0.3.

36      Interview with the BaFin, March 2006.



                                                                                                                     Page 53
Costs of production components of the value chain

September 2006                                                                      CRA International




        were believed to be particularly complex and this may add to the audit costs as well.
        Other differences in cost may simply reflect the cost of labour differing between member
        states.


        3.3.     CLIENT ADMINISTRATION

        Turning to administration costs associated with the clients of the fund the most significant
        elements are:

        •   Transfer agency involves the maintenance of records of shareholders’ accounts and
            transactions, disbursing and receiving funds from shareholder transactions, the
            preparation and distribution of account statements and tax information, handling
            shareholder communication, and providing shareholder transaction services.
            Reclaiming tax refunds also falls in this sub-function.

        •   Administration of shares arises when a client wishes to buy or sell units, this results in
            various transactions occur leading to associated administration including the issuing
            of contract notes, opening accounts for clients and sending them information
            regarding this.

        •   Shareholder services resulting from client decisions regarding their investments such
            as the reinvestment of dividends, payment of dividends to a separate account, etc.
            The cost of these services is likely to vary depending on whether units are
            accumulation units in which there is an automatic reinvestment of any income earned,
            or whether they are distribution units in which the income earned is distributed to unit
            holders.

        These services are billed to the fund under a variety of arrangements, but average
        account size is, in most cases, the most important determinant of transfer agent billings.

        In the data collected for this assignment, there are significant differences in the
        importance given to these three elements of client administration. This appears to reflect
        how costs have been allocated rather than meaningful differences useful for our analysis.
        Therefore we have focused on client administration as a whole.


        3.3.1. Client administration by type of fund

        Many fund managers and service providers indicated that client administration activities
        represented relatively low value work because of the limited added value services
        possible in this part of the production of investment funds. Generally, transfer agency
        activities were seen as activities that had to be conducted in order to undertake business
        rather than an area of the value chain that could result in any advantage to a particular
        fund manager.




                                                                                              Page 54
Costs of production components of the value chain

September 2006                                                                      CRA International




        Transfer agency activities are also believed to be very IT intensive with substantial
        investments in technology required in order to offer good services and to deal with
        scanning, work-flows and trailer fees.37 However, because of the various aspects of the
        role including different tax filing requirements, it is difficult to achieve economies of scale
        across borders as there are different needs for customers in different locations. It is
        argued that this is preventing this area from being one where cost efficiencies can be
        exploited. Differences caused by time zones leading to there being different market
        closures in different countries also constrain the transfer agency role when dealing with a
        cross-border fund.

        Transfer agents charge according to their activity. They will typically have a charge per
        transaction in terms of account maintenance fees, system support fees, fees for tax
        purposes. The cost therefore depends on investor activity – i.e. new subscription and
        redemptions.

        Evidence from the interviews supports the case that costs do not vary according to the
        type of fund per se i.e. whether the fund is focused on particular securities, but rather the
        costs would vary according to factors such as whether or not there is a nominee account
        that is used for client records.

        Costs of transfer agency will vary by fund because of investor activity. Only if activity
        varies by type of fund will there be substantial costs. For, example if equity funds are
        more likely to have people go in and out, the transfer agency cost will be higher.

        Overall we expect transfer agency costs to vary by the location of the activity but not to
        vary by the type of fund (at least after allowing for scale).


        3.3.2. Client administration by member state

        The cost of client administration vary significantly across countries because of:

        •    Omnibus or nominee accounts: In general, transfer agency costs are higher if
             individual names are kept on a register compared to only keeping the name of the
             global owner (e.g. when the bank keeps a global account with the TA). Hence, where
             the costs lie depend on the distribution model used. In models where there are many
             distributors and it is important to maintain the record for the purposes of rewarding
             the distributor, we would expect transfer agency costs to be consequently higher.

        •    Technology is another important factor. For example, in Luxembourg transfer agency
             could cost €20-40 per account but if both the order flow and the settling of cash
             management were automated, it could be around €5.38




37      Interviews with service providers, March 2006.

38      Interviews with service providers, March 2006.



                                                                                               Page 55
Costs of production components of the value chain

September 2006                                                                                   CRA International




        Finally, participants in some member states argued that costs were likely to vary due to
        competition. There were concerns about the degree of competition in transfer agency in
        some markets. For example, in Poland, a restrictive data protection act makes it
        impossible to use centralised transfer agency arrangements across a management group
        because they would need to ask every client to agree to have their data given to someone
        outside the country in order to enable this arrangement.39

        Looking at the results from the survey undertaken for this project, the cost of transfer
        agency is about 10% of the total administration costs. This results in a cost of around 1-2
        basis points.

        Figure 13: Client administration costs for different member states




                                 30.0



                                 25.0
          Cost in basis points




                                 20.0



                                 15.0



                                 10.0



                                  5.0



                                  0.0
                                        France   Italy   Luxembourg   Poland   Spain    Sweden          UK


        Source: CRA analysis of survey data. Sufficient information on other countries was not available on client
        administration. In later analysis Belgium and Germany have been estimated as the average of France, Italy and
        Spain (0.6 bp) as they are mainly banking models. Ireland has been set equal to Luxembourg (5.8bp) as they
        are both centres of cross-border fund activity.




39      In Poland, it is also not possible to use omnibus or nominee accounts because the civil law does not allow for
        the nominee principle. Thus, if the nominee went bankrupt, the ultimate client would not be protected since
        under the law the owner of the units would be the distributor and not the ultimate client. This affects both
        domestic funds and foreign funds.



                                                                                                             Page 56
Costs of production components of the value chain

September 2006                                                                    CRA International




        As is shown in the figure above, in many countries client administration is thought to cost
        around 1bps. However, we find that transfer agency is significantly more expensive in a
        number of countries, as is the case, in Poland, the UK and Luxembourg. The evidence
        from the interview programme supports these differences. It seems likely that the reasons
        for these differences vary from country to country. However, it is significant that the UK
        models rely on a fragmented distribution model has significantly higher costs. This could
        represent:

        •   A real cost to fragmentation of the distribution activity. That is, it is simply not
            necessary to maintain complex systems for recording remuneration of many advisers
            in other distribution models;

        •   It could represent that in bancassurance models this cost is spread across other
            activities; or

        •   Alternatively that this cost is captured in other parts of the value chain. For example,
            this might be recorded as a cost for the distributor in bancassurance model.




                                                                                            Page 57
Costs of production components of the value chain

September 2006                                                                     CRA International




        3.4.     ASSET MANAGEMENT

        The asset management activity encompasses: cash management; research; strategic
        asset management; operational asset management and dealing decisions; and guarantee
        provision.

        •   Cash management - involves the management of clients’ funds especially cash
            pending investment which is usually placed in money market and other short-term
            funds before it is invested in line with the aims of the fund. In addition, it will also
            include the conversion of foreign currencies, and the settlement of transactions
            carried out in different currencies.

        •   Research - A significant issue for all actively managed funds is the decision as to
            which of the securities to invest in. As part of this decision making process it is
            necessary to undertake research into issues including macro-economic conditions,
            sectoral trends, and company events which may affect fund performance. One of the
            areas that has been of concern in many member states is the way in which research
            is paid for since in many cases the cost of research is covered through trading costs.
            Although this is clearly an important issue in terms of the incentives that it causes, it
            is out of scope of our project and hence we do not focus on this issue.

        •   Strategic or tactical fund management - Strategic asset management focuses on the
            long term view and includes long-term asset allocation (the distribution of investment
            in the portfolio across various asset classes, currencies, sectors and geographical
            areas) and risk management. The freedom with which investment managers can
            make this decision will depend on the aims of different funds. For example, a fund
            which is described as an index tracker fund will not have any requirement to make
            asset allocation decisions since the assets in which the fund is invested have already
            been pre-determined. By contrast, another fund may be a “balanced” investment
            fund in which the managers have far more discretion and may need to take into
            account the trade-off between risk and return and the position in the economic cycle.

        •   Operational fund management and dealing decisions - Operational asset
            management includes the day-to-day selection of stocks and other securities to be
            included in the fund portfolio, and the corresponding decision-making process and
            implementation. The responsibility for making dealing decisions and choosing who to
            deal with can be undertaken by the fund manager, alternatively the decision as to
            how best to fulfil any transactions can be delegated to a centralised dealing function
            within the fund manager and shared across funds (and potentially for the managers
            own dealing decisions as well). Monitoring the transaction process and settlement
            will also be necessary. This also includes post-trade liaison with brokers and
            custodians.

        •   Guarantee provision - some investment funds will offer guarantees to consumers,
            most commonly relating to capital protection such that consumers are guaranteed to
            receive at least their original investment at the maturity of a product. Providing these
            guarantees will impose a cost.



                                                                                             Page 58
Costs of production components of the value chain

September 2006                                                                      CRA International




        These activities are paid for directly by the annual management charges that are applied
        to the fund. Trading costs that would be paid directly from the fund fall outside of the remit
        of this assignment.

        In the rest of this section we provide information on the total of the cost of asset
        management.


        3.4.1. Asset management costs by type of fund

        Asset management is one of the key areas in the value chain where interviewees
        consistently indicated that the cost varied significantly depending on the particular type of
        fund that was being considered.

        It was universally acknowledged that, for a fund of a given size, money market funds were
        cheaper than bond funds, which were cheaper than balanced funds, which were cheaper
        than equity funds. There was some disagreement regarding the relative cost of
        guaranteed funds in terms of whether they were more or less expensive than equity
        funds, with the majority that they were more expensive. Index tracking funds were also
        believed to be less expensive than all other types of funds. This is simply because of the
        use of passive management which means that no analysts are required to monitor the
        stock selection since this is automated with little research being required.

        The reasons typically given for the relative expense of equity funds compared to other
        funds were:

        •   Requiring more expensive staff due to need for a variety of skills and experience, for
            example, needing staff focused on particular sectors or regions;

        •   Heterogeneity of equities compared to bonds or money market vehicles has
            implications for research costs. Research needs to be focused on individual stocks
            and companies compared to research on government bonds which would more
            typically focus on macro-economic factors or on yields;

        •   Higher risk management costs and issues regarding corporate governance. If fund
            management companies own a large proportion of a particular company or market
            which would not arise in bond funds even if they own a substantial value of bonds i.e.
            fund management companies would not interact with central bankers in the same
            way as they do with individual companies; and

        •   Diversification issues meaning that doubling the size of an equity fund does not
            simply lead to doubling the size of all of the underlying investments since in some
            cases this could lead to owning very large proportions of these underlying
            investments and hence additional stock selections would often be made (although
            fund managers disagreed on the extent this is a significant issue).

        It was also mentioned that asset management costs depend on the efficiency of the
        clearing and settlement system in Europe. The data from the survey undertaken for this
        assignment supports the interview evidence regarding the relative cost of asset
        management by fund type as presented in Figure 14 below.


                                                                                              Page 59
Costs of production components of the value chain

September 2006                                                                        CRA International



        Scale

        In terms of the cost structure of fund management, there was a clear consensus that
        asset management is one of the areas where there are substantial economies of scale
        that can be exploited. For example, one fund manager suggested that increasing assets
        under management from €1-5 billion to €100 billion might lead to a 4 bp cost saving (in a
        multi-location set up).40This is also thought to vary significantly by the type of fund:

        •   Index tracking equity funds – economies of scale continue even after the fund gets
            large but the reduction in costs is relatively small;

        •   Actively managed equity funds – economies of scale are more significant but are
            exploited more quickly; and

        •   Bond and money market funds – where the extent of economies of scale lie
            somewhere between the index tracking and actively managed funds.

        However, it is clear that even within markets there is considerable variation in the size of
        funds. This suggests that there are likely to be significant economies of scale that remain
        unexploited in this part of the value chain.

        Figure 14 below shows that evidence of considerable economies of scale was found from
        the survey data for the five types of fund. As anticipated from interview evidence, the
        economies of scale are quickly exploited in an equity fund, with limited differences in the
        asset management cost once funds have reached a size of around €250 million assets
        under management. The economies of scale are much less significant for money market
        and capital protected. We were not able to identify economies of scale in index tracking
        funds.




40      Interview with a Belgian fund manager, who referred to a study by McKinsey.



                                                                                                Page 60
Costs of production components of the value chain

September 2006                                                                                                                  CRA International




        Figure 14: Asset management costs versus size of funds (€m) by types of fund


                                                                  Index tracker     Capital protected    Equity      Bond       Money market

                                                       45
            Cost of asset management in basis points



                                                       40

                                                       35

                                                       30

                                                       25

                                                       20

                                                       15

                                                       10

                                                       5

                                                       0
                                                            25   50   75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500 525
                                                                                           Assets under management


        Source: CRA analysis of survey data. Note that the data presented in this figures represents the statistical
        relationship between cost of assets management costs and scale estimated using data from the quantitative
        survey. This was estimated using an OLS based approach. Data on actual costs is not presented to retain
        confidentiality.


        The evidence suggests that for small funds, asset management costs of equity funds
        could be less that bond funds. This is surprising but not inconsistent with the evidence
        earlier that bond funds are cheaper. This reflects that on average bond funds are
        significantly larger than equity funds.

        It is also worth questioning where economies of scale in asset management come from. A
        number of possibilities were discussed in the interviews:

        •                                          Having only one fund manager – although a fund manager could manage two similar
                                                   funds, this still involves two sets of order, two processes for compliance, and the fund
                                                   objectives are never identical;

        •                                          Needing two set of trades – although in many fund managers centralised dealing
                                                   desks mean that economies of scale are already exploited, in smaller fund managers
                                                   the dealers are still responsible for trades and consolidation across managers does
                                                   not occur. Furthermore, in the case of fixed income securities, better prices might be
                                                   achieved when more assets are being offered or purchased;

        •                                          Using research – large fund managers are able to undertake research internally.
                                                   However, it is possible to use research provided by investment banks which is
                                                   commonly accessed by smaller fund managers. It should be noted that if research is
                                                   undertaken by the fund manager it is paid for through the management charge,




                                                                                                                                          Page 61
Costs of production components of the value chain

September 2006                                                                       CRA International




            whereas if research is acquired through soft commissions paid to investment banks,
            this will be paid directly by the fund and will not be part of the TER.

        •   Cash management – the need to keep a reserve of cash for consumers redeeming
            their shares is reduced for larger funds. The larger the fund the more predictable the
            pattern of redemption and subscriptions and the lower the level of cash that needs to
            be held as a percentage of assets.

        There is a counter argument that funds can become too big. In particular, it is theoretically
        possible that funds become so large, that trading decisions involve larger trading impacts,
        i.e. to sell shares without changing market prices requires more and more sophisticated
        trading strategies. Some cases were highlighted where concerns about the fund
        becoming too large (amongst other factors) has led to a fund being split between two fund
        managers adopting similar but not identical investment objectives. To the extent that the
        size of the fund is limited by trading costs, this suggests that the size of the fund is related
        to the liquidity of the underlying capital markets.

        Similarly, as mentioned above, if funds increase in size it may lead to funds owning large
        proportions of particular underlying investments and hence additional stock selections
        may need to be made rather than simply doubling the investment in the same stocks.
        The importance of this would again depend on the relative size of the fund compared to
        the market capitalisation of the underlying stocks.


        3.4.2. Asset management costs by member state

        Asset management is still predominantly undertaken in-house since many fund
        management companies see this as their core competency. However, the location of the
        fund management activity was often unrelated to the domicile of the fund. Instead, it is
        common to structure the asset management activity in one location, frequently London,
        Paris and Frankfurt for European investments whereas Asian and US stocks are often
        managed locally to the investment of assets.

        Where asset management is not possible within a group, it is common to buy-in this
        expertise. This can arise through outsourcing the management of the assets or through
        the structure of the fund e.g. by using funds-of-funds or manager-of-manager products.

        The reasons for asset management taking place in specialist locations is well known:

        •   The ISD directive means that there is largely a single market in the provision of asset
            management services;

        •   There are thought to be advantages in undertaking asset management in hubs – due
            to the access to a pool of skilled expertise. (Interestingly, this could lead to there
            being a number of hubs for example, some fund managers indicated that they split
            their asset management activity according to the type of asset in the funds for
            example equity funds might be managed in one location and bond funds managed in
            another); and




                                                                                                Page 62
Costs of production components of the value chain

September 2006                                                                                                                                CRA International




        •                                              Being closer to the market is important in two ways. First, there are trading
                                                       advantages in being closer to the market where trades take place. Second, this
                                                       make getting access to the companies being invested in easier e.g. it is possible to
                                                       access Japanese CEOs if the asset management is located in Japan. Hence asset
                                                       management is often located where most stocks are listed on exchanges.

        Results from our survey suggested that there were no differences in asset management
        costs that could be clearly identified across countries based on the domicile of the fund
        for equally sized funds. This is likely to reflect the fact that asset management is a global
        market and the activities are already taking place in the member state thought to have a
        competitive advantage. This is consistent with the view of the Expert Group on
        Investment Fund Market Efficiency that asset management is a global market.41 This
        supports a view that asset management activities are already exploiting cost efficiencies
        through the way that businesses have structured and centralised this part of the value
        chain.

        Figure 15: Asset management costs by member state



                                                       25.0
            Cost of asset management in basis points




                                                       20.0



                                                       15.0



                                                       10.0



                                                        5.0



                                                        0.0
                                                                                    Germany




                                                                                                        Italy




                                                                                                                                                         UK
                                                                 Belgium




                                                                                              Ireland




                                                                                                                Luxembourg



                                                                                                                             Poland



                                                                                                                                      Spain
                                                                           France




                                                                                                                                                Sweden




        Source: CRA analysis of survey data. Note that the difference in costs reflects the different average size of
        funds between member states.



        3.5.                                                  MANAGEMENT OVERHEADS AND SYSTEMS COSTS

        The final element of the costs is management overheads and systems costs, which can
        be broken into systems development and systems maintenance. These include:




41      Report of the Expert Group on Investment Fund Market Efficiency, European Commission, July 2006.



                                                                                                                                                         Page 63
Costs of production components of the value chain

September 2006                                                                   CRA International




        •   Management overheads covering overhead allocation for premises, senior
            management and cost of capital requirements at the management company level.
            The management company provides various corporate services to the fund including
            the provision of premises and senior management. In addition, funds face capital
            requirements and may hold capital beyond that which they might otherwise choose
            imposing an additional cost on the fund. To some degree this may substitute for
            other forms of investor protection; and

        •   Systems overheads covering planning and implementation of new IT and major
            enhancements to existing systems, and operational and technical maintenance of
            existing IT systems. In a technologically advanced sector there are costs associated
            with providing the underlying infrastructure on which investment funds can function. In
            addition to developing systems, they must be maintained in order to ensure the
            smooth functioning of the various operations involved in running an investment fund.

        Most fund managers believed that these costs were fixed costs which would be best
        allocated according to the value of assets under management. However, it was clear
        from interviews that few fund management companies would in fact undertake such an
        allocation for their own management information systems.

        Interviews also indicated that fund managers did not believe that either management
        costs or overheads would vary by fund type, or indeed by member state.

        There are a number of cautions that need to be made with regards to these numbers. In
        particular, many of the actual costs (including premises and IT systems) are likely to be
        shared across other products as well as UCITS funds including investment funds which
        do not meet the UCITS requirements, other products that might require fund management
        services, as well as other products in a wider management group. For example, the
        offices in which the fund management team reside may also be used by the rest of a
        bank.

        There was considerable variation in the cost of overheads between the various fund
        management companies who responded to our survey, although the sample on this part
        of the survey was smaller than elsewhere. Nonetheless, there was no systematic
        difference between member states that could be identified, and neither were there
        differences depending on the fund types offered.

        Overall, we estimate that the average costs were as follows:

        •   Systems costs of around 1 basis point; and

        •   Management costs of around 6 basis points.

        Despite suggestions by fund managers that these costs are fixed costs and therefore we
        should expect to see economies of scale, this was not borne out by the data. There
        appeared to be a very slight decline in costs as the value of assets under management
        increased, although this effect was minimal and we have therefore modelled this part of
        the value chain by applying the figures above to the assets under management of the
        fund.


                                                                                           Page 64
Costs of production components of the value chain

September 2006                                                                                                                                          CRA International




        3.6.                                        SUMMARY OF PRODUCTION COSTS

        In all of the sections above, we have detailed the cost of production for the different
        stages of the value chain. In this section we bring all of those components together to
        assess the overall cost of production for investment funds.


        3.6.1. Summary of production costs by fund type

        Bringing the cost elements together we can estimated the average cost of producing
        different types of fund. The results from this analysis are shown in Figure 16 below.

        Figure 16: Comparison of total production costs by member state based on current average
        fund sizes


                                                     Regulatory compliance of the fund   Audit   Fund Accounting   Custody   Transfer Agency   Asset management    Overheads

                                               60




                                               50
          Cost of production in basis points




                                               40




                                               30




                                               20




                                               10




                                                0
                                                     Money market                   Bond             Index tracking equity       Capital protected                Equity



        Source: CRA analysis


        As can be seen above, this results in equity funds having the highest costs, reflecting
        both higher asset management but also administration costs. Money market having the
        lowest costs, with this being primarily driven by the small asset management component.
        There is a much smaller difference between the costs of bond, index tracker and capital
        protected funds with the higher administration costs associated to capital guaranteed
        being outweighed by bonds having higher asset management costs.


        3.6.2. Summary of production costs by member state

        Turning to differences between member states, the first comparison we make is the cost
        of production for funds depending on the current average fund size of funds in that
        country. This is shown in Figure 17 below for an equity fund (the detailed results for other
        asset categories are shown Annex A).



                                                                                                                                                                           Page 65
Costs of production components of the value chain

September 2006                                                                                                                                                           CRA International




        Figure 17: Comparison of total production costs for equity funds by member state based on
        current average fund sizes


                                                      Regulator compliance of the fund     Audit             Fund Accounting   Custody         Transfer Agency   Asset management    Overheads

                                               90.0

                                               80.0
          Cost of production in basis points




                                               70.0

                                               60.0

                                               50.0

                                               40.0

                                               30.0

                                               20.0

                                               10.0

                                                0.0
                                                                                 Germany




                                                                                                                      Italy




                                                                                                                                                                                         UK
                                                       Belgium




                                                                                                   Ireland




                                                                                                                                  Luxembourg



                                                                                                                                                    Poland



                                                                                                                                                                 Spain
                                                                    France




                                                                                                                                                                            Sweden
        Source: CRA analysis


        Figure 17 shows that there are quite large differences between member states. Poland is
        found to be the most expensive country at 85 basis points with Sweden and France the
        cheapest at 42 and 43 basis points respectively.

        Examining the differences between countries, it is clear that the transfer agency cost is
        one area in which there are considerable differences in the level of costs between
        countries with Poland and the UK being relatively expensive in this area. As noted in
        section 3.3 above, this may be due to the different ways of structuring the transfer agency
        role in different countries which in turn is partly linked to distribution channels and whether
        external channels have been prevalent. Since the main function of the transfer agent is to
        hold customer level information, it may be the case that in some countries this role is
        performed by the distributor of even the central securities depositary which may therefore
        reduce the cost of transfer agency within the country concerned.

        As noted above, Figure 17 is presented on the basis of the total cost of production using
        the actual average fund size in each of the member states which varies from €44 million
        in Poland to €253 million in the UK. Hence part of the reason for the difference in the cost
        of production will reflect the extent to which different member states are more or less able
        to exploit any gains from economies of scale.

        For this reason, it is also useful to compare the cost of production for member states on
        the basis of a similar fund size. Whilst this could be considered to be a measure of
        efficiency, it is worth noting that for it to be an accurate reflection of efficiency would imply
        the assumption that countries are equally (in)efficient at producing a fund at their current




                                                                                                                                                                                        Page 66
Costs of production components of the value chain

September 2006                                                                                                                                                           CRA International




        average fund size as they would be at producing a fund at the average across the
        member states.

        Figure 18: Comparison of total production costs for equity funds by member state based on
        average fund size for all member states


                                                      Regulator compliance of the fund     Audit             Fund Accounting   Custody         Transfer Agency   Asset management    Overheads

                                               80.0

                                               70.0
          Cost of production in basis points




                                               60.0

                                               50.0

                                               40.0

                                               30.0

                                               20.0

                                               10.0

                                                0.0
                                                                                 Germany




                                                                                                                      Italy




                                                                                                                                                                                         UK
                                                       Belgium




                                                                                                   Ireland




                                                                                                                                  Luxembourg



                                                                                                                                                    Poland



                                                                                                                                                                 Spain
                                                                    France




                                                                                                                                                                            Sweden
        Source: CRA analysis


        Figure 18 compares the cost of production for an average fund size of €148 million in
        each of the member states. France and Sweden remain the cheapest two countries
        although the gap between France at 38 basis points and Sweden at 43 basis points has
        increased. Poland also remains the most expensive country although the cost of
        production has fallen from 85 basis points to 69 basis points reflecting the fact that Poland
        currently has the smallest average fund size and therefore makes considerable gains
        from economies of scale on this basis.

        What is also clear is that the difference in cost between member states is also much
        reduced in Figure 18 compared to Figure 17. Excluding the most expensive and the least
        expensive (Poland and France in both cases), we can compare the difference in
        production costs across the other member states. We find that in Figure 17, the
        difference between the second most expensive (Belgium) and the second cheapest
        country (Sweden) is 18 basis points, but when the comparison is done for the same fund
        size across countries the difference (between the UK and Sweden) falls to only 10 basis
        points.

        This suggests that much of the current differences in the cost of production in investment
        funds is mainly due to scale effects, with much smaller differences observed once scale
        has been taken into account.




                                                                                                                                                                                        Page 67
Costs of production components of the value chain

September 2006                                                              CRA International




        If we consider the cost of production for bond funds, we find a very similar pattern
        emerges compared to that for equities.




                                                                                      Page 68
Costs of distribution components of the value chain

September 2006                                                                      CRA International




4.      COSTS OF DISTRIBUTION COMPONENTS OF THE VALUE
        CHAIN

        In this section, we use the data collected in our quantitative survey along with evidence
        collected through our interviews to assess whether there are differences in the cost of
        distribution for funds of different types or for funds sold in different countries.

        It is important to note that in this section we focus on eight of the member states that we
        have examined and we do not provide information regarding the cost of distributing funds
        in Ireland and Luxembourg. The reason for this is that these countries focus on cross-
        border funds and we have focused on collecting data for funds of this type. As these
        funds are typically sold in other countries any information regarding the cost of distribution
        for Ireland and Luxembourg domiciled funds actually reflects the cost of distributing funds
        in other countries.

        As with the previous chapter, in this section we consider the cost of distributing funds by
        the type of fund and also by member state and then briefly examine the cost of
        distributing funds through other channels and whether the cost of distribution varies
        depending on whether a fund is domestically or foreign domiciled. In the final section we
        examine the overall cost of funds, bringing together the cost of production from section
        3.6 as well as the cost of distribution from this chapter.

        This part of the value chain covers:

        •   New product development - including considering which products are likely to be
            attractive to investors and identifying changes in regulation or the wider market place
            that allow for new opportunities. This activity may require a trade off to be made
            between the use of complex instruments to attain a desirable outcome for investors,
            and the additional regulatory and control costs that these may impose. In addition, it
            is also important to ensure that terms and conditions of the product will meet
            regulatory requirements.

        •   Promotional activity - marketing is undertaken so as to alert potential investors to the
            presence of the fund. In some cases, as well as marketing the investment fund
            directly to clients and external distributors, this may also involve marketing the
            product to an in-house distributor.

        •   Compensation to the distributor for reaching the end customer - gathering assets
            from investors occurs through the use of a number of different channels e.g. banks
            branches, financial advisers (FAs), direct purchases and fund supermarkets. The
            payment for the sales process can come direct from the consumer in the form of an
            upfront fee or via the fund manager through commission which is eventually paid for
            through the charges imposed on the customer. This is the largest component of this
            part of the value chain and it is where the rest of the chapter focuses.




                                                                                              Page 69
Costs of distribution components of the value chain

September 2006                                                                                   CRA International




        •   Documentation provision – supplying documents such as the simplified prospectus
            serves both to market the fund and also to provide information about the fund that
            can be used to understand the characteristics of the fund and potentially in comparing
            it to other funds.

        •   Distribution compliance - as well as imposing requirements on the product, most
            countries also impose requirements on the sale of investment funds. In many cases
            this will overlap with either the provision of advice or may be the responsibility of the
            distributor and hence included in the cost of distribution to end consumers.

        •   Advice - financial advice on whether or not a consumer should invest in a particular
            investment fund could be seen as a separate product to that of the actual investment
            fund itself. However, most funds will bundle the provision of advice into their cost of
            distribution.

        Distribution is the stage of the value chain at which consumers are contacted and in
        which assets from investors are gathered. The activity of distribution is multi-faceted and
        to some extent subjective. From the provider’s perspective distribution often involves
        “selling”, that is persuading consumers to decide to save, to save in a particular type of
        instrument and in a particular provider’s product. From the consumer perspective this
        may provide both information and also advice. It is extremely difficult to disentangle the
        cost of selling from the provision of advice and therefore we have not sought to do so in
        this report.42 Instead we have measured the payment made for distribution across
        different channels, by type of fund and between member states. There are two common
        forms of payment for distribution:

        •   An initial commission – typically this would represent a percentage of the
            contributions that the consumer is investing. It would normally come from the front
            end loaded charges paid by the consumer, with distributors often receiving all, or a
            significant proportion, of this charge;43 and

        •   An ongoing commission – typically this would represent a percentage of assets under
            management on an annual basis. In most countries it is negotiated as a percentage
            of the management fee.

        In many cases the front end load is intended to cover the cost of distribution and hence
        would be seen as cost reflective. However, it also acts as a deterrent to switching (the
        cost of which would otherwise fall on those shareholders of the fund that did not switch).
        Yet it is recognised that charging structures based on an initial load and an on-going



42      Even after significant regulatory attention it has not been possible in the UK to work out the degree to which
        commission is a payment made by the providers for sales activity or a payment for advice. See for example
        “Study of intermediary remuneration” for the Association of British Insurers, CRA International, February 2005
        and “The effect of commission based remuneration on financial advice” for the Financial Services Authority,
        CRA International 2002.

43      In some cases the payment made to the distributor may even exceed the front end loaded charges imposed on
        the consumer, with the fund manager recovering this over time.



                                                                                                             Page 70
Costs of distribution components of the value chain

September 2006                                                                                     CRA International




        charge are more complicated for consumers to compare, and that front-ended loads
        impose a cost on consumers who make mistakes and switch out of a fund after a short
        period of time.

        Differences in structure are often observed across member states and there are various
        reasons for this including:

        •    Regulatory: In some countries, such as the UK, the market has been encouraged to
             move to a level load structure because of concerns regarding consumer detriment for
             those customers who do not hold the fund for a long time; and

        •    Historical: If the fund industry has grown because of the greater popularity of money
             market funds then it is less likely that front-end loads will become the norm. For
             example, in France, the restriction regarding interest payments on current accounts is
             attributed as one of the reasons for the early growth in money market investment
             funds. By contrast, if investment funds have developed in competition with investment
             based insurance products that typically have front-end loads then there may have
             been less resistance to similar charges being applied on investment funds.

        The payment for distribution is complicated by negotiations that arise between the
        intermediary and the ultimate consumer. This can occur by not charging the consumer the
        initial load as occurs in many countries (especially in private banking) or by rebating the
        initial commission into the consumer’s investment (this appears relatively rare but is more
        common in the UK). It is equally possible that the distributor charges an amount on top of
        the initial charge for the services they provide. This is commonly the case in private
        banking where a charge may be applied across the consumer’s whole portfolio. Our
        interviews suggest fee based advice is relatively rare in the markets under consideration.
        The UK is likely to have the largest use of fees but even there less than 10% of advice
        costs are paid for on a fee basis and significantly less as a percentage of consumers
        receiving advice.44

        The rest of this section considers the variation in distribution costs between fund types
        and member states, as well as examining whether there are links with the distribution
        channels that are used.


        4.1.      DISTRIBUTION COSTS BY TYPE OF FUND

        In many cases, the main cost of distribution is the time spent in face to face discussions
        with consumers. Since much of this cost would arise regardless of the product taken out it
        might be expected that the cost of distribution would be invariant to the type of fund being
        sold. However, there are a number of reasons that would suggest that the cost of
        distribution might vary by type of fund including that:




44      “Study of intermediary remuneration” for the Association of British Insurers, CRA International, February 2005.



                                                                                                                Page 71
Costs of distribution components of the value chain

September 2006                                                                                                                                                     CRA International




        •                                                                          The sales process for more complex products such as equity instruments or
                                                                                   guaranteed products might be more lengthy or cumbersome both in terms of
                                                                                   persuading the client or from a regulatory perspective;

        •                                                                          Alternatively if the sales cost for a given consumer is fixed, this would represent a
                                                                                   lower proportional cost for larger contributions. To the extent that the average size of
                                                                                   contributions varies by types of fund this would lead to distribution costs as a
                                                                                   percentage of funds differing by fund types; and

        •                                                                          Finally, the cost of distribution will depend on the structure that is negotiated between
                                                                                   fund managers and distributors. It is common that in order to keep negotiations
                                                                                   tractable a relatively simple rule is used for any given distributor such as receiving a
                                                                                   fixed percentage of the annual management charge, with the result that some cross-
                                                                                   subsidisation occurs.

        Based on the evidence from the interview programme, active equity funds generally face
        the highest distribution charges in all countries and money market funds face the lowest.
        The results from our survey bear these views out and Figure 19 below, shows that there
        is a considerable difference in the ongoing cost of distribution by fund type, with equity
        funds facing a distribution cost over three times the cost of distribution for money market
        funds.

        Figure 19: Ongoing cost of distribution by type of fund
            Ongoing distribution cost as a percentage of assets under management




                                                                                   1.00%

                                                                                   0.90%

                                                                                   0.80%

                                                                                   0.70%

                                                                                   0.60%

                                                                                   0.50%

                                                                                   0.40%

                                                                                   0.30%

                                                                                   0.20%

                                                                                   0.10%

                                                                                   0.00%
                                                                                           Money market /   Government bond   Index tracking   Capital protected     Equity fund
                                                                                             cash fund            fund            equity             fund


        Source: CRA analysis of survey data


        Although Figure 19 shows a large difference between the distribution costs by type of
        funds, during the interviews many fund managers indicated that distribution costs are
        commonly set as a percentage of the management fee rather than being set as a fixed
        number of basis points for a particular type of fund. The implication of this is that since


                                                                                                                                                                             Page 72
Costs of distribution components of the value chain

September 2006                                                                                                                                          CRA International




        the annual management fees are lower on bond and money market funds than on equity
        funds, this leads to a lower charge being made for the distribution costs for these types of
        funds. In order to consider this, we present the costs of on-going distribution as a
        percentage of the annual management charge in Figure 20 below.

        Figure 20: Ongoing cost of distribution as a percentage of the annual management charge
          Ongoing commission as a percentage of annual management charge




                                                                           80%


                                                                           70%


                                                                           60%


                                                                           50%


                                                                           40%


                                                                           30%


                                                                           20%


                                                                           10%


                                                                           0%
                                                                                  Money market /   Index tracking   Capital protected Government bond     Equity fund
                                                                                    cash fund          equity             fund              fund


        Source: CRA analysis of survey data


        The impact of presenting the cost of distribution on this basis is to significantly reduce the
        differences between the types of fund. Indeed there is only a 4% point difference in the
        distribution cost as a percentage of the annual management charge for all funds but the
        capital protected fund which is the outlier of the set of funds.

        Even though the ongoing commission as a percentage of the annual management charge
        is constant, this nonetheless leaves equity funds with the highest distribution costs.
        Given the data collected from this project it is not possible to ascertain whether this
        represents a higher cost for distributing these products or whether higher profits are being
        made on the distribution of equity funds than on the distribution of money market and
        bond funds (although some fund managers did agree that this was the case). Whether the
        additional commission reflects higher costs or enables higher profits will determine
        whether distributors have a greater incentive to sell equity funds than other types of
        funds.


        4.2.                                                                     DISTRIBUTION COSTS BY MEMBER STATE

        As noted above, one of the factors determining the level of commission paid to
        distributors is the negotiation that occurs between the fund manager and the distributor
        and hence the relative bargaining power between the two parties is of importance. Most
        fund managers interviewed agreed that the level of commission that was typically paid to


                                                                                                                                                                  Page 73
Costs of distribution components of the value chain

September 2006                                                                                        CRA International




        a distributor would depend on the likely volume of business they would bring and whether
        it resulted in exclusive access to a particular channel.

        From the consumer’s perspective the way that commission is determined could cause
        concern. In a number of fund managers, there was little resistance to higher levels of
        commission - if commission was higher, the annual management charge through that
        distributor could also be higher. In some cases, this could require a different share class
        to be established or would lead to a share class with a high maximum level of commission
        built in. Although it is not necessarily the responsibility of the fund manager to restrain the
        costs of distribution, this is at least indicative that there is little competitive pressure on the
        charges for distribution. In turn this implies that fund managers believe that there is
        limited price sensitivity of consumers regarding the management fee or the TER that
        consumers then have to pay.

        Despite this being the majority view, in a small number of cases, fund managers had
        turned down distribution agreements because the resulting charge to the customer would
        have resulted in a low value product for consumers which the fund manager felt would be
        inappropriate given their brand values. This was particularly the case where distributors
        wanted to charge high front end loaded charges for money market products.

        The net result of this is that a “market norm”, of what fund management companies are
        paying for their distribution services, is established. From the interviews Italy and Spain
        were consistently mentioned as locations in which distribution is expensive whereas the
        UK was frequently mentioned as a location in which distribution is relatively less
        expensive. It is useful therefore to examine the percentage of the fund that a consumer
        effectively pays for the services of the distributor and this is presented in Figure 21 below
        for equity funds.45




45      Focusing on a single fund type is necessary to make meaningful comparisons. If there are differences in the
        costs of distribution for different funds, for example, there is a higher cost for distributing an equity fund than a
        money market fund, then if these aren’t taken into account it is difficult to make meaningful comparisons across
        countries. For example, countries with more equity funds will look more expensive.



                                                                                                                   Page 74
Costs of distribution components of the value chain

September 2006                                                                               CRA International




        Figure 21: Comparison of the costs of distribution of an equity fund by member state


                                       Initial commission          Ongoing commission

          6.00%



          5.00%



          4.00%



          3.00%



          2.00%



          1.00%



          0.00%
                  Belgium     France     Germany        Italy     Poland       Spain      Sweden        UK


        Source: CRA analysis of survey data.     Note that the percentage relates to contributions for the initial
        commission, and assets under management for ongoing commission.


        It is clear from Figure 21 that there are considerable differences in both the typical level of
        initial commission as well as the typical level of ongoing commission that is paid between
        different member states. Evidence in our survey indicates that countries such as
        Germany have a high initial commission but relatively low ongoing commission, whereas
        in Sweden and Spain it does not appear to be typical to pay any initial commission, but
        compared to Germany, the ongoing commission would be slightly higher.

        It is also important to note that the relative cost of initial and ongoing commission may
        vary depending on the type of fund that is being considered. For example in many
        countries there will be either no, or a very low, initial commission on money market
        products, whereas initial commission on equity products are much more common. This
        could reflect the products that these fund compete with do not typical have an initial
        charge or that the way consumers use the product means an initial fee would be
        prohibitive.

        This difference in the relative use of initial compared to ongoing commission means that
        comparisons between member states are difficult since examining only initial commission
        or only ongoing commission may give a mis-leading impression of the relative expense of
        different countries. Because of this difficulty, some countries have developed measures
        that attempt to account for both initial and ongoing payments. Typically this is done in
        respect of charges in order that consumers can compare products with different charging
        structures for example through using measures such as reduction in yield (in the UK) or
        total shareholder costs (as in Italy).




                                                                                                         Page 75
Costs of distribution components of the value chain

September 2006                                                                                                                                     CRA International




        A similar approach can be used to compare the distribution component of costs and we
        use the reduction in yield approach. In order to do this comparison it is necessary to
        make a number of further assumptions including:

        •                                                                         Whether the amount paid in is a single payment or regular payments – we assume a
                                                                                  one-off investment is made;

        •                                                                         The duration of the contract – since we are comparing the cost of distribution for
                                                                                  equity products we assume a holding period of 7 years (although if bond or money
                                                                                  market products were being considered then shorter periods would need to be
                                                                                  examined); and

        •                                                                         The investment return that is being made where we assume this is 7% per annum
                                                                                  (again if bond funds or money market funds were being considered then smaller
                                                                                  investment returns might be needed).46

        These assumptions are used in Figure 22 below.

        Figure 22: Comparison of the average distribution costs by member state for an equity fund
            Average distribution costs as a percentage of funds under managment




                                                                                  2.50%




                                                                                  2.00%




                                                                                  1.50%




                                                                                  1.00%




                                                                                  0.50%




                                                                                  0.00%
                                                                                          Belgium   France   Germany   Italy   Poland   Spain    Sweden     UK


        Source: CRA analysis of survey data


        Figure 22 supports the view expressed in interviews that Italy and Spain have relatively
        high distribution costs, although we also find that Poland and Germany have relatively




46      Using a reduction in yield approach means that differences in the investment returns have a relatively small
        impact on the outcomes and hence if somewhat lower or much higher returns were used, this would have only a
        limited impact on Figure 22.



                                                                                                                                                             Page 76
Costs of distribution components of the value chain

September 2006                                                                                                                     CRA International




        high costs. Similarly, that the cost of distribution is low in the UK is also supported by
        interview evidence, although France is found to have still lower costs.

        It is important to note that the comparison provided in Figure 22 depends on the
        assumptions. The longer the duration of the contract the lower the costs will be for funds
        with an initial commission. Based on the comparison above, this means that assuming a
        longer holding period would be less favourable to Sweden and Spain and more
        favourable to Belgium and Germany. This analysis also uses an average period in which
        the fund is held. To look at the actual costs in each market we would need an average
        duration that a consumer hold the fund for in each of the member states, this data is not
        currently available, so instead we have focused on other metrics (which we have then
        used in the quantitative analysis in the next chapter).

        The most common metric used during the interviews was the percentage of the annual
        management fee that is paid to the distributor on an on-going basis. When comparing
        between countries on this basis it is important to remember that this does not take into
        account any payment of initial commission and hence member states with low initial
        commission may look relatively expensive on this basis. In addition, it should be noted
        when considering Figure 23 below, that, because the chart is calculated as a percentage
        of the ongoing management charge, for any given level of commission paid to distributors
        calculated as a percentage of assets under management, this will represent a lower
        proportion of the ongoing management charge the higher is that management charge.
        Hence this figure could be interpreted as providing information on the relative bargaining
        power of the fund managers compared to the distributors.

        Figure 23: On-going distribution as a percentage of the annual management charge for
        equity funds



                                                                          90%
          Ongoing commission as a percenage of annual management charge




                                                                          80%


                                                                          70%


                                                                          60%


                                                                          50%


                                                                          40%


                                                                          30%


                                                                          20%


                                                                          10%


                                                                          0%
                                                                                Italy   Sweden   Spain   Germany   Poland   UK   Belgium   France


        Source: CRA analysis of survey data




                                                                                                                                             Page 77
Costs of distribution components of the value chain

September 2006                                                                                 CRA International




        4.3.        DISTRIBUTION COSTS BY CHANNEL

        An important factor to keep in mind when making comparisons between member states is
        that the distribution channel that is used in different member states. Across the whole of
        Europe, banks are by far the most commonly used distribution channel for investment
        funds. Indeed, in all the member states examined, with the exception of the UK, bank
        distribution is the most common channel within the country. Figure 24 below highlights
        the distribution channels used in each of the member states.

        Figure 24: Distribution of investment funds by channel for all funds


                      Retail bank    Private bank   IFA-advised    Supermarket   Direct   Funds of funds   Other

            100%

             90%

             80%

             70%

             60%

             50%

             40%

             30%

             20%

             10%

               0%
                      Belgium       France   Germany       Italy      Poland     Spain      Sweden         UK


        Source: European fund market data digest 2006, FERI Fund Market Information. Belgium, Poland, Sweden are
        based on estimates from the surveys and interviews. Where fund supermarkets are part of a retail bank it has
        not been possible to determine how funds have been categorised, i.e. in Poland, mBank is a subsidiary of BRE
        bank but we have not identified data supermarket in Poland.


        Within member states the distribution of funds varies dramatically and it is common to
        differentiate between the private banking markets, the retail bank and financial advisers.
        Each of these has quite different economics and types of customers and therefore it
        would be unsurprising if the funds offered through these different channels and the terms
        under which consumers pay for them also varies.

        Taking all of the member states together, an analysis of the survey data suggests that
        typically, the ongoing commission payment is greater for proprietary banks compared to
        other distribution channels (including distribution through banks outside of the financial
        groups and other channels) with equity funds paying ongoing commission on average of:

        •   137 basis points for proprietary banks; and

        •   77 basis points for other channels.



                                                                                                           Page 78
Costs of distribution components of the value chain

September 2006                                                                     CRA International




        However, this hides very different dynamics between the different member states where:

        •   In Belgium there is a lower percentage of on-going costs paid to proprietary banks
            than non-proprietary banks;

        •   In Italy and Spain there is a higher percentage of on-going costs paid to proprietary
            banks than non-proprietary banks; and

        •   In Sweden there is no clear difference between the percentage of on-going costs paid
            to proprietary banks and non-proprietary banks.

        However, what was clear from the interviews was that although the amount paid might
        vary between distribution channels, this was not in fact due to underlying cost differences.
        Indeed, for those fund managers that were part of banking groups and hence distributed
        primarily through bank branches, no interviewee believed that the underlying cost of
        distribution for the distributor varied according to whether the fund was proprietary or non-
        proprietary i.e. it would not cost a bank any more to distribute an external fund compared
        to an internal fund. Furthermore, there was consensus view that no additional costs or
        compliance checks would be required. Therefore the differences in the commission paid
        seem primarily to reflect the bargaining position between fund manager and distributor
        rather than underlying cost differences.

        If the distributor is in a strong position, then they would be able to negotiate a higher
        payment from those fund managers in need of a distribution outlet than they would need
        to take from their in-house fund manager. Interestingly, it was not in fact always found to
        be the case that a non-proprietary fund would be charged more than a proprietary fund.
        This is likely to be because different banking groups will take different views regarding the
        internal pricing for these services, with some groups preferring to make more money on
        the fund management side (and hence in the actual fund management company) and
        some groups preferring to make more money on the distribution side (and hence in the
        bank). Indeed, some interviewees suggested that within banks as much as 90% of the
        management charge could be going to distribution.

        The counter argument was also expressed that non-proprietary funds were those likely to
        have a consumer brand and hence would be able to negotiate a lower level of ongoing
        commission compared to the funds of the in-house provider. There appears to be some
        evidence that those with strong consumer brands are gaining more bargaining power over
        time and that this is resulting in a lower percentage of the annual management charge
        being paid to the distributor.

        Finally, it was clear that the proportion of the management fee received depends not only
        on the channel, but more particularly it depends on the value of contributions that the
        distributor can offer to the fund manager. Larger distributors typically receive a higher
        proportion of the management fee than smaller distributors.


        4.4.     DISTRIBUTION COSTS BY DOMICILE OF THE FUND

        So far this chapter has focused on the cost of distribution depending on the location in
        which funds are being distributed irrespective of where those funds are actually domiciled.


                                                                                             Page 79
Costs of distribution components of the value chain

September 2006                                                                    CRA International




        In both interviews and from our quantitative data, it is possible to examine whether there
        are differences in the cost of distribution depending on the domicile of the fund and in
        particular depending on whether the fund was a domestically domiciled fund or a cross-
        border fund.

        There is no evidence from the interview programme that the domicile of the fund affected
        either the cost of distribution or the payment made to distributors. That is, for any given
        distributor, they would expect to receive the same ongoing commission for a particular
        type of fund regardless of whether that fund was domiciled domestically or was domiciled
        in another member state. Similarly, for any given fund manager distributing in a particular
        location, they would expect to pay the same amount of ongoing commission for a
        particular type of fund to a particular distributor regardless of whether the fund was
        domiciled domestically or in another member state.

        Examining the data finds that the ongoing commission received on an equity fund on
        average across all member states is:

        •   64 basis points for a cross-border domiciled fund; and

        •   113 basis points for a domestically domiciled fund.

        However, it is clear from the discussion above that any differences that are observed in
        the payments for distribution for cross-border funds compared to domestic funds are not
        in fact, due to differences in the cost of distributing a cross-border fund per se. Instead,
        the primary driver of these differences is the channel choice and especially whether or not
        the fund management company has a proprietary distribution channel in the particular
        member state. The higher payment for domestically domiciled funds results from them
        being more likely to be distributed through proprietary banks than other channels. Further,
        when the data is examined in greater detail, there is no difference in the level of ongoing
        commission received by non-proprietary banking channels when we consider funds
        domiciled domestically or abroad – this again supports the view that it is bargaining power
        rather than the location of the fund that drives the level of ongoing commission.

        One of the concerns expressed is cross-border funds will often be marketed in a
        particular country by management companies who do not own local distribution networks.
        Therefore they could face buyer power on the part of the distributor who is therefore able
        to extract some of the economic rents from them. By contrast, many of the domestically
        domiciled funds will be managed by companies who also own proprietary distribution
        channels. The prices charged for distribution are then most likely to be due to a
        negotiation within the group. We find no support for the concern above that cross-border
        funds pay a higher margin for distribution than domestic funds.

        A final issue that may differentially affect cross-border funds is the structure of
        commission. As set out above some countries may have a front-end-load and a low
        ongoing management charge, others may have no front-end load and a higher ongoing
        management charge. One of the issues for distribution is that foreign funds may not have
        the same pricing structure as domestic funds and therefore may struggle to be sold.
        However, this seems to be solved through use different sub-funds and does not impose a
        significant barrier.


                                                                                            Page 80
Costs of distribution components of the value chain

September 2006                                                                                                                                  CRA International




        4.5.                                   SUMMARY ON THE TOTAL COST OF THE VALUE CHAIN

        In chapter 3 we detailed the cost of production for the different stages of the value chain
        and the sections above have provided detailed information on the cost of distribution in
        each of the member states. In this section we bring both of these components together to
        assess the total cost for investment funds.


        4.5.1. Total cost of the value chain by fund type

        Adding the distribution to the production costs, we can calculate the total cost of different
        fund types. This is shown in Figure 25 below.

        Figure 25: Comparison of total costs by fund type based on current average fund sizes


                                                         Regulatory compliance of the fund and overheads   Administration    Asset management   Distribution

                                       140.0



                                       120.0



                                       100.0
          Total cost in basis points




                                        80.0



                                        60.0



                                        40.0



                                        20.0



                                         0.0
                                                 Money market               Bond             Index tracking equity          Capital protected            Equity



        Source: CRA analysis


        The impact of introducing distribution costs is to increase the differences between the
        fund types. However, equity remains the most expensive and money market remains the
        cheapest. There is now a greater differential between capital protected, index tracker and
        bond with the lower distribution cost associated with bonds, resulting in being the
        cheapest of the three.


        4.5.2. Total cost of the value chain by member state

        It is important to note a number of assumptions that feed into this section. First, as
        discussed above, information on the cost of distribution in Luxembourg and Ireland has
        not been gathered due to the focus of these two countries on production rather than
        distribution. In Figure 26 below, the distribution cost for these two countries is a simple
        average of the distribution cost in the other eight member states. This cost has been



                                                                                                                                                                  Page 81
Costs of distribution components of the value chain

September 2006                                                                                                                                    CRA International




        included rather than leaving only the production cost in order to prevent a distorted
        comparison from being shown.

        Second, we have only used the ongoing cost of distribution in the chart since to do
        otherwise would involve making assessments of the likely length of time for which
        investors would hold their investment funds – something which may differ between
        member states.

        Thirdly, as in section 3.6, we first compare the total cost for funds depending on the
        current average fund size of funds in each member state and this is shown in Figure 26
        below for an equity fund.

        Figure 26: Comparison of total costs for equity funds by member state based on current
        average fund sizes


                                                            Regulatory compliance of the fund and overheads   Administration   Asset management   Distribution

                                          250.0
          Cost of funds in basis points




                                          200.0



                                          150.0



                                          100.0



                                           50.0



                                            0.0
                                                                           Germany




                                                                                                    Italy




                                                                                                                                                                 UK
                                                  Belgium




                                                                                        Ireland




                                                                                                                 Luxembourg



                                                                                                                               Poland



                                                                                                                                         Spain
                                                               France




                                                                                                                                                      Sweden




        Source: CRA analysis


        Figure 26 shows that there are very large differences between member states when
        distribution costs are included. Poland is found to be the most expensive country at 235
        basis points with France the cheapest by some margin at 80 basis points.

        Examining the differences between countries, it is clear that the differences in distribution
        cost is the main area which drives differences in the total level of costs between
        countries. Hence France, and the UK which were both found to have low distribution
        costs are also found to have low total costs. Similarly, Spain and Italy were found to have
        relatively high distribution costs and this then leads to them being the second and third
        most expensive countries overall.

        As noted above, Figure 26 is presented on the basis of the total cost of production using
        the actual average fund size in each of the member states. Hence part of the reason for



                                                                                                                                                                 Page 82
Costs of distribution components of the value chain

September 2006                                                                                                                                    CRA International




        the difference in the cost of production will reflect the extent to which different member
        states are more or less able to exploit any gains from economies of scale – however, we
        found no evidence to suggest that larger funds were any cheaper to distribute than
        smaller funds and hence there is no difference in the cost of distribution in basis points as
        the size of the fund changes.

        Nonetheless, for completeness we provide a comparison of the total costs for investment
        funds for member states on the basis of a similar fund size in Figure 27 below.

        Figure 27: Comparison of total production costs for equity funds by member state based on
        average fund size for all member states


                                                            Regulatory compliance of the fund and overheads   Administration   Asset management   Distribution

                                          250.0
          Cost of funds in basis points




                                          200.0



                                          150.0



                                          100.0



                                           50.0



                                            0.0
                                                                           Germany




                                                                                                    Italy




                                                                                                                                                                 UK
                                                  Belgium




                                                                                        Ireland




                                                                                                                 Luxembourg



                                                                                                                               Poland



                                                                                                                                         Spain
                                                               France




                                                                                                                                                      Sweden




        Source: CRA analysis


        Figure 27 compares the cost of production for an average fund size of €148 million in
        each of the member states. As can be seen from a comparison between Figure 27 and
        Figure 26, given that distribution costs represent the largest share of costs in all member
        states, the pattern of cost differences is the same as in both charts.

        If we consider the total cost of funds for bond funds in addition to equity funds, we find a
        somewhat similar result, although there is less variation in the cost of distribution for bond
        funds across member states compared to equity funds. Spain is found to be the most
        expensive country which mainly reflects the fact that Poland has a significantly lower cost
        of distribution for bond funds compared to equity funds and hence goes from being the
        most expensive equity fund to being cheaper than Italy and Spain. The detailed results
        are presented in Annex A.

        Further consideration of these differences is undertaken in chapter 5 where we consider
        whether there are any potential cost savings in different scenarios.




                                                                                                                                                                 Page 83
Quantification of potential cost savings

September 2006                                                                                    CRA International




5.      QUANTIFICATION OF POTENTIAL COST SAVINGS

        This section examines a range of different scenarios designed to test the scale of
        potential cost savings that could be made if the cost of investment funds in one country
        became more like that in another country. The cost estimates are aggregated for the ten
        member states examined for this project based on the market size as at 2005 and
        represented in both monetary terms (in euros) and in basis points. To illustrate the size of
        these cost savings we have also calculated the resulting savings as a percentage of the
        total expenses ratio of the fund and how this might impact the returns to savers if they
        were passed on.

        It is important when considering these results to remember that they represent a “static”
        estimate of potential annual saving that could occur if cost changes were to occur
        immediately and to all funds in a given market. These represent the savings that result
        from a convergence in costs and we have not quantified any loss that would result from
        lower quality, reduced product differentiation or a corresponding reduction in service
        levels (where these are relevant these are considered in the next chapter). Equally we
        have not included any impact on the market, for example, that the demand for funds could
        change as a result of cost savings leading to lower prices.47

        We first focus on the cost of producing funds rather than the cost of distributing them.
        Although it is clear from the preceding chapters that distribution represents the largest
        cost and is the area where there are the most significant differences by country, this is
        also the area where it is most difficult to see how it will be affected by integration of the
        investment fund market.

        In the next chapter, we look at the degree to which these cost saving may occur due to
        on-going trends, the speed at which this might occur and both whether any savings would
        be passed onto consumers and the mechanism this would take.


         5.1.     COMPARISONS WITH TODAY’S FUND SIZES

        The first set of comparisons is based on the cost of production given the existing pattern
        of fund sizes observed in the member states. As is seen in the country summaries to this
        report there are considerable differences in the average size of funds from country to
        country.




47      Note that any increase in demand resulting from lower prices could result in a higher total cost associated with
        the production of investment fund but would also result in higher consumer benefits. Therefore the savings are
        likely to represent a lower bound of the benefits that could be passed onto consumers.



                                                                                                               Page 84
Quantification of potential cost savings

September 2006                                                                                                                                                            CRA International




        Figure 28: Comparison of total production costs for equity funds by country based on
        current average fund sizes


                                                        Regulator compliance of the fund         Audit   Fund Accounting        Custody    Transfer Agency      Asset management            Overheads

                                                 90.0

                                                 80.0
            Cost of production in basis points




                                                 70.0

                                                 60.0

                                                 50.0

                                                 40.0

                                                 30.0

                                                 20.0

                                                 10.0

                                                  0.0
                                                                             Germany




                                                                                                         Italy
                                                         Belgium




                                                                                                                                                                     UK


                                                                                                                                                                             Weighted-avg



                                                                                                                                                                                                 US
                                                                   France




                                                                                           Ireland




                                                                                                                   Luxembourg


                                                                                                                                  Poland


                                                                                                                                             Spain


                                                                                                                                                       Sweden




                                                                                                                                                                                 fund
        Source: CRA analysis


        Figure 28 above provides the production costs for an equity fund by country based on the
        current average fund sizes. As well as giving cost details on the ten member states
        investigated in this project, we have also included the weighted average fund cost and the
        cost of funds produced in the US. The weighted average fund cost is calculated using the
        value of equity funds under management in each of the ten member states. In total there
        were around €1.3 trillion assets under management in equity funds in the ten member
        states. To examine the potential for cost savings we have used the model (described in
        Annex A) to test hypothetical scenarios where we change the costs in a given country to
        reflect those observed elsewhere. The four scenarios we have considered are:

        •                                        Average cost - For those countries where the total cost of production is greater than
                                                 the weighted average total cost, we reduce the cost of production to the weighted
                                                 average across the ten member states. For countries with costs below this average
                                                 we leave them as they are;

        •                                        Cross border cost – Here we assume that the funds in a given country adopt the
                                                 costs associated with cross-border funds. Funds domiciled in Luxembourg and
                                                 Ireland have been used to estimate the cost of cross-border funds. In this scenario,
                                                 where the cost of production is greater than the average cost for Luxembourg and
                                                 Ireland, we reduce the cost of production in other countries to this cost. For countries
                                                 with cost below the average cost of cross-border funds we leave them as they are;

        •                                        Lowest cost - For all countries we reduce the cost of production to the total cost of
                                                 production in the lowest cost country (which is Sweden for equity funds as shown in
                                                 Chapter 3); and


                                                                                                                                                                                               Page 85
Quantification of potential cost savings

September 2006                                                                      CRA International




        •     US cost – For all countries where the cost of production is greater than the cost of
              production in the US (in this case all ten member states), we reduce the cost of
              production to that observed in the US for a similar type of fund. For countries with
              cost below the cost of US funds we leave them as they are

        Table 9 below highlights the cost savings of these four scenarios compared to today’s
        cost base (which based on the markets today for equity funds in the ten member states is
        equal to about €6.4 billion).

        Table 9: Cost savings in production based on current fund size for equity funds

                                    Cost of production      Cost savings in €     Cost savings in bp

            Today                          €6,420 million            -                      -

            Average cost                   €6,166 million      €254 million               2 bp

            Cross-border cost              €6,369 million       €50 million               <1 bp

            Lowest cost                    €5,390 million      €1,030 million             8 bp

            US cost                        €4,635 million      €1,785 million             14 bp

        Source: CRA analysis


        Table 9 shows that the impact of reducing the cost of production for those countries with
        above average costs would be a saving of approximately €254 million per year, which is
        equivalent to a reduction in the costs of the fund by 2 basis points. Putting this in context,
        the average TER of equity funds across the ten countries is around 180 basis points, so
        this represents less than 1% cost saving. The small cost saving reflects that where
        countries have costs substantially above the average (Poland and Belgium) but these
        tend to be the smaller markets for equity funds.

        Considering the second scenario, the benefits from moving to a production cost of cross-
        border funds is found to be very small indeed. This result is not surprising when we
        consider that Ireland and Luxembourg are not found to be among the cheapest of
        countries in terms of total production costs as seen in Figure 28 above.

        However, we see more significant cost savings resulting from a move to the lowest cost
        member state or to US costs. Moving to the cost of production observed in Sweden would
        lead to a cost saving of some €1,030 million or 8 basis points. However, this is dwarfed
        by the potential savings that would arise if all member states could replicate the current
        costs of the US where nearly €1.8 billion of cost savings or 14 basis points could be
        achieved. However, even assuming a wholesale move to US costs, this represents less
        than a 10% reduction in the total expenses currently charged to the funds.


         5.2.         COMPARISONS WITH EU AVERAGE FUND SIZES

        As noted throughout this report, one of the main drivers of differences in cost is that the
        size of the average fund in member states varies significantly. Since economies of scale
        exist within many components of the investment fund value chain, this means that some



                                                                                                  Page 86
Quantification of potential cost savings

September 2006                                                                                                                                                           CRA International




        member states are currently able to exploit economies of scale to a greater degree than
        others. In this set of scenarios we therefore consider the costs of producing an average
        sized EU fund (with assets under management of €148 million) in each of the countries.
        After making the adjustment regarding average fund size, we can test if there are cost
        savings that result from an increase purely in efficiency. On this basis we examine a
        similar scenarios to those considered above.

        Figure 29: Comparison of total production costs for equity funds by country based on
        average EU fund sizes


                                                       Regulator compliance of the fund         Audit   Fund Accounting        Custody    Transfer Agency      Asset management            Overheads

                                                80.0

                                                70.0
           Cost of production in basis points




                                                60.0

                                                50.0

                                                40.0

                                                30.0

                                                20.0

                                                10.0

                                                 0.0
                                                                            Germany




                                                                                                        Italy




                                                                                                                                                                    UK


                                                                                                                                                                            Weighted-avg
                                                       Belgium




                                                                                          Ireland




                                                                                                                  Luxembourg


                                                                                                                                 Poland


                                                                                                                                            Spain




                                                                                                                                                                                                US
                                                                 France




                                                                                                                                                      Sweden




                                                                                                                                                                                fund

        Source: CRA analysis


        It is important to note that moving to the average EU fund involves the average fund size
        going up in some countries but falling in other countries. In the latter case, the cost of
        production actually increases in these countries in terms of setting the baseline in
        comparison to the figures for today in Table 9 above (although as seen in Table 10 the
        net result is a reduction in the total costs of production to €6.3 billion). Note we have not
        included the US in this scenarios as we are not able to model how US costs would
        change if we applied the EU average fund size to the US.




                                                                                                                                                                                              Page 87
Quantification of potential cost savings

September 2006                                                                      CRA International




        Table 10: Cost savings in production based on average EU fund size for equity funds

                                    Cost of production      Cost savings in €      Cost savings in bp

          Average EU fund size             €6,284 million            -                      -

          Average cost                     €5,980 million      €304 million                2 bp

          Cross-border cost                €6,186 million       €98 million               <1 bp

          Lowest cost (France)             €4,946 million      €1,338 million             10 bp

        Source: CRA analysis


        Table 10 shows that once we have removed the impact due to scale, the impact of
        reducing the cost of production for those countries with above average costs would be a
        saving of €304 million or 2 basis points. This is very similar to the cost saving identified in
        the previous scenarios suggesting efficiency improvements alone could yield the same
        cost savings as moving to average fund. The saving related to cross-border funds is twice
        as large as found above but remains very small. This result is somewhat surprising given
        that one of the advantages of cross-border funds is that the fund can serve multiple
        jurisdictions and hence realise benefits of scale.

        Moving to the cost of production observed in the lowest cost country (which after
        adjusting to European average fund size is now France) would lead to a cost saving of
        some €1,338 million or 10 basis points. So after accounting for scale, the cost savings
        resulting purely from efficiency would be as large as the cost savings from moving to
        today’s cheapest fund.

        It is important when considering these numbers to remember that the scale results are not
        simply a subset of the cost savings identified above. To look at scale effects, we have
        adjusted the funds to be the same size and this affects total costs. Secondly, after
        adjusting for size, the member states with the lowest cost funds has changed. So instead
        of looking at the cost saving associated with moving to today’s Swedish fund, we are
        looking at what would happen if we adopted the cost structure of the French industry. The
        intuition behind this is that a country may be very efficient but not have lowest cost as it
        does not exploit scale economies.


         5.3.    COMPARISONS BASED ON TODAY’S COSTS BUT DIFFERENT FUND SIZES

        To understand the potential cost savings that would result from increased scale we have
        used the model to look at how the costs would change in each country when we change
        the average funds sizes. In Table 11 below, we look at the aggregate European cost after
        we have changed the fund size in each of the countries to the average European fund
        size (€148m), the maximum European fund size (€253m) and the average US fund size
        (€886m).




                                                                                                  Page 88
Quantification of potential cost savings

September 2006                                                                    CRA International




        Table 11: Cost savings due to increasing fund size

                                    Cost of production       Cost savings in €   Cost savings in bp

            Today                          €6,420 million             -                   -

            Average EU fund size           €6,284 million       €136 million             1bp

            Maximum EU fund size           €5,614 million       €806 million             6bp

            Average US fund size           €4,183 million       €2,237 million          17 bp

        Source: CRA analysis


        As described above the impact of moving all funds to the European average size is
        relatively small. This is to be expected, in countries with low average fund sizes, we have
        increased fund size lowering costs, however this is offset by countries with greater than
        average fund size, where we have reduced the size of the funds thereby raising costs.
        When we move all funds to the European maximum (in this case the UK) there is a cost
        savings of 6 bp but this is only slightly more than a third of the reduction in costs that
        would be achieved from moving to the US fund size. (Note that the full set of scenario
        results relating to funds of these sizes can be found in Annex A).

        One of the most interesting implications of this analysis is that the impact of moving
        European countries to the US fund size (as seen in Table 11) has a larger impact on
        costs than moving to US costs (as seen in Table 9). This suggests that although there are
        significant scale benefits in the US markets, based on this analysis, European funds are
        as efficient as their US comparators.

        It is, however, important to note that considerable caution should be placed on
        interpreting this result since it involves forecasting the cost functions for the various
        components of the value chain far outside of the current range in which the modelling was
        undertaken. For this reason the uncertainty surrounding the likely cost of production
        would be expected to be quite considerable and hence these estimates may not be
        robust.


         5.4.       COMPARISONS OF SAVINGS RESULTING FROM EFFICIENCY VERSUS SCALE

        In the sections above we have investigated the impact of moving to the costs of different
        countries and looking at this in terms of scale and efficiency. Table 12 brings together
        these results. Here we compare the savings resulting from:

        Scenarios investigating efficiency:

        •     Reducing the costs of funds above average European costs after assuming away
              scale effects – this increases the fund size in each member state to the weighted
              average and also reduces the cost of production for those countries that have above
              average production costs at this level of assets and this is taken from Table 9;




                                                                                                Page 89
Quantification of potential cost savings

September 2006                                                                                    CRA International




        •      Reducing the costs of funds to the minimum European costs after assuming away
               scale effects – this increases the fund size in each member state to the weighted
               average and also reduces the cost of production to the costs of cheapest European
               investment fund markets. This is also taken from Table 9;

        Scenarios investigating scale:

        •      Maximum EU fund size with country cost – this increases the fund size in each
               member state to the EU average fund size but keeps the country specific cost
               function and this is taken from Table 11; and

        •      Average US fund size with country cost – this increases the fund size in each
               member state to the US average fund size but keeps the country specific cost
               function and this is taken from Table 11.

        The comparison of these scenarios allows us to examine both scale and efficiency
        effects.

        Table 12: Cost savings in production based on current fund size for equity funds

                                                      Scenario                              Cost savings in bp

                                                      Average cost                                   2 bp
            Efficiency effects48
                                                      Lowest cost                                    8 bp

                                           Maximum member state fund size                            6 bp
            Scale effects49
                                                 Average US fund size                               17 bp

        Source: CRA analysis


        It is clear from Table 12 above that the efficiency gains within Europe are significant at
        around 8 bp. There are also significant gains to be made from increased scale, that is, if
        we increase all European funds to the average US size fund we would gain 17bps. This
        clearly shows that the cost savings arising due to scale are even bigger than those arising
        from increased efficiency. However, it is important to note that these savings result from
        increasing the size of average funds whilst still using European cost conditions.


         5.5.       COMPARISONS INCLUDING DISTRIBUTION COSTS

        In all of the scenarios so far in this chapter we have focused on the cost of producing the
        investment fund rather than including the cost of distributing the product. In this scenario,




48      Efficiency effects are taken from Table 10 based on moving to the average and lowest cost of production for a
        fund size at the member state weighted average.

49      Scale effects are taken from Table 11 based on moving to different fund sizes but assuming that the current cost
        function in each member state is retained.



                                                                                                               Page 90
Quantification of potential cost savings

September 2006                                                                                                                                              CRA International




        we consider the impact of different scenarios examining the total cost of fund production
        including distribution.

        Figure 30: Comparison of total costs for equity funds by country based on current average
        fund sizes


                                                             Regulatory compliance of the fund and overheads        Administration       Asset management    Distribution

                                           250.0
           Cost of funds in basis points




                                           200.0



                                           150.0



                                           100.0



                                            50.0



                                             0.0
                                                                       Germany




                                                                                            Italy
                                                   Belgium




                                                                                                                                                        UK


                                                                                                                                                                  Weighted-avg



                                                                                                                                                                                  US
                                                             France




                                                                                 Ireland




                                                                                                       Luxembourg


                                                                                                                      Poland


                                                                                                                                 Spain


                                                                                                                                               Sweden




                                                                                                                                                                      fund
        Source: CRA analysis


        It is clear from Figure 30 above that there are substantial differences in the total cost of
        both production and distribution for equity funds between countries and this is reflected in
        the results that are found below in terms of the potential cost savings resulting from the
        same set of scenarios examined above.50




50      In calculating the total cost differences, because they now include the cost of distribution, we need to take into
        account the difference between retail and institutional funds since the distribution methods are substantially
        different between the two. In particular, we estimate from Fitzrovia data that 78% of funds are retail funds and
        hence we include in the cost savings below 78% of the potential savings related to distribution along with all of
        the savings associated to production.



                                                                                                                                                                                 Page 91
Quantification of potential cost savings

September 2006                                                                        CRA International




        Table 13: Cost savings in production and distribution based on current fund size for equity
        funds

                                           Total cost       Cost savings in €        Cost savings in bp

          Today                        €16,979 million               -                        -

          Average cost                 €15,451 million         €1,528 million               12 bp

          Lowest cost                  €11,627 million         €5,352 million               41 bp

          US cost                          €9,270 million      €7,710 million               60 bp

        Source: CRA analysis


        It is clear from Table 13 above that there are very substantial potential cost savings once
        distribution costs are included in the analysis. If those countries that have total costs
        above the current weighted average reduced to the current average, this would lead to an
        annual potential cost saving of €1.5 billion or 12 basis points, if costs fell as far as the
        lowest cost country, the figures would be a huge €5.4 billion or 41 basis points and if they
        fell as far as the current US cost figures then there would be potential for cost savings of
        some €7.7 billion or 60 basis points. Comparing this to the average TER this represents a
        33% reduction in costs of total costs of the fund.

        Given that the variation in the cost of distribution was seen to be much greater than the
        variation in other parts of the value chain, and since distribution is the largest component
        of costs in all member states, it is not surprising that when distribution is included in the
        scenario analysis the potential cost savings are substantial. An important caveat with
        these results relates to the services offered by the distributor. In these scenarios we have
        assumed that the higher cost countries simply adopt the cost structure of the lower cost
        countries. However, higher costs may reflect greater provision of advice or face to face
        time with the salesman that would not be possible in a low cost environment. It is not
        possible to quantify the costs associated with different parts of distribution, i.e. sales effort
        versus advice costs, so we have not been able to test this proposition. The implication of
        this is that the cost savings relating to convergence in distribution costs should be used
        with care.

        In this next chapter, we examine whether on-going trends mean that the potential costs
        savings may be realised due to on-going developments in the market or whether existing
        models of distribution and differences in consumer preferences mean that it would be
        difficult to realise these cost savings.


         5.6.       TRANSLATING COST SAVINGS INTO POTENTIAL CONSUMER BENEFITS

        To understand the potential cost savings generated by the scenarios above it may be
        useful to think about what these would mean to consumers if they were passed on. To do
        this we need to make a number of assumptions. For the purposes of comparison we look
        at an average European retail investor investing a lump sum of €5,000 in an equity
        investment fund. We assume that they hold the fund for seven years and get investment
        returns of 7% per annum and that the fund has average charges (i.e. a TER if 1.8%).



                                                                                                    Page 92
Quantification of potential cost savings

September 2006                                                                       CRA International




        Without any change in the costs of the fund, this would result in nominal investment
        returns of around €2,100. In Table 14 below we look at the change in the investment
        returns under different cost scenarios.

        Table 14: Impact of cost savings on investment returns

               Scenario            Lowest cost    Average US        Lowest cost        Average US
                                  European fund       fund        European fund            fund
                                   (production    (production    (total production        (total
                                      costs)         costs)       and distribution   production and
                                                                       costs)          distribution
                                                                                          costs)

          Investment returns          €2,167         €2,177            €2,326             €2,419

          Increased in nominal             2%         3%                9%                 13%
          investment returns as
          a % of returns in the
          base case

        Source: CRA analysis




                                                                                                 Page 93
Competition, integration and costs

September 2006                                                                     CRA International




6.      COMPETITION, INTEGRATION AND COSTS

        Differences in cost across different countries may arise because of a variety of reasons.
        In this chapter we examine the impact that cross-border integration and competition have
        on costs. The aim is to assess first the degree to which these differences explain the
        variation observed in cost and, second, whether the on-going trends identified will result
        in greater integration and competition reducing cost differences.

        In common with other sectors we would expect:

        •   Markets with greater cross-border integration to have lower costs. This is because
            integration exposes providers to a wider range of competitors than only those in the
            domestic market and allows providers to take advantage of scale and specialisation.
            Hence we might expect that those member states that have the most cross-border
            integration would have the lowest costs.

        •   More competitive markets to be those where providers exploit cost efficiencies in
            order to offer their customers the best value products. More intense competition
            could result in either lower priced products or higher quality products. In this study,
            we are focusing on cost efficiency and although we have attempted to standardise on
            particular product types we have not considered in detail if the products offer
            consumers different attributes or result in higher performance.

        When looking at the benefits from competition and cross-border integration, it is important
        also to take into account the fact that competition can occur for the product itself, in this
        case a retail investment fund, but also:

        •   At different stages of the value chain – for example, although in a given market retail
            investment funds may not be a competitive market, the provision of a particular
            activity such as transfer agency could be; and

        •   Between different products - for example, competition between traditional collective
            investment products and exchange traded funds, hedge funds or unit-linked
            insurance vehicles. It is possible to argue that where these products compete more
            intensively, we would also expect them to be more efficient.

        In this study, we have examined competition for different stages of the value chain but
        have not considered the impact of substitute products.

        After reviewing the relationship between competition, integration and costs today, we then
        examine on-going trends in the investment fund market and whether these are likely to
        realise the potential cost savings identified in the previous chapters.

        The primary method for identifying these relationships and their interaction with on-going
        trends was to undertake structured interviews in a sample of representative Member
        States with stakeholders at each stage of the value chain and to combine this with
        background research collected on the market. We have not sought to test whether these
        trends can be supported through robust quantitative data as this is being examined more



                                                                                             Page 94
Competition, integration and costs

September 2006                                                                                   CRA International




        closely by another assignment being undertaken for the European Commission.
        Coordination has arisen between the projects and the data available at the time of writing
        supports the trends highlighted in this section.


        6.1.     INTEGRATION AND COSTS – AT THE FUND LEVEL

        As set out above we would expect markets that are more open to cross-border activity to
        be more efficient. To look at the advantages of greater cross-border integration we can
        either look at:

        •   the differences in cost of countries that have a high degree of cross-border activity to
            those that do not; or

        •   the differences in costs between countries that import funds and those that export
            funds.

        An important issue to be addressed in looking at cross-border integration is the type of
        cross-border activity taking place. For example, we could look at the degree to which
        consumers in one country choose to buy products sold in other countries, however this
        type of cross-border integration is currently very small. Over time consumers could
        become more amenable to purchasing investment products distributed outside the
        consumers’ traditional domestic market, however, given the current behaviour of
        consumers this is a long-term possibility (and is distinct from consumers purchasing
        products that are domiciled elsewhere which they may or may not even be aware of).51

        Instead we focus on traded funds. It is common to discriminate between “true” cross-
        border activity and “roundtrip” trade, where funds are designed for a particular member
        state but domiciled in another. However, there is good reason to look at both types of
        trade, round-trip activity means that a fund provider in one country domiciles a fund
        abroad for its home market. To the extent that this involves locating regulatory activity
        and administration services in the exporting country, this may still be exploiting
        advantages of cost differences between member states. However, this type of activity
        does not increase the number of fund promoters in a particular country and therefore may
        not increase the intensity of competition.




51      See for example evidence in “Qualitative study among cross-border buyers of financial services in the European
        Union Final Report”, Optem, for the European Commission, December 2003.



                                                                                                             Page 95
Competition, integration and costs

September 2006                                                                                                           CRA International




        Figure 31: Cross-border integration versus costs


                                                  Percentage of notified funds domiciled domestically           Total production costs
                                            80%                                                                                    90

                                            70%                                                                                    80
            Measure of cross-border trade




                                                                                                                                   70
                                            60%




                                                                                                                                         Total product costs (bps)
                                                                                                                                   60
                                            50%
                                                                                                                                   50
                                            40%
                                                                                                                                   40
                                            30%
                                                                                                                                   30
                                            20%
                                                                                                                                   20

                                            10%                                                                                    10

                                            0%                                                                                     0
                                                   Belgium    France Germany       Italy    Poland      Spain   Sweden     UK


        Source: FERI FMI data and CRA survey data. The percentage of notified funds domiciled domestically is equal
        to the number of funds domiciled domestically divided by the number of funds available for sale in the country
        (which in turn is equal to the sum of domestically authorised funds and non-domestically notified funds).


        As a proxy for cross-border integration we use the proportion of funds that are available
        for sale in a given country that are domestically domiciled. Countries where most of the
        funds notified for sale are domestically domiciled would be seen as being less integrated
        than countries where most of the funds notified for sale are domiciled in another country.

        In Figure 31 we can examine whether this measure of integration is related to the average
        cost of production. The evidence supports a view that countries with few foreign funds (a
        high proportion of funds domiciled domestically), such as Poland, have high costs and
        those with many foreign funds have low costs. However, we need to recognise that this is
        based on a relatively small number of countries and it is not a relationship that holds in all
        cases since France has low cross-border activity (by this measure) and also has relatively
        low costs.

        Considering our second approach (whether the countries that export funds are cheaper
        than countries that import funds), as was highlighted in chapter 3, we find that those
        countries that are generally associated with cross-border exporting (Ireland and
        Luxembourg) are not actually found to be the locations where production costs are
        lowest. This is surprising since theory would suggest that where there is greater
        integration there should be greater competition and hence lower costs. There are a
        number of reasons that may partly explain why we do not see such a direct relationship.

        •                             The reason for location is preferable for the provider but not the consumer. For
                                      example there may be tax advantages that accrue to the fund promoter rather than




                                                                                                                                    Page 96
Competition, integration and costs

September 2006                                                                                       CRA International




             the unit holders. Alternatively, tax advantages may lead to an improved performance
             rather than lower costs;

        •    The market is in transition and the advantage of location will become apparent over
             time. It may be the case that fund managers have set up their fund range in
             Luxembourg or Ireland and the benefits of this will only materialise over time as funds
             become larger and if mergers are possible;

        •    That the domestic markets are not competitive, resulting in higher profits for the
             provider which are not passed on to the retail consumer (we examine this further in
             section 6.5);

        •    That restrictions on the locations of activities enable high costs to persist because
             countries do not face the full force of competition that might otherwise be expected; or

        •    That location is driven by historic reasons and that location effects or switching costs
             cause providers to remain even if there are no longer specific cost advantages form
             doing so.

        Given that costs are found to be higher in Ireland and Luxembourg, it is a reasonable
        question to ask why it is that companies choose to locate in these two countries despite
        the higher costs that they incur. Some fund managers made specific comments relating
        to the example of Luxembourg as a fund centre. They noted that the cost of labour is
        relatively expensive there and hence, custody and other administrative functions could be
        done more cheaply if it was done elsewhere, but that the tax implications of changing the
        domicile meant this would not be done. Furthermore, fund managers noted that
        Luxembourg has established a reputation such that funds domiciled in Luxembourg are
        accepted elsewhere without concern by continental investors as it has a good track
        record. Detailed consideration of this is outside of our scope of analysis.

        The results regarding the relationship between integration and costs are therefore mixed.
        Although, we find on aggregate integration reduces costs, we find that an increase in
        funds, ceteris paribus, being domiciled in Ireland and Luxembourg, may not lead to a
        reduction in the cost of investment funds. This suggests that there could be other
        benefits from integration, such as increased competition for the domestic market, that
        need to be taken into account.52

        To consider whether ongoing trends in the market will increase integration, and with it
        cost convergence, we need to consider factors affecting the level of cross border
        integration. General barriers to integration such as tax and legal restrictions will be
        important in determining the level of integration that arises. Similarly, consumer
        preferences regarding a “home bias” might affect the choices that consumers are willing



52      An alternative explanation is that the types of fund that are located in Ireland and Luxembourg should not
        compared to the average fund sold domestically. For example, they might represent investment funds with
        added services or might be used when it is not possible to get sufficient scale within the domestic market. In this
        case, we should not compare these funds directly with the average fund in a market as this would under-
        estimate the cost savings that would result.



                                                                                                                  Page 97
Competition, integration and costs

September 2006                                                                     CRA International




        to make. Interviewees agreed that it is not the domicile if the fund that makes a difference
        but recognition of the consumer facing brand. It seems like that this home bias will only
        be overcome with the development of European wide financial services brands. We do
        not consider these issues here but instead focus on issues specific to investment funds
        such as: authorisation and notification; cross-border mergers; and access to customers.


        6.1.1. Authorisation and notification

        The difficultly in getting a fund notified for sale in different member states was identified
        as frustrating cross-border integration. Examining whether there are trends in
        authorisation and notification, we found no evidence of trends in terms of the pure
        regulatory costs that local regulators charge for either authorisation or notification. The
        basis on which regulatory costs are charged varies between member states (some
        charge a fixed fee, others charge according to the size of the fund) and there appears to
        be no pressure on regulators for these fees to convergence (in turn because of different
        ways of funding regulators as a whole in each of the member states). However, there
        was some, weak, evidence of a reduction in costs associated to translation of documents
        in certain countries where either documents could be provided in English or only the
        simplified prospectus needed to be translated rather than all documents regarding the
        fund, but this was not believed to be a major trend.

        Of much greater importance was the speed of authorisation and notification rather than
        the direct cost of it. This was identified by market participants as the most significant
        trend in terms of regulation relating to fund creation and maintenance. Many fund
        managers highlighted that their national regulator was improving in the speed of
        authorisation (Germany, Poland, Sweden) although many also stressed that there was
        still a long way to go. During the main interview programme there was guarded optimism
        that CESR would make progress on the agenda to bring additional harmonisation
        between member states, later interviews, however, suggest that the current proposals
        adopted by CESR were unlikely to solve this and differences between member states
        seems likely to persist. Furthermore, a reduction in the time to market, is unlikely to
        reduce the cost differences in the regulatory compliance of the fund between member
        states. Instead its main impact will come through additional dynamic competition
        between funds which are able to access markets more easily.

        The cost of regulation relating to fund creation and maintenance was found to be small at
        around 2 basis points. Trends in this area are therefore unlikely to have direct significant
        impacts on the overall cost of investment funds faced by consumers and there were very
        few areas where interviewees or other evidence suggested strong trends that would drive
        cost changes in this part of the value chain.


        6.1.2. Fund consolidation

        A second mechanism for increased integration is the consolidation of existing funds. By
        using the same fund to serve multiple markets it is possible to have larger funds,




                                                                                             Page 98
Competition, integration and costs

September 2006                                                                                       CRA International




        spreading the fixed costs across more consumers and a larger value of assets resulting in
        lower charges.53

        In examining the evidence on fund mergers it is important to distinguish between
        domestic and cross-border mergers since lessons from the former can be used in
        understanding the benefits from the latter.

        Much has been made of the difficulty of fund consolidation for example regarding the
        need for shareholder agreement and issues of whether mergers represent a taxable
        event.54 Furthermore, mergers are not without cost since some unit holders will take the
        opportunity to sell their investment at the point of mergers because they can do so without
        exit charges being imposed. Consumers selling units reduces the overall level of assets
        under management and therefore reduces the gains from consolidation

        However, over the last five years, there have been significant numbers of domestic
        mergers with around 8% of funds merged or closed. Indeed, the consistent view from the
        interviews undertaken for this assignment was that mergers within a domicile were
        straightforward and barriers have been reduced over time:

        •    Where mergers have been difficult, alternatives, such as schemes of arrangement
             have been used as in Ireland;

        •    Most jurisdiction have set out the rules under which mergers are possible; and

        •    Where tax rules have been problematic, such as capital gains tax in France, the rules
             have been changed to facilitate consolidation.

        Nonetheless, in some member states domestic mergers of funds are not always
        straightforward and Sweden was highlighted here due to constraints regarding the
        investment objectives of merging funds. This is apparently because of a stronger concern
        in Sweden than other countries that consumers may end up being invested in a fund with
        different aims to that which they originally invested in. In other countries such as Spain,
        mergers may have occurred because fund sizes are too small to meet regulatory
        requirements and hence the fund is forced to close or to merge.

        Figure 32 below provides information on merged and closed funds by member state and it
        is clear that mergers are occurring in all the markets under examination. The difficulty of



53      Some fund managers (mainly in the UK and Sweden where there is a greater focus on equity funds) noted that
        diseconomies of scale may result where very large funds own substantial proportions of individual stocks and it
        becomes increasingly difficult to make rapid trades. In addition, in larger funds, it is more difficult to beat the
        market, or as some interviewees described it “exploit the alpha”, because the transactions may be of such a size
        that they begin to move the market and this may place a limit on the benefits of fund consolidation. (The
        performance return made by a fund is equal to the market return (beta) and the divergence from the market
        return (alpha); the larger the fund is, the more it is likely to replicate the whole market and hence the closer the
        return would be expected to be to the overall market return and the smaller would be the divergence.)

54      For example: In Luxembourg, 100% (in Dublin, 75%) of the votes cast must agree to the merger and these
        votes must represent 50% of the outstanding shares.



                                                                                                                   Page 99
Competition, integration and costs

September 2006                                                                                   CRA International




        merging funds in Sweden due to the requirement for funds to be more homogenous is
        seen in the lower proportion of mergers that arise.

        Figure 32: Mergers/closures as a percentage of funds on the market


            12%




            10%




            8%




            6%




            4%




            2%




            0%
                   Belgium   France    Germany      Italy     Poland      Spain      Sweden       UK      International



        Source: European fund market data digest 2006, FERI FMI. Note that this refers to mergers or closures in 2005
        as a percentage of funds in 2004. International is a category defined by FERI as relating to those groups with
        funds domiciled in Luxembourg and Dublin that are not otherwise included within the data of specified European
        markets.


        Although mergers are arising, in all of the countries bar France, Italy and the UK, fund
        launches are larger than mergers and closures leading to an increase of around 600
        funds in the countries under examination between 2004 and 2005 (compared to a total of
        20270 funds).55 Hence mergers are not resulting in a decline in the number of funds or
        significant growth in average fund size. There are a number of explanations for this:

        •     Competition may be focused on new fund creation because consumers (or
              distributors) are constantly looking for new products. Frequent changes to the tax
              regimes were also said to have encouraged this; and

        •     Fund managers may prefer to have new funds as this allows them flexibility regarding
              how performance is reported and allows them to attract new consumers even if their
              existing products are not performing as well as the rest of the market.

        Where large scale mergers of funds have been seen, this has typically arisen when
        domestic banks have merged and their fund management arms have then also been
        merged after this. For example, in Spain, the largest banks, Santander and BBVA are the



55      European fund market data digest 2006, FERI FMI.



                                                                                                             Page 100
Competition, integration and costs

September 2006                                                                                        CRA International




        results of various domestic banking mergers with the fund management groups being
        merged on the back of this.

        However, when considering cross-border mergers, many of the changes observed in
        relation to easing domestic mergers do not carry across to cross-border mergers. For
        example, when considering tax implications, while domestic mergers would often not be
        identified as a taxable event, a cross-border merger would be. In addition, some
        countries have restrictions preventing the transfer out of funds from their jurisdiction.

        Similarly, although domestic banking mergers have arisen, the EC has noted the lack of
        cross-border banking mergers.56 Since banking mergers have often led to the merging of
        fund management companies and in turn the individual firms, a lack of cross-border
        banking mergers is likely to be acting as a barrier to cross-border fund mergers.

        In addition, restrictions imposed on cross-border funds mergers may also be preventing
        merger and acquisition activity from arising at the level of the fund management
        company. In turn these restrictions reduce the competitive dynamic that arises from
        replacing a poorly performing or high cost management company with a more efficient
        one.

        Furthermore, factors preventing consolidation from arising at the level of the fund reduce
        competitive pressure by preventing fund managers from exploiting economies of scale in
        order to reduce costs and offer lower priced or higher quality products to consumers.
        They therefore lead to higher costs in comparison to what would otherwise be observed.

        Cross-border mergers have proved extremely difficult with relatively few cases where this
        has been possible and with some countries where these are seen as impossible. Even
        where fund management groups have merged, this has still led to many fund managers
        with ranges of funds in multiple countries. In particular, fund ranges in Luxembourg or
        Dublin alongside domestic ranges in Italy, Spain and Germany are very common.

        With the exception of in distribution, fixed costs or economies of scale were identified in
        all other areas of the value chain. Merging funds has the clear potential to release very
        significant cost savings generating by economies of scale and the removal of duplicate
        fixed costs. An increase in mergers would be equivalent to examining the impact of the
        cost savings from producing larger funds as explained in Table 17 in section 5, where
        simply moving from today’s fund size to the average EU fund size for those member
        states who are below average would save some €150 million.57 A similar cost saving
        would be made if there was a reduction in the number of funds across all countries by
        10% leading to an equivalent increase in the funds under management.

        This is of particular concern for existing or legacy funds. While new funds can be set up
        in one jurisdiction and hence benefits from economies of scale exploited as that one fund




56      See for example, Sector Inquiry under Article 17 Regulation 1/2003 on retail banking.

57      This is calculated by comparing the starting point for the production cost in Table 17 with that in Table 10.



                                                                                                                  Page 101
Competition, integration and costs

September 2006                                                                                       CRA International




        is sold across the whole of Europe, existing funds that are already in multiple jurisdictions
        can not be merged. Given that new funds take time to establish and to grow, the speed
        at which cost savings can be gained through new funds alone is reasonably long. By
        contrast reducing barriers to cross-border mergers could realise immediate cost savings
        across the whole fund sector. However, without any changes in regulation, we would not
        expect an increase in cross-border mergers, the current trend of few such mergers would
        persist with the result that cost savings will not be realised.


        6.1.3. Cross-border distribution

        Fund management companies are polarised regarding their behaviour on cross-border
        distribution. Some companies are clearly aiming to be pan-European providers of
        investment funds and are actively seeking to market their funds in a number of member
        states with this number typically increasing over time. More usually these will be the
        larger groups and they will often have followed strategies such as domiciling all of their
        funds in one country (typically Luxembourg or Ireland) in order to centralise all of their
        administrative functions.58

        Other companies are simply focused on their own domestic market and have not notified
        their funds in other countries. In some cases this was because the domestic fund
        management industry was providing sufficient opportunities for expansion. In the main,
        however, their reason for doing this was often linked to the difficulties involved with
        accessing distribution networks in other countries or because different regulatory
        conditions meant that it was not worth the cost of doing so. Indeed, difficulty accessing
        distribution was also mentioned by those companies who were already active in many
        markets.

        The ability for new entrants to gain access to distribution is not specifically related to
        cross-border provision of funds. As explained in section 4.4, there was no evidence that
        gaining access to non-proprietary distribution channels was more expensive for cross-
        border funds compared to domestically domiciled funds. This suggests that access to
        distribution is a more general issue for new entrants and for those who do not own
        proprietary distribution channels rather than for cross-border provision per se. Hence we
        examine this issue further in section 6.4.3, where the potential for cost savings is
        considered.


        6.2.      INTEGRATION AND COSTS – ALONG THE VALUE CHAIN

        Even if funds are not being traded from one member state to another, gains from
        specialisation and scale can occur within different stages of the value chain. For



58      Some fund managers appeared to have a regional focus. For example, the Benelux countries are increasingly
        seen as a regional market (with banks and their fund managers operating in all three countries). Swedish fund
        managers were focused on Scandinavia (including Norway despite it being outside of the EU) and the Baltic
        states of Estonia, Latvia and Lithuania. In addition, cross-border integration is not limited to the European Union
        and as well as Swedish fund managers integrating with Norway, Spanish fund managers focus on Latin America
        and the UCITS brand is used in multiple other locations.



                                                                                                                Page 102
Competition, integration and costs

September 2006                                                                                    CRA International




        example, if one particular country has a competitive advantage in transfer agency, this
        could result in better value funds thus prompting cross-border trade in the fund itself.
        Alternatively, in theory, fund providers in other countries could use transfer agency
        services from that country even though funds remain domestically domiciled. Both
        mechanisms offer the opportunity to reduce costs. In the previous section we considered
        integration of the fund and here we consider integration along the different stages of the
        value chain.

        In some parts of the value chain integration seems to have already occurred.               For
        example, the asset management component of the value chain is the area where there is
        the least relationship between location and the domicile of the fund and where it is
        already seen as a globalised market. Gains from integration in regulatory compliance
        related to fund creation and maintenance already seem to have been exploited by some
        companies who choose to use a centralised team to undertake this activity (although they
        continue to use local lawyers). Other firms prefer to use local regulatory teams. The
        difference is strategy seems to vary more by company than by member state. Finally,
        distribution, by its very nature, is based in the country where it is sold and thus integration
        of distribution across countries is not a relevant concept.59

        However, we find that there are locational barriers to integration in the administration part
        of the value chain and these relate to:

        •    Regulatory barriers; and

        •    Barriers due to market standards

        To understand whether we might see more cross-border activity in particular activities we
        need to consider the barriers that prevent this activity today and whether these barriers
        might change over time.


        6.2.1. Regulatory restrictions on the location of activities

        One of the rules of UCITS funds is that the depositary must be located in the same
        country as that in which the fund is domiciled. In some member states this role can be
        carried out through a local branch of a depositary bank, which reduces, but does not
        eliminate, the barrier. In addition, although activities such as oversight must be
        undertaken in the country in which the fund is domiciled, in effect many other activities
        can be delegated and hence efficiencies may be achievable through delegating custody
        to a subsidiary or branch in the relevant jurisdiction.

        There may be regulatory reasons for this restriction to ensure that the regulatory authority
        with responsibility for regulating the fund also has the ability to regulate the custodian
        because they are in the same domicile. However, in terms of integration of the custodian




59      In theory and in the very long term, consumers may be willing to purchase investment funds through distribution
        channels in a different member state to the one where they are located. However, there is no evidence of this at
        present.



                                                                                                              Page 103
Competition, integration and costs

September 2006                                                                    CRA International




        market, it functions as a barrier to entry since custodians in other countries are prevented
        from offering their services direct to the fund.

        In an integrated market, we would expect to see overlap of providers in each member
        state. However, in many markets, large domestic banks dominate custody provision,
        although in other markets a range of international banks offer custody services. The
        difference is likely to be linked to the size of the market opportunity:

        •   In many countries fund management companies are part of larger banking groups
            and typically use the bank within the same group for custodian services. This
            reduces the potential market that is contestable and hence reduces the incentive to
            enter a particular domestic market. This would be expected to remain the case even
            if regulatory barriers were removed; and

        •   Global custodians may be progressively entering markets and hence have focused
            on the large and growing markets before they enter smaller markets. However, the
            speed of this approach may be reduced because of the need to achieve a certain
            volume of business in each individual country rather than in Europe as a whole in
            order to justify entry on commercial grounds.

        The restrictions on custodians was identified as particularly frustrating to fund managers
        where they have funds that invest in a different location to the one in which the fund is
        domiciled since the custodians in their domestic market may be less good value in other
        locations because they do not have scale elsewhere, yet, due to the reasons highlighted
        above, global custodians have not entered their domestic market.

        Nonetheless, many fund managers expected custody markets to become increasingly
        concentrated because of the increased fiduciary duties towards shareholders under
        UCITS III. Hence some small providers of custody services may exit the market.

        In addition to restrictions on the location of the depositary, some member states also
        impose additional restrictions regarding the activities that need to be undertaken in their
        member state in the form of “central administration” or “minimum activities”. Again these
        operate as barriers to integration, although outsourcing of functions (including within the
        same company rather than an external provider) reduces its impact.

        Since both of these issues (restrictions on the depositary and restrictions on the location
        of other administrative activities) are regulatory barriers, market trends are not expected
        to change them. Thus, while there may be potential cost savings arising from these
        activities (custody costs around 8 basis points and all of fund administration around 23
        basis points for an average fund), no cost savings are expected here on the basis of
        current market trends.


        6.2.2. Market standards restricting the location of activities

        A reduction in integration can arise through the lack of standardisation of processes
        throughout the value chain. Thus it is possible that, given the processes that are in place,
        the value chain may be competitive in each individual market, but gains from competition
        are not being fully exploited because of an inability to standardise across markets. This is


                                                                                           Page 104
Competition, integration and costs

September 2006                                                                                   CRA International




        seen as a particular issue in regard to fund order processing and accessing the Central
        Securities Depositary (CSD).

        For example, different processes are used in different countries regarding the transfer
        agency role and how it interacts with the CSD, daily cut off times, trade dates, number of
        decimal places used in cash orders etc. The custodian of the fund needs to be able to
        interact with the CSD (such as Euroclear and Clearstream) which undertakes the
        centralised settlement of transactions and is ultimately responsible for collective safe-
        custody of assets i.e. the holding and transferring of securities.60 Depending on the local
        rules, this may require having a local sub-custodian who itself has access to the local
        CSD.61

        For funds that are traded cross-border this may mean that processes must be in place to
        meet the requirements of all the countries in which the fund is traded This is makes
        undertaking cross-border trades more difficult since interfaces between different
        processing approaches are not always straightforward with differences arising in business
        practices such as valuation times that may vary from country to country as well as
        different communication methods including both manual and electronic methods being
        used. As well as adding to the difficulty of the trade, this also potentially leads to
        duplicative costs being incurred to conduct similar, but not identical, processes.

        Standardising processing protocols across countries is not an easy or straightforward
        approach to take. In the short term, solutions that allow the interaction or interfacing of
        different processes such that the different processes in different countries in order to
        enable transactions to arise may bring gains. In particular, solutions that ensure that
        consumers are not unduly impacted by these different approaches would appear to be the
        preferred solution i.e. where electronic solutions enable the interaction rather than forcing
        consumers to transact in a different way for cross-border trades compared to the way that
        they are used to transacting for domestic trades. However, such solutions would
        inevitably involve adding additional costs to allow the interaction, although it would bring
        the benefit of increased competition and access to a wider selection of funds for
        consumers in any given member state.

        In the longer term, standardisation of processes is preferable. This would remove the
        need for those costs that are incurred in order to allow the interaction of processes (since
        all processes would be the same) and it would also be expected to increase competitive
        pressure since all member states would be using the same processes and suppliers of
        these processes would be competing in a larger market.




60      Euroclear acts as the CSD for the UK, Ireland, France and the Netherlands with the Belgium CSD having joined
        the Euroclear group in January 2006. Clearstream acts as the CSD for the German and Luxembourg domestic
        markets.

61      The use of sub-cutodians as opposed to gaining direct access to the CSD would also depend on commercial
        issues such as whether the sub-custodian would offer additional value added services such as speaking in a
        non-local language or providing their own capital. Interview with service providers offering custody services,
        March 2006.



                                                                                                            Page 105
Competition, integration and costs

September 2006                                                                               CRA International




        There is also a concern that standardisation will not result in the standards of the most
        efficient country being adopted. Hence there is a risk that processes are standardised on
        the countries which currently have the most cross-border trade even though these
        countries may not have the processes which are either the lowest cost processes or the
        best processes to use.

        Notwithstanding this concern, the standardisation of processes is a clear goal among
        market participants and progress is being made to this end.62 It is therefore clear that
        existing market trends are likely to lead to cost savings over the medium to long term
        through the initiatives of the fund industry and the industry believes that there is no need
        for additional intervention in order for this to arise.63 However, on the basis of quantitative
        information collected for this project, it has not been possible to estimate the value of
        these cost savings as a separate component from other functions of administration.


        6.3.     COMPETITION AND COSTS – AT THE FUND LEVEL

        As the objective of this study is to look at the relationship between competition and costs
        rather than identify competition issues in a particular country, we have not sought to
        undertake a market definition exercise or conducted analysis that would allow country-
        specific conclusions of this kind to made. Rather we are looking for evidence of whether
        competition encourages efficient models of investment fund provision.

        Competition would be expected to lead to cost efficiencies through both the supply-side
        and the demand side. On the supply side, competition between companies who providing
        investment fund services increases efficiency (such as through increased productivity and
        reductions in costs) through new entry and its threat as well as the potential for capital
        market competition through exposing financial services companies’ management teams
        to the threat of replacement by other teams from different companies. On the demand
        side, competition driven by consumers would involve consumers searching out the
        appropriate products and the providers best able to deliver these services as well as
        those providers offering services that consumers value. Issues to do with demand-side
        competition are considered in section 6.5 where we examine whether cost savings will be
        passed through to consumers


        6.3.1. Intensity of competition

        To measure the intensity of competition it is common to look at the level of prices in a
        market or the degree of price dispersion between countries. Although we have collected
        data on TERs (and the comparison of them can be found in the country appendices), we
        have not put substantial weight on this analysis. This is because differences in TERs may
        reflect differences in product types with some markets having products that are more
        differentiated others and hence some of these products may be more costly to produce.



62      For example, the EFAMA Fund Processing Standardisation Group, established in 2003 published
        Standardisation of Funds Processing in Europe in February 2005.

63      See Report of the Expert Group on Investment Fund Market Efficiency, European Commission, July 2006



                                                                                                        Page 106
Competition, integration and costs

September 2006                                                                                        CRA International




        Simple comparisons of TERs across countries are therefore not a robust indicator of the
        degree of competition. Hence we have examined more indirect measures of the intensity
        of competition.

        Most markets were perceived to be competitive by industry participants and regulators.
        The box below sets out particular issues arising in the each of the member states.

        Country-specific issues64

        Belgium: the number of new fund products coming on to the market, the degree to which they are quickly
        copied and asset management companies considering charges of competitors when setting their own charges
        supported a view of strong competition. On the other hand, captive channels, the lack of the education of
        investors and the difference in structures of different products making them difficult to compare (especially for
        guaranteed funds) might reduce the level of competition.

        France: The industry is seen to be competitive, with both large and very small players doing well in the market.
        Large players include global players such as BNP Paribas whereas the smaller players are usually SME fund
        managers which cover certain niche funds. Overall around 500 management companies compete in France.
        The French asset management industry is also seen as very diversified and comprises many entrepreneurial
        firms and asset managers, subsidiaries of banks, insurance companies, financial and asset management
        groups, French and foreign.

        Germany: There are around 80 different investment fund companies in Germany. A large number of these
        companies distribute German domiciled funds and funds domiciled in Luxemburg.                 Evidence from fund
        providers, suggests competition in the investment fund market is especially strong because the percentage of
        German households’ assets in investment funds is relatively low compared to other European countries and
        investment fund companies are competing fiercely to capture a share of the market potential.

        Italy: Banks dominate the sector since the largest fund management companies are part of banking groups (e.g.
        Intesa, Sanpaolo – IMI, and Fineco) and use the proprietary network of bank branches as the main distribution
        channel. However, this is changing because Credit Agricole completed the acquisition of Nextra in 2005, which
        is one of the main Italian fund promoters in Italy and was previously affiliated to the banking group Intesa.

        Poland: At the end of 2005 there were 23 fund managers operating in Poland. However, this market is evolving
        quickly, with 5 players entering the market in the last few years as the Polish market is growing very fast and
        growth in demand is seen as strong.

        Spain: Banks and savings banks dominate the fund management industry, since fund management companies
        are often part of a banking group (as in the case of the four largest fund promoters Santander, BBVA, La Caixa,
        and Caja Madrid).

        Sweden: The four largest fund management companies are owned by the four largest Swedish banking groups,
        which are “universal” banks and carry out their activity in several financial sectors (e.g. investment funds,
        mortgages, insurance, etc), as well as in the core “retail banking” sector. After the launch of the pension reform
        and the creation of the PPM, Swedish investors have gained greater familiarity with funds promoted by
        international players and fund supermarkets. Their market share in terms of assets under management is still
        small, but fund management companies which are not part of a banking or insurance group are being
        successful in terms of net sales and managed to slightly erode the dominant position of banks.




64      See country appendices for further details.



                                                                                                                  Page 107
Competition, integration and costs

September 2006                                                                                   CRA International




        UK: The UK remains unusual in that the majority of funds are distributed through advisers who are not part of a
        banking group and instead independent advisers continue to flourish resulting in the most fragmented fund
        management industry of the ten countries under investigation.


        Drawing the evidence from across member states suggests there were two main areas of
        common concern (excluding those relating directly to consumers):

        •    Where banking markets were concentrated and banks sell predominantly proprietary
             products there were concerns regarding the concentration of the industry; and

        •    Banking groups owning the fund managers and administrators and custodians were
             thought to put relatively little pressure on costs.


        6.3.2. Industry concentration

        In some markets, a smaller number of providers might indicate that the intensity of
        competition is low. Hence markets that are highly concentrated might be at risk of having
        lower cost efficiency. There are substantial differences in the fragmentation of the fund
        management industry in different member states which largely reflects the importance of
        the banking distribution channel in many markets. Where distribution is focused on a
        relatively small number of banks, who have traditionally sold the product of their
        proprietary fund manager, this has resulted in these fund managers having an equally big
        market share.

        Figure 33 shows the relationship between concentration and production costs. The figure
        shows there is no clear relationship between the markets with the greatest concentration
        and those markets we have found to have the highest costs. This is likely to reflect the
        fact that while concentration is a useful indicator it does not directly measure the intensity
        of competition. In addition the lack of relationship may reflect different countries gaining
        from economies of scale to a greater or lesser extent. For example, given the size of the
        Swedish market (€ 53 billion funds under management) and the level of concentration this
        may have led to the main Swedish fund providers exploiting economies of scale to a
        greater extent than in Belgium which is even more concentrated but has a much smaller
        market (€18 billion assets under management).




                                                                                                             Page 108
Competition, integration and costs

September 2006                                                                                                                          CRA International




        Figure 33: The relationship between concentration and production costs


                                                                    5 player concentration             Total production costs

                                  90                                                                                                               100%


                                  80                                                                                                               90%


                                                                                                                                                   80%
                                  70

                                                                                                                                                   70%
                                  60




                                                                                                                                                          % player concentration
          Total production cost




                                                                                                                                                   60%
                                  50
                                                                                                                                                   50%
                                  40
                                                                                                                                                   40%

                                  30
                                                                                                                                                   30%

                                  20
                                                                                                                                                   20%

                                  10                                                                                                               10%


                                   0                                                                                                               0%
                                       Belgium   France   Germany    Ireland          Italy   Luxembourg    Poland         Spain   Sweden    UK




        Source: CRA analysis of survey data and EFAMA data


        6.3.3. Market structure

        Another factor that might effect the intensity of competition is the market structure. The
        structure of the market regarding the level of vertical integration varies between member
        states.     Hence it is worth considering whether vertical integration or vertical
        disintermediation is related to cost structures. As highlighted in chapter 4, there are
        significant differences in the method of distribution between member states.

        In particular, the UK is unusual because none of the top five fund promoters primarily
        distribute through a proprietary distribution channel and hence the UK displays a level of
        vertical disintermediation not seen in other countries. In contrast, the majority of products
        in Italy and Spain are distributed through bank branches of the same group as the fund
        manager.

        At current fund sizes, the UK is one of the cheaper countries for both distribution and total
        production costs (although it has a similar cost structure to other countries that have a
        vertically integrated structure). Given that the UK is the only country in which non-
        integrated channels are the majority, there is only a limited amount of evidence on this
        issue, but it does not provide strong evidence for the suggestion that separation of the
        fund management and distribution activity necessarily results in more cost efficiency.
        Indeed, chapter 5 does not find that the UK is the cheapest country when the impact of
        scale is removed. Arguably, this finding could reflect costs of distribution being lower
        where there are a smaller number of distributors to deal with in universal banking models.

        In distribution, there are very large differences in the cost of distribution between member
        states which is likely to reflect the activity being undertaken varying by country as well as
        the intensity of competition. Interviewees universally agreed that distributors, because
        they have the consumer relationship, are typically in a strong position and that this
        component of the value chain was the least competitive. Indeed concentration in


                                                                                                                                                  Page 109
Competition, integration and costs

September 2006                                                                      CRA International




        distribution was believed to be higher in most countries than the level of concentration in
        other parts of the value chain. However, differences in the cost of distribution will also
        reflect different consumer preferences in each member state regarding the method of
        distribution as well as consumer protection measures. Furthermore, it is not possible to
        work out the proportion of distribution costs that relates to issues such as advice and
        therefore caution should be taken when comparing distribution numbers especially when
        calculating cost savings.

        Therefore it is unclear whether changes in the market such as trends to more open
        architecture and vertical separation in some markets (e.g. in Italy we are seeing the
        separation of banking and fund management) will directly lead to large gains to cost
        efficiency.


        6.4.     COMPETITION AND COSTS – ALONG THE VALUE CHAIN

        So far, we have considered the relationship between competition in the overall investment
        funds market and cost. However, competition also arises within the different stages of the
        value chain. In asset management and administration there is evidence that fund
        managers are using the services of other providers and even outsourcing this entire
        activity. In addition, there is evidence of trends in the use of different distribution
        channels which may have an impact on costs overtime. We examine these issues in turn.


        6.4.1. Fund administration and client administration

        The main trend in terms of fund and client administration that has been identified is that of
        outsourcing activities to other companies. Linked to this is the development in which
        distributors, particularly fund supermarkets, are increasingly undertaking more of the
        administrative roles – we discuss this in section 6.4.3 below. Outsourcing is arising
        because many of the fund and client administration activities are gradually being seen as
        commodities for which IT intensive solutions can be found. Related to this is that these
        activities tend to have considerable economies of scale that can be exploited.

        However, outsourcing is not without cost since transaction costs arise in selecting the
        provider of services and arranging contracts with them. For example, some fund
        managers argue that cost efficiencies gained through outsourcing would not compensate
        for the loss of close proximity of all activities. In addition, in some markets it is necessary
        to pay VAT on the outsourced services (which at the moment is not paid if services are
        done internally since financial services companies do not charge VAT).

        Furthermore, outsourcing itself may not be competitive if there are large switching costs.
        For example, in the custodian market, switching costs would be incurred in areas such as
        the board, re-registration, audit, having to transfer all the assets from local markets to the
        new custodian, updating records including underlining shareholder records, and
        regulatory time. The manner in which this cost would be paid would vary, but some fund
        managers estimated that it might cost around 3-4bps to change the custodian. In some
        countries this could not be charged to the fund, or charged back to the shareholders of
        the fund and hence it would be the fund manager itself that bears the cost of the change
        where the fund receives the benefit in terms of lower custodian prices. In such


                                                                                              Page 110
Competition, integration and costs

September 2006                                                                                         CRA International




        circumstances the incentive to switch custodian for the benefit of consumers may be
        reduced.

        The degree to which outsourcing is currently undertaken varies considerably between
        companies and between countries (although these may also be linked). Outsourcing will
        also depend upon whether or not there are economies of scope as well as scale.

        Company size

        Small companies have most to gain from the economies of scale that can be achieved by
        outsourcing. Indeed, on average 33% of administrative activities is outsourced by
        independent asset managers compared to 17% for dependent asset managers (i.e.
        where the asset manager is part of a wider financial group).65 Furthermore, additional
        requirements under UCITS III has made administration more complex which may
        therefore induce small and medium companies to outsource activities further in the future.
        However, amongst the larger fund managers differences in strategy are seen:

        •    Some fund management companies want to retain all parts of the value chain in-
             house including administrative functions because they prefer an integrated approach
             and so that systems are designed specifically around their own requirements. They
             also believe that they can exploit many of the economies of scale internally by
             centralising their activities.

        •    By contrast, other fund managers do not see back office functionality as part of their
             core competence. They argue that outsourcing offers a competitive choice of
             providers (unlike an integrated approach) and switching (or its threat) is sufficient to
             ensure that they receive good quality services.

        Given that these differences in outsourcing among larger companies appear to be
        coherent strategies, it seems likely that such differences in business models will persist.

        Member states

        Differences in outsourcing are also observed between member states, although much of
        this is linked to the size and concentration of fund management as a whole. For example,
        in both Spain and Sweden the degree of outsourcing is limited with administration mostly
        carried out in-house. In both cases this may reflect large fund managers being part of
        banks that may have already reached critical mass in administrative activities.
        Furthermore, the fact that large players dominate fund management may also explain
        why foreign players have not entered some markets to offer outsourced activities.

        In other markets, such as Italy, there have been regulatory constraints on the degree to
        which outsourcing can be used as (in the past the NAV had to be calculated internally).
        Similarly until recently there were restrictions on outsourcing parts of the value chain in




65      This is based on the weighted mean for those countries in the European Union for which data is available and
        relates to back-office activities, marketing, printing of prospectuses, etc. It is based on Feri FMI data.



                                                                                                                     Page 111
Competition, integration and costs

September 2006                                                                                         CRA International




        Germany. Where such regulation has been removed allowing outsourcing, we have seen
        fund managers take advantage of this relatively quickly. However, there was no evidence
        that these countries had significantly higher administration costs than countries without
        such restrictions – this may suggest that there are country specific administrative costs.

        In both Luxembourg and Ireland outsourcing is very common with over 80% of back office
        administration outsourced in the latter.66 In both Luxembourg and the UK there has been
        a considerable increase in the use of outsourcing. For example, in Luxembourg, the
        percentage of funds whose administration is outsourced rose from 20% in 1994 to 57% in
        2002.67 In the UK, the use of third party administrators has increased from 30% in 1994
        to over 75% in 2004.68 Despite the apparent cost savings from outsourcing, there was no
        evidence that those countries with more outsourcing had lower administrative costs.

        Economies of scope

        Economies of scope arise when the cost of undertaking two activities reduces because
        they are done together. In the context of this study, economies of scope might arise
        through combining separate stages of the value chain together. Many fund managers
        and administrators believe that there are economies of scope in the provision of
        administrative services. However, there is less agreement on the precise activities that
        lead to the economies of scope arising.

        For example, some service providers argue that combining fund accounting and custody
        saves costs. Bundling these elements enables the service provider to leverage off a
        better relationship with the fund manager, and to offer large discounts compared to when
        only one of the services is provided. In particular, they argue that undertaking custody
        services means that there is a fiduciary responsibility for the calculation of the NAV (the
        main element of fund accounting). If this duty involves checking the methodology or
        implementation of the NAV calculation then the custodian may end up replicating part of
        the activities of the fund accountant. Hence the fund may be in a better position if the
        custodian also performed the role of fund accounting. Furthermore, custody and fund
        accounting services may have synergies between them as they are both transaction
        based – undertaking both of these activities within one service provider enables any one
        trade to go through an IT system only once.

        Others argued there were synergies between transfer agency and fund accounting
        although these are smaller than between custody and fund accounting because there is
        not a need for the transfer agent to understand each portfolio trade which is required by
        the fund accountant.




66      This relates to back-office activities, marketing, printing of prospectuses, etc. It is based on Feri FMI data.

67      Key dynamics affecting the outsourcing market, Mc Cumiskey, E. European Fund Administration, 2003. The
        article is available under http://www.efa.lu/EN/research/articles.html.

68      http://www.ftmandate.com/news/fullstory.php/aid/27/Evolution_of_solutions.html



                                                                                                                   Page 112
Competition, integration and costs

September 2006                                                                      CRA International




        One way to look at economies of scope is to look at the proposition offered by different
        out-sourcers where multiple different models are observed:69

        •    BNP Paribas offers a modular package of services but believe some bundling is
             important. They would want to combine custody and fund accounting and would not
             favour doing transfer agency on its own.

        •    European Fund Administration by contrast does separate out the provision of transfer
             agency (TA) services arguing that offering TA services independently of fund
             accounting is advantageous due to the increasingly specialised nature of the TA
             service. They note that different clients take up different service offerings in
             combination.

        •    Brown Brothers Harriman offers different services to different clients combining TA
             with fund accounting for some of their clients but not for others.

        That there are different models being used suggests either that the economies of scope
        are small (since the disadvantage of picking the wrong model are small) or, more likely,
        that the market is currently in a state of transition. It seems plausible that most of the
        focus in administration so far has been on obtaining cost reductions through the benefits
        of economies of scale. As competition continues to exert pressure on costs, we might
        expect to observe greater focus on obtaining further cost reductions through exploiting
        economies of scope and thus to see a convergence towards the models which exploit
        such economies, or alternatively to see increased specialisation where such economies
        do not exist.

        Together fund and client administration cost between 20 and 32 basis points for average
        equity funds in member states with the exception of Poland where high transfer agency
        costs increase this to 53 basis points. The variation between member states and the not
        insubstantial level of costs suggests that the potential for cost reduction exists.

        Overall, it is likely that outsourcing will continue. During the course of the project, no fund
        management company that was currently outsourcing its activities indicated that they
        were likely to bring these functions back into in-house provision (with the exception of
        those under going mergers) while many companies indicated that they were actively
        considering outsourcing opportunities. Furthermore, since a number of countries have
        only relatively recently relaxed the rules on outsourcing, this will lead to renewed interest
        in outsourcing in a number of countries. However, it is not realistic to believe that all
        companies will adopt an outsourcing approach since they may be able to exploit
        economies of scale internally.

        The continuation of outsourcing is likely to reduce costs. Despite the result that
        outsourcing and administrative costs are not linked at the country level, a reduction in
        costs is nonetheless expected because individual companies would not outsource unless
        they believed significant cost savings could be made. Therefore, although we would



69      Information is taken from websites of the providers.



                                                                                              Page 113
Competition, integration and costs

September 2006                                                                     CRA International




        expect to see a reduction in costs because of outsourcing, we would not necessarily
        expect to see convergence in costs between countries.


        6.4.2. Asset management

        The market for the provision of asset management services appears to be competitive.
        As discussed in section 3.4, there was no discernable difference in the cost between
        member states once scale has been taken into account. Whilst surprising, this may well
        reflect the fact that asset management is now a global business, outsourcing arises in
        certain funds and hence competitive forces are already being brought to bear on this part
        of the value chain. Alternatively, it is possible that a minority of fund managers continue
        to use domestic fund management that results in a detrimental performance impact not
        measured directly for this assignment rather than an effect on cost per se. Pooling offers
        the opportunity to manage these funds on a more efficient basis, however, cross-border
        pooling appears unlikely to occur without significant encouragement from the regulatory
        authorities.

        There are a number of significant trends that will encourage further efficiencies in asset
        management:

        •   Increased use of centralised dealing desks appears inevitable;

        •   Convergence of asset management on the specialist financial centres within Europe
            and globally appears likely to continue; and

        •   Increased use of outsourcing for fund managers who cannot themselves exploit
            international specialism.

        First, there is a trend to the use of centralised dealing desks which involve all trades from
        all funds being centralised within the fund management company and then this desk
        sending the orders for transactions to arise. The advantage of doing this is that larger
        trades can be sent with better prices being extracted due to fewer transaction-based
        costs and also because larger trades themselves typically attract better prices on the
        actual stock (which in turn may lead to better performance).

        Second, there has been a substantial trend to consolidating asset management functions
        in a small number of locations. This is arising both within companies (as they centralise
        their activities) and also across companies (as regional hubs have developed such as
        London.) Hence many of the benefits of competition in asset management may already
        be exploited through these methods.

        Within companies, group-wide strategies are common in which, for example, all French
        investments would be managed in Paris, all German investments would be managed in
        Frankfurt, all UK investments would be managed from London and all US investments
        would be managed from New York regardless of either the domicile of the fund or of the
        domicile of the customer. In this way, larger fund management companies gain the
        benefits of economies of scale through the way in which they structure their businesses
        internally.



                                                                                            Page 114
Competition, integration and costs

September 2006                                                                      CRA International




        Across companies, regional hubs such as London, and Frankfurt, Paris are well
        established centres for asset management. Given that there appear to be network effects
        from having centres of asset management (the necessary infrastructural arrangements
        and large numbers of highly trained staff) this means that new companies deciding where
        to locate are likely to choose existing centres rather than to move to new centres and
        hence we would expect this centralisation to continue. This would be likely to involve the
        domicile and the locations in which the funds were sold being different to the location of
        asset management.

        Third, despite the fact that asset management is viewed as the core competence required
        in the investment fund management industry, there is a trend to increased outsourcing.
        This has arisen because customers, particularly high net worth and private banking
        customers, have become more demanding and some fund managers have developed a
        specific reputation in some areas of asset management.

        Outsourcing typically occurs for specialist funds e.g. offering a fund with a different
        investment style compared to the in-house approach or because the fund manager does
        not believe that they have the necessary competence in a particular area. Typically these
        funds would be those with a geographical investment focus outside of western Europe
        such as Asian funds, Eastern European funds and American funds. Outsourcing these
        funds seems a logical approach while there is relatively small demand for these funds
        from European consumers.

        It is clear that outsourcing these funds is a global market in which national boundaries
        play no role in the choice of fund manager. The market is thought to be highly
        competitive as global fund management companies compete for the mandates. In turn
        these asset managers might be exploiting economies of scale in stock selection by
        running their own fund with a particular investment strategy and then running funds for
        other companies as mirror funds.

        Although outsourcing may enable fund managers to access the skills of alternative
        providers, as with any kind of outsourcing, this is not without cost. In particular, there are
        transactions costs associated to identifying the provider of outsourced activities and
        reaching agreements on servicing. As such the cost of delegation is not zero, although
        fund managers did not highlight this as of particular concern during interviews.

        In the short to medium term we would expect the use of outsourcing of asset
        management to continue as firms seek to offer wider ranges of funds to their client base
        and as that client base becomes more demanding in terms of the investment
        opportunities that are sought. However, once there is sufficient demand we might expect
        some of the outsourcing to cease with fund managers bringing the asset management
        activities back in-house once they have sufficient scale and experience to have built up
        the necessary competence to manage the assets themselves.

        Currently asset management costs around 15 basis points for the average EU equity
        fund. Although much of the gains from efficiency improvements in asset management
        appear to have already been exploited, the continuation of some of these trends would be




                                                                                             Page 115
Competition, integration and costs

September 2006                                                                    CRA International




        expected to continue to lead to cost savings. However, as discussed in section 6.1.2
        above, the more significant cost savings seem likely to arise from increased scale.


        6.4.3. Distribution

        As noted in the previous chapter, proprietary bank distribution remains the most important
        distribution channel in the vast majority of member states included in this study.
        However, there are three main trends that have been noted in nearly all member states.

        Omnibus or nominee accounts

        With the use of omnibus or nominee accounts, the owner of the units according to the
        records kept by the fund manager and the transfer agent is the distributor. The distributor
        then maintains records itself regarding the actual underlying customers. There are
        advantages from a trend to the use of omnibus accounts. In particular, it allows for a
        reduction in costs for the fund manager: Since there is only one client (the distributor)
        record keeping costs are lower and it is possible to aggregate trades. In part this reduces
        the overall costs (e.g. having one buy transaction and one sell transaction rather than
        many of each type) and in part this simply transfers the costs from the fund manager to
        the distributor (keeping client records).

        However, in some countries, it was argued that using nominee accounts brought
        additional systems costs in order to deal with different distributors. This was particularly
        raised by fund managers that were part of banking groups where the bank branches
        represented their primary distribution channel.

        Fund of funds

        Fund of funds are vehicles where the investment fund invests money into other
        investment funds. These have become a more commonly used vehicle in many countries
        over recent years and now represent 9.2% of assets distributed in France and 8.7% of
        assets distributed in Spain.

        For many companies fund of funds are seen as a market entry strategy when they do not
        have alternative distribution arrangements in a particular country. By reaching an
        agreement with one fund of fund provider (typically a bank), the underlying fund is
        distributed to a large number of end customers of the bank.

        This approach may be especially valuable for foreign companies facing consumers with a
        “home bias” i.e. where consumers prefer a domestic brand on the product that they are
        buying. For example, banks may offer a fund of funds product that has their brand name
        on it, but invests in other products outside the group.

        Open architecture

        Open architecture refers to the situation in which non-proprietary funds can be accessed
        through a particular distribution channel. For example, open architecture is seen when
        Schroders’ funds are available in Deutsche Bank’s branch network in Germany. The
        degree to which this is being used varies across countries with locations such as


                                                                                           Page 116
Competition, integration and costs

September 2006                                                                                    CRA International




        Germany, Sweden and the UK areas where it is relatively common compared to locations
        such as Italy and Spain.

        From the distributor’s perspective, third party funds often fill a gap in their own range.
        They may face consumer demand (starting in their private banking operations) for
        additional fund choice or access to the best funds. If this causes banks to search for the
        best funds for their clients because the distributor is acting as an informed buyer on
        behalf of their customers, this is likely to impose competitive pressure on the fund
        management companies.

        Fund supermarkets and platforms

        An extension of open architecture is the use of fund supermarkets and platforms. These
        allow consumers to access a wide range of funds from across the market. In some
        countries, the downturn of stock markets in the late 1990s slowed the growth of fund
        supermarkets, but the great majority of fund managers believed that recent strong
        performance would cause fund supermarkets to attract many new customers over the
        next few years.70

        In Sweden the use of fund supermarkets has been given a particular boost because of
        the reforms that have been made in the pensions arena, where compulsory contributions
        to the Swedish pensions system (PPM) are invested by consumers through what is
        effectively a fund supermarket approach. This is likely to lead consumers to be confident
        to follow a similar approach in their direct investments into investment funds.

        Fund supermarkets usually have search tools available for consumers to access
        information regarding the funds, with comparisons on cost and performance among the
        most common. These tools are likely to aid competition since consumers are more likely
        to compare the various different offerings that they can access and hence make choices
        according to their preferences rather than simply invest in the one option that they are
        offered from a bank. In as far as consumers compare performance indicators, some fund
        managers believed this would lead to increased competition in the whole value chain
        since the overall costs determine the net returns that consumers receive.

        For fund managers, the widening access to distribution that open architecture and fund
        supermarkets allows is seen as very beneficial in the short to medium run. In particular,
        where countries are dominated by banking distribution, fund managers have stressed the
        importance of open architecture or fund supermarkets in order for them to have any
        opportunity to distribute their products in a particular member state.

        So far, many small fund supermarkets have been willing to accept low payments for
        distribution compared to market norms in order to attract large numbers of fund managers
        to their distribution channel, and because their bargaining power is small as their volumes




70      A small number of fund managers dissented from this view believing that fund supermarkets were no longer
        fashionable and that their day had already gone with other fashions likely to take their place. The weight of
        opinion, however, was very much in favour of the rise of fund supermarkets over the next few years.



                                                                                                              Page 117
Competition, integration and costs

September 2006                                                                    CRA International




        are small. Since fund supermarkets, at least in some countries, operate on a direct
        purchase arrangement using either the internet or telephone purchasing, their costs of
        distribution are likely to be smaller than other channels which use face to face salesmen
        or advisers.

        However, over the longer term, it is unlikely that this dynamic will remain. In particular,
        since there are likely to be network effects (because both consumers and fund managers
        want to be on the fund supermarket that offers the widest choice), it is likely that most
        markets will converge on a small number of large fund supermarkets. Indeed, in the US
        although there are thought to be around 80 fund supermarkets, Fidelity Funds Network
        and Schwab are believed to have approximately 80% of the market.71 Similarly, in the
        UK, Co-funds and Fidelity Funds Network are the two major fund supermarket offerings,
        although many more also exist.

        Over time, fund supermarkets will gain greater bargaining power and extract a greater
        proportion of the management fees for the distribution element than currently. Whether
        this leads to a greater or smaller proportion of the management fee compared to today is
        difficult to assess. This issue is also likely to be confused by the fact that some fund
        supermarkets are undertaking record keeping and servicing on behalf of the fund
        manager and therefore undertaking more parts of the value chain. In addition, as fund
        supermarkets grow they are able to aggregate transactions with the fund management
        company. In both of these activities, economies of scale may therefore be realised.

        The degree to which a trend to fund supermarkets results in changing the bargaining
        power of distributors over producers rather than lower cost funds to consumers depends
        on the degree to which consumers are actively choosing fund supermarkets because they
        present a low cost competitor to existing channels. Given the relatively low numbers of
        consumers who actually search on the basis of price (as discussed below) its impact on
        the mass market retail place seems likely to occur slowly over time.

        The cost of distribution for the average equity fund is around 80 basis points. The trends
        in distribution that have been identified (omnibus accounts, fund of funds, open
        architecture and fund supermarkets) are all likely to work in the direction of imposing cost
        efficiencies and greater competition. However, the different distribution models are
        currently reasonably small and it seems likely that it will take some time for these trends
        to actually result in lower distribution costs. The success of these channels and the
        resulting pressure for costs to be lowered is also related to consumer behaviour and their
        price sensitivity, which is discussed below.


        6.5.     PASSING ON COST SAVINGS TO CONSUMERS

        Chapter 5 found that there are potential cost savings to be made in investment funds.
        Chapter 6 so far has examined whether competition, integration and existing market
        trends would cause these to be realised. This section now considers whether, if such
        cost savings are made, they would be passed on to consumers.



71      Interview with UK fund manager, March 2006.



                                                                                           Page 118
Competition, integration and costs

September 2006                                                                                         CRA International




        In all markets, firms have the incentive to reduce costs wherever possible since then they
        are able to increase their profits (irrespective of whether this involves reducing prices to
        customers). If companies did not do this, we would expect the management to be
        replaced and new managers put in their place to target cost savings and increase
        profitability. Even where these cost savings are not passed onto consumers, the
        production of funds would then be productively efficient.72 However, if cost savings are
        not passed through then prices would be too high and allocative efficiency would not
        arise.

        Other things being equal, those providers who make cost savings are able to offer low
        prices, and those providers with low prices gain additional market share. Thus
        competition on prices focuses business on the providers who are the most efficient and
        can offer products at the lowest price in order to be able to attract additional custom while
        making a profit. In the context of investment funds, it is therefore important to consider
        whether consumers are price sensitive and hence whether we would expect any cost
        savings that arise to be passed on to consumers (either through price reductions or
        improvements in quality).

        There are four sources of evidence that we consider:

        •    Interview evidence from fund managers and regulators;

        •    Evidence on price sensitivity in financial services and investment funds in particular;

        •    CRA’s own quantitative research on prices and TERs; and

        •    The impact of existing trends in distribution.


        6.5.1. Interview evidence

        There was considerable scepticism expressed regarding whether consumers were price
        sensitive with respect to the cost of fund management. This was the prevailing view
        amongst nearly all fund managers and regulators interviewed across the different
        member states.

        Many interviewees argued that this lack of price sensitivity among consumers was firstly
        because consumers do not necessarily understand the price of investment funds either
        because they are unaware of management charges (because they may not be
        transparent) or because they do not understand how they impact the value of
        investments.

        In addition, even if consumers do look at a measure of prices, they may not consider all of
        the relevant costs. In particular, there was believed to be a distinction between
        management charges (which some customers might examine) and the total expenses




72      We note that there is also the potential for X-inefficiency although we do not consider it further in this section but
        focus on the prospect for consumer pass through when efficiency savings are made.



                                                                                                                   Page 119
Competition, integration and costs

September 2006                                                                                   CRA International




        incurred by the fund (which was believed to be much less commonly considered even
        though it would be disclosed in the simplified prospectus).

        One of the implications of this may be that those costs that typically feed into the
        management charge (such as asset management and distribution) may receive more
        consumer focus than those costs that do not (such as custody). There were exceptions in
        different member states, for example, in Sweden custody costs are separately disclosed
        to consumers which might be expected to lead to more competitive forces being imposed
        on custody costs than would otherwise be the case.73

        More generally there might be a distinction between those costs charged directly to the
        fund and those charged through the management fee. If custody costs are charged
        directly to the fund, it could be argued that cost savings in this area must be passed
        directly to consumers. However, we found considerable scepticism this would be the
        case, given that these services were often being purchased from within the same financial
        group, interviewees believed that cost savings would if possible be retained by the
        provider rather than passed on.

        However, probably of more significance from the interview evidence was the view that
        consumers focus on performance rather than price. Again there was overwhelming
        agreement on this both from fund managers and regulators. Many fund managers argued
        that this was a reasonable position for consumers to hold since it is the net performance
        of the fund that matters to consumers as this is the return that they receive.

        In particular, fund managers argued that differences in performance were more likely to
        make a significant difference to the net performance compared to differences in prices
        and hence it was reasonable for consumers not to focus on price differences in
        isolation.74 Interviewees argued that where there were very good returns, consumers
        were unlikely to notice or mind whether they received a return of 16.3% or 16.4% and
        hence there was limited pressure on companies to make the extra 10 bps of savings
        because it would not be likely to attract significantly more custom. However, some
        interviewees believed that when returns were poor consumers would focus more on cost
        savings because they represent a larger proportion of the return. Therefore we might
        expect price sensitivity to be somewhat cyclical.

        In addition to performance other measures of the “quality” of investment funds was also
        believed to be important to consumers, this may include aspects such as the investment
        style of the fund management provider, the reputation of the provider and aspects to do
        with distribution and service standards.



73      Another potential result of this is to incentivise fund managers to recover costs through categories directly
        charged to the fund rather than through the annual management charge. This is one of the arguments used to
        criticise the use of soft commissions to pay for investment research.

74      The persistency of past performance is a controversial area and outside of the scope of this report. Regardless
        of the evidence on this, we note that prices can be observed ex ante whereas returns are always only revealed
        ex post. Interested readers could examine, Performance persistence in UK equity funds – A literature review,
        CRA International, January 2002 available at www.crai.co.uk for a discussion of the academic evidence on this.



                                                                                                             Page 120
Competition, integration and costs

September 2006                                                                                      CRA International




        It is also important to note that there is general agreement that it is not necessarily in the
        best interests of consumers to search out a product purely on the basis of cost. Other
        factors such as the investment objectives of the funds and the expected performance are
        also of importance. As with most products there may be a trade-off between price and
        quality and different consumers may have different preferences about where on this
        trade-off they would like to be.

        Finally, it is useful to look at what has happened when cost efficiencies have been
        exploited. Evidence from interviews found many examples where consolidation of funds
        has taken place (and hence cost savings realised) without changes in the charges to
        consumers. While we note that this could be because funds were not profitable before
        they were merged, this evidence is consistent with the views from interviews that cost
        savings would not be passed on to consumers.

        Overall, therefore it was clear that interviewees did not believe that consumers would
        bring strong competitive pressures to bear on the fund management industry at least in
        the short run and hence even if potential cost savings were realised, they did not believe
        that the majority of these savings would be passed to consumers. Even where
        consumers use intermediaries there were concerns regarding whether cost savings would
        be passed on or whether this would simply result in greater rewards for distributors. This
        is further discussed below.


        6.5.2. Evidence on price sensitivity

        There is considerable academic and consumer research evidence suggesting that
        consumers are not particularly price sensitive in financial services products. This is
        thought to be due to a range of factors especially those linked to problems of information
        asymmetry such as complexity of characteristics, bounded rationality, difficulty in
        assessing quality and switching costs.75 In the UK, consumer research has found that
        unit trusts are perceived to be among the more complicated of financial services products.
        Products characterised by difficulties of information asymmetry are typically also those in
        which consumers are not particularly price sensitive since they are unable to clearly
        assess the value of the products.

        Academic evidence explicitly examining price sensitivity in investment funds is relatively
        limited and US based although some interesting results have nonetheless been found.
        For example, Sirri and Tufano assess the degree to which net flows into investment funds
        are related to expenses (among other factors).76 They find that:




75      See section 3 of An assessment of the extent of an identified need for simplified, standard products, A report for
        the European Commission, CRA International, December 2004 for a more detailed further discussion of these
        issues.

76      See Sirri, E. and P. Tufano (1998) “Costly Search and Mutual Fund Flows,” Journal of Finance, 53 (5), pp. 1589-
        1622.



                                                                                                               Page 121
Competition, integration and costs

September 2006                                                                                    CRA International




        •    Consumers prefer funds with lower fees. In particular, fee differences of 100 basis
             points between funds are associated with 2.9% differences in fund flows; and

        •    Flows are inversely related to fee changes, although investors’ response is not
             symmetric. For example, a 20 basis point decrease in fees would lead to an
             increase in fund size by 4.2%. However, fee increases do not have a significant
             impact on flows.

        The explanation provided for the asymmetrical result is linked to the effect of marketing.
        Sirri and Tufano argue that increased fees lead to greater marketing efforts to sell funds –
        which in turn reduces search costs for investors and make it easier for them to identify
        and buy a fund (even if such fund is less attractive). These two effects – reduction in
        search costs and increase in fees – then cancel each other out.

        More recently, Barber et al. also find evidence of a negative impact of expenses on
        mutual fund flows.77 In particular, they find that investors tend to invest less in funds
        which have high front-end load charges, since these are

                   “transparent and thus salient in-your-face expenses”.

        Moreover, by comparing mutual fund purchases of first-time buyers with those made by
        experienced buyers, Barber et al. also find direct evidence of learning. For example, the
        empirical results show that experienced fund buyers pay, on average, about half the front-
        end load fees of first-time purchasers.

        In contrast, the relationship between fund flows and (on-going) operating expenses is not
        significant although excluding 12B-1 fees (paid in the US for distribution and marketing),
        other operating expenses have a negative impact on fund flows. This supports the
        previous view that marketing expenditures have an impact on investors’ choice of mutual
        funds. Barber et al. conclude that

                   “investors have learned by experience to avoid mutual fund expenses. However, they
                   have learned more quickly about front-end load fees, which are large salient one-time
                   fees, than operating expenses, which are smaller ongoing fees that are easily masked by
                   the volatility of equity returns.”

        The academic evidence therefore does in fact provide some support for price having
        some impact on fund flows, although the economic impact of the price sensitivity is
        reasonably small and the evidence comes predominantly from the US. Given the greater
        disclosure that is available in the US, this could be used to argue that efforts to improve
        transparency will also result in price sensitivity in Europe.




77      See Barber, Brad M., Odean, Terrance and Zheng, Lu (2003), "Out of Sight, Out of Mind: The Effects of
        Expenses     on   Mutual    Fund   Flows",      forthcoming   Journal   of   Business.   Available   at   SSRN:
        http://ssrn.com/abstract=496315.



                                                                                                              Page 122
Competition, integration and costs

September 2006                                                                                           CRA International



        6.5.3. Quantitative evidence on prices and TERs

        To look at the situation in Europe we have also examined the trend in prices over time.
        As described above, inferring price sensitivity from comparing of annual management
        charges or TERs is problematic. However, in this case we have focused on a particular
        part of the market, index trackers, where the problems discussed in section 6.3.1 are
        mitigated. By focusing on this market segment, issues to do with quality and different
        objectives of funds are minimised.

        Figure 34 looks at the management fee of index tracker funds in a small number of
        member states to identify whether prices to consumers have changed over time. If there
        is no incentive to pass on cost reductions to consumers then we would not expect to see
        any decrease in prices over time since providers gain no additional market share from
        doing so.

        Figure 34: Index tracking weighted average management fee


                                                                     France     Germany   Spain    UK
                                            1.60%

                                            1.40%
          Weighted average management fee




                                            1.20%

                                            1.00%

                                            0.80%

                                            0.60%

                                            0.40%

                                            0.20%

                                            0.00%
                                                    1997   1998   1999   2000     2001    2002    2003   2004     2005


        Source: CRA International based on Fitzrovia data


        It is clear from Figure 34 above that there is in fact evidence of index tracking prices
        falling over time. This is especially the case in France and Spain, but is also identified in
        the UK. A similar result is found if the TER rather than the management fee is examined,
        although this evidence is slightly weaker.

        This evidence supports a view that potential cost savings would in fact be passed onto
        consumers since it suggests that cost savings are already being passed on in index
        trackers where less differentiation in products compared to other types of investment
        funds makes the comparison possible. It also gives some support to the view expressed
        by Barber et al that consumers are more sensitive to charges which are more transparent
        (in this case the management fee compared to the TER).



                                                                                                                  Page 123
Competition, integration and costs

September 2006                                                                              CRA International




        In addition to impacting average prices, passing on cost savings and competitive
        pressures would also be expected to reduce the spread of prices. Again using index
        tracking funds we examine whether there is any evidence of changes in the spread of
        prices as measured by the management fee or the TER.

        Figure 35: Index tracking standard deviation of management fee


                                                        France     Germany   Spain    UK
                               0.80%

                               0.70%

                               0.60%
          Standard deviation




                               0.50%

                               0.40%

                               0.30%

                               0.20%

                               0.10%

                               0.00%
                                       1997   1998   1999   2000     2001    2002    2003   2004     2005


        Source: CRA International based on Fitzrovia data


        Figure 35 above shows very strong evidence of a reduction in the variation in prices in the
        UK, France and Spain. In Germany, the sample size before 2000 is small and hence it is
        appropriate to examine the period after 2000 when there is a larger sample size and
        during which a reduction in the variation of prices is also observed.

        Overall, therefore, within index trackers there is evidence that would support the view that
        cost savings would, at least in part, be passed onto consumers because of the evidence
        of both falling prices and convergence of prices currently. One caution that must be
        applied is that index trackers are likely to be the area in which there is greatest price
        competition since there is less differentiation in quality as all funds are passively
        managed. Hence evidence of price competition here does not necessarily indicate that
        price competition should be expected to the same degree in other fund types where
        quality indicators may be more important to consumers.


        6.5.4. Cost pass through and distribution models

        Finally, it is important to look at distribution for two reasons. Firstly, the model of
        distribution could provide evidence regarding financial sophistication. There is no
        comparable measure of the degree of consumer sophistication to use across member




                                                                                                     Page 124
Competition, integration and costs

September 2006                                                                                      CRA International




        states.78 It is, however, possible to other proxies such as the use of direct purchases as
        an indicator of consumer sophistication since they, along with fund of funds, may reflect a
        measure of consumer choice. This would support France and the UK as having relatively
        sophisticated consumers due to the high proportion of consumers using fund of funds and
        buying directly respectively. Examining production costs, we find that the UK and France
        also have relatively low costs at current fund sizes. In addition, we find that Sweden,
        where the use of funds has been encouraged by the development of the PPM in which
        consumers can make choices over funds, has low costs (even though concentration is
        high). This is therefore supportive of the relationship between consumer sophistication
        and the cost of funds.

        Figure 36: The proportion of new business sold via direct and fund of funds channels


                                                         Direct   Funds of funds
          18%

          16%

          14%

          12%

          10%

           8%

           6%

           4%

           2%

           0%
                       France             Germany                 Italy               Spain               UK


        Source: FERI data. Note that data is not available for other countries covered in our report.


        Looking to the future the movement to purchasing through the internet and the use of
        fund supermarkets are often associated with search tools that allow consumers to search
        for funds on the basis of the charges that are imposed, with some of these tools providing
        definitions of the charges. This would suggest that the likelihood of cost savings being
        passed onto consumers will improve in the future.

        Secondly, even if consumers are not price sensitive, it is possible that they use
        intermediaries to shop around with the result that savings are passed on to them. In this
        case, this role could be undertaken by a financial adviser or even a retail bank choosing
        which funds to include in their range. It was argued that this could result in a greater



78      The types of product that consumers buy in different markets could be used as a measure of financial
        sophistication. However, this is likely to be very misleading. For example, in France the market is dominated by
        money market investment funds compared to markets such as the UK and Sweden where equity products are
        more common. However, this appears to reflect historical differences between the markets, for example that
        interest was capped on deposit accounts in France rather than purely consumer behaviour.



                                                                                                               Page 125
Competition, integration and costs

September 2006                                                                     CRA International




        pressure for the fund managers to pass on cost savings or offer better service standards.
        However, from the consumer’s perspective they still face the issue of how to choose the
        best intermediary and determine whether they are acting in their best interest. Indeed,
        there is little evidence that consumers are price sensitive when choosing their adviser.
        Indeed, in markets where this model is being used we found a concern regarding the role
        of commission, the information that was revealed to consumers and whether this resulted
        in cost savings being passed onto the distributor rather than the final consumer.

        The amount of information being given to consumers does appear to be increasing. In
        some countries there are requirements on the disclosure of cost information at the point
        of sale in a variety of different formats in order to allow consumers the opportunity to
        compare funds. Indeed, the impact of MiFID has been for many countries to review the
        information given the customers. However, the impact of MiFID here is currently
        unknown with many market participants indicating that this will simply add costs to
        investment funds without bringing any benefits (and may even in a number of cases
        prevent domestic measures to improve disclosure). There is considerable scepticism that
        consumers actually read much of the information that they are given which suggests that
        enforcing more transparency would actually have relatively little impact. Furthermore,
        some fund managers thought that transparency could be bad for distribution negotiations
        because it becomes harder to change commissions over time when these charges are
        disclosed in prospectuses. As an example, it was noted that in Japan, the fund
        prospectus gives the annual management charge and the proportion of this going to each
        named distributor and this has led to their being no negotiations as all distributors receive
        the same amount of commission.

        There is general agreement across the ten member states that consumers do not exert as
        much discipline of providers or distributors as would be liked by regulators. An on-going
        trend in many member states is the provision of more information regarding the fund’s
        characteristics, the cost of the fund and the remuneration of advisers. However, there is
        little evidence that disclosure alone brings substantial benefits. Therefore we do not
        believe that there will be a significant increase in the sophistication of consumers in the
        short-term that would lead to a sharp reduction in cost differences.


        6.5.5. Conclusion regarding cost pass through

        There is also some reason to have cause for optimism regarding consumer behaviour.
        We found that:

        •   The trend to offering choice through previously closed distribution channels appears
            to be continuing. Although this started with private clients and discretionary fund
            managers, this is moving into branch networks and increasingly being made available
            to retail consumers.

        •   There is direct evidence from our interview programme that although historically
            distributors have been able to negotiate their remuneration without reference to end-
            consumer fees, there are recent examples of providers seeking to preserve their
            reputation in the market by pushing back on some commission structures they see as
            “being too high”.



                                                                                            Page 126
Competition, integration and costs

September 2006                                                                  CRA International




        However, based on interview evidence this is starting from a low base. The regulators in
        all 10 member states believe that consumer’s capability to choose between fund offerings
        could be significantly improved.

        •   Even where product disclosure is relatively high, for example, in the UK and Sweden
            there is an increasing concern that consumers face too much choice and are unable
            to make informed choices;

        •   There is concern in all countries that consumers place to much value on past
            performance rather than on the cost of the underlying fund; and

        •   There is a concern in many countries about the influence of the sales process and
            degree to which consumers understand whether they are paying for advice. In many
            countries this has led to increased disclosure of the remuneration structure.

        Empirical evidence on cost savings being passed on to consumers is weak. Interviewees
        were sceptical regarding cost savings being passed on and point to asymmetric
        information, complex products and poor financial literacy as reasons for this as well as a
        lack of price reduction following mergers of funds. Nonetheless, there is some evidence
        that cheaper funds attract additional market share along with evidence of price falls and
        price convergence in index trackers which would support consumers gaining from cost
        reductions.

        Overall, this suggests that cost savings are likely to be passed onto consumers only very
        slowly, although as consumers become increasingly familiar with investment funds over
        time and as the financial literacy of consumers increases over the longer term we would
        expect at least some of the cost savings to be passed on to them.




                                                                                          Page 127
Annex A: Technical Details on the model

September 2006                                                                     CRA International




        ANNEX A: TECHNICAL DETAILS ON THE MODEL

        In this annex we describe the model used to estimate the potential savings arising from
        the integration scenarios, more detailed results for bond funds and the result from
        additional scenarios.


        STRUCTURE OF THE MODEL

        The purpose of the model is to estimate potential cost savings resulting from a series of
        scenarios designed to test what would happen in a fully integrated investment fund
        market. The scenarios under investigation vary depending on the choice of:

        •   The comparator under investigation: we wished to test the savings resulting from the
            production costs becoming equal to those in other member state, the cost of a cross-
            border fund, and to a similar US fund.

        •   The assumption regarding the impact on scale: Each of these scenarios had to be
            calculated assuming the scale of funds in each of the countries stayed as of now, and
            if they moved to the European average.

        •   Whether they apply to production or distribution activities: as distribution is likely to
            remain a local activity we wished to distinguish between this and production costs.

        To estimate the potential cost savings from these scenarios it was therefore important to
        distinguish between efficiency improvements and the impact of scale (as illustrated in
        Figure 4) and between efficiency improvements resulting from the production of the fund
        and the cost of distribution. Therefore we constructed a model that estimated how
        different costs components vary with the size of the fund.

        To look at the cost savings resulting from a US based cost structure we have also
        included the cost of an average US fund. This used Morningstar 2006 data for retail
        funds. This has been calculated in a different manner to that described below and does
        not allow us to vary the size of the fund.


        Data

        The model uses cost data described in chapters 3 and 4 of the main report. It is based
        on:

        •   The 28 quantitative surveys (as described in section 2.4) collected for this assignment
            during the period March to June 2006. These provided us with a detailed description
            of the cost of the various different stages of the value chain for each of the ten
            comparator countries. The information from different providers was used to calculate
            the cost of a “generic” fund of average size.

        •   Data provided by Fitrozia. In addition to using Fitzrovia data during the interviews with
            market participants we have used Fitrovia data on audit, custody and fund
            administration.


                                                                                            Page 128
Annex A: Technical Details on the model

September 2006                                                                    CRA International




        •   Data provided directly by regulatory authorities regarding the costs of fund
            authorisation.

        The data provided has allowed us to investigate the cost differences for different part of
        the value chain, different fund types and different member states. However, the richness
        of the data varied significantly and we were only able to develop the scenario based
        model for equity and bond funds.

        As the aim of the model is to calculate the cost savings in terms of both euros saved in
        the cost of providing retail investment funds and in terms of a reduction in bps charged to
        the fund we need to scale the result. To calculate the aggregate savings we used
        average fund size data provided by EFAMA.


        Calculating average costs

        To calculate the average cost of a fund in the ten reference countries we used data from
        the responses to the quantitative questionnaire. This was used firstly to assess the
        nature of cost, i.e. whether it varied by type of fund, by member state or was driven by the
        size of the fund.

        The assumptions used for each of the cost categories are set out in Table 15 below. As
        with any modelling approach a number of pragmatic assumptions were required when the
        data was too volatile or appeared to have been submitted on different basis by different
        participants. A good example of this is the costs of overheads and systems costs.
        Although many respondents suggested these were largely fixed costs, the estimate of
        these costs varied significantly between market participants. We have chosen to apply
        these as a fixed percentage of assets. This was the method thought most appropriate for
        sharing fixed costs within the company.




                                                                                           Page 129
Annex A: Technical Details on the model

September 2006                                                                                   CRA International




        Table 15: Description of the cost components in cost model

             Category                 Function                          Method of quantification

             Regulatory              Authorisation    Estimated using the cost provided by member state regulators
          compliance of the                                               as set out in Table 6.
                fund

                                Other elements of           Estimated using the responses to the quantitative
                                    regulatory             questionnaire as a fixed cost as in Figure 6, page 36.
                                   compliance

                                        Audit         Fixed cost per fund set out in Figure 12 p.53 invariant to type
           Fund and client
                                                         of fund. This varies from €4,500 in Spain to €13,000 in
            administration
                                                                                Luxembourg

                                Fund accounting        Varies depending on assets under management and by type
                                                                   of fund, does not vary by country

                                                                  For equity funds the relationship is
                                                                                                          (-0.2178)
                                                          Fund accounting cost in bps = 43*(Assets size)^

                                       Custody          Relationship between cost and scale determined through
                                                      econometric methods. The specification of the equation is set
                                                                             out in below.

                              Client administration     Variable cost by type of fund and domicile as in Figure 13,
                                                                                  page 56.

                                                        Relationship between cost and scale determined through
                                                         econometric methods (as illustrated in Figure 14 p.61)
               Asset
            management
                                                                  For equity funds the relationship is
                                                                                                        (-0.3325)
                                                         Asset management cost in bps = 71*(Asset size)^

                                                         Country and fund specific costs applied as a variable as set
                                                         out in Figure 23, p. 77. Ireland and Luxembourg are treated
             Distribution                             differently in the model as these represent exporting countries.
                                                      It is therefore appropriate to apply average distribution costs to
                                                                 these funds for the purposes of comparison.

                                  Management                 Calculated as average of all industry responses
             Overheads         company overheads                                 6bps

                                       Systems               Calculated as average of all industry responses
                                                                                 1 bps

        Source: CRA International.


        The relationship between custody, size and domicile

        In order to assess the extent of economies of scale and cost differences in the custody of
        European-focused equity funds in different member states, we have carried out a simple
        econometric analysis where custody fees paid by each fund (expressed in basis points –
        bps) are regressed against fund size and several country dummies. Table 16 below
        presents our results, where in column (i) and (ii) funds domiciled in Luxembourg and
        Ireland are included and excluded from the dataset, respectively.




                                                                                                             Page 130
Annex A: Technical Details on the model

September 2006                                                                                     CRA International




        Table 16: Linear regression results for custody fees of equity funds

             Independent variables                 (i) Including Ireland and              (ii) Excluding Ireland and
                                                          Luxembourg                             Luxembourg

                                                              -0.005                                -0.003
          Net Assets (m)
                                                             (5.22)**                               (2.43)*
                                                              1.147                                    -
          Dummy for Luxembourg
                                                              (1.39)
                                                              0.729                                    -
          Dummy for Ireland
                                                              (0.73)
                                                              2.988                                  2.905
          Dummy for Germany
                                                             (3.10)**                               (3.41)**
                                                              4.274                                  4.098
          Dummy for Italy
                                                             (4.26)**                               (4.61)**
                                                              4.102                                  4.225
          Dummy for Spain
                                                             (4.49)**                               (5.22)**
                                                              -1.920                                 -2.084
          Dummy for Sweden
                                                              (0.83)                                 (1.02)
                                                              -2.005                                 -2.182
          Dummy for the United Kingdom
                                                              (2.18)*                               (2.67)*
                                                              -5.721                                 -5.901
          Dummy for France
                                                             (4.56)**                               (5.31)**
                                                              9.072                                  8.827
          Constant
                                                             (11.72)**                             (12.79)**
          No. of observations                                  691                                    384

          Adjusted R-squared                                   0.22                                   0.37

          Dublin and Luxembourg                                YES                                    NO
          included?

        Source: CRA analysis on Lipper/Fitzrovia data. Notes: only funds with custody fees between 1 and 25 bps and
        net assets between 1 million and 1,000 million included in the sample. T-statistics indicated in parentheses; *
        (**) denotes that the coefficient is significant at 5% (1%). The dummy variables have a value of 1 if the fund is
        domiciled in the specified location, and 0 otherwise; Belgium is the omitted country. Luxembourg and Ireland
        are omitted in the second model because this improves the fit of the modelling.


        In both specifications, our estimates confirm the existence of a negative relationship
        between the level of custody fees and fund size, i.e. there are economies of scale in
        custody. However, although the relationship is statistically significant, its magnitude in
        economic terms is relatively low: on average, if the size of a fund increases by a hundred
        million, custody fees only fall by half a basis point (according to our first model).

        More importantly, there are sizeable differences in custody fees across countries, even
        after controlling for fund size. For example, compared to Belgium, custody fees are lower
        in France, the UK, and Sweden (although the coefficient for Sweden is not significant)




                                                                                                               Page 131
Annex A: Technical Details on the model

September 2006                                                                                            CRA International




        and higher in Germany, Spain and Italy. This is useful to identify differences between
        countries at the overall level of the costs of custody.79

        The results of the model for equities

        The results of the model for equities have been provided in the main body of the report.
        Here we provide a table of the costs in basis points to go alongside the graphical
        representation of the costs provided earlier.

        Figure 37: Comparison of total costs by member state based on current average fund sizes
        for equity funds

                                  Belgium   France   Germany   Ireland   Italy   Luxembourg   Poland   Spain   Sweden      UK
        Regulator compliance of      4.0      0.9       3.5       1.4      1.9        0.5       4.7      2.5      1.2       1.1
        Audit                        1.7      0.7       0.5       0.9      0.6        1.0       1.4      0.8      0.5       0.3
        Fund Accounting             17.7     15.5      13.0      15.7     13.8       14.2       18.1    16.8     13.4      12.3
        Custody                      8.8      2.8      10.9       9.3     12.4        9.4       9.2     12.8      6.1       5.5
        Transfer Agency              0.6      0.3       0.6       5.8      0.6        5.8       24.0     1.0      0.8      10.9
        Asset management            19.5     16.0      12.3      16.3     13.5       14.0       20.2    18.2     12.8      11.3
        Overheads                    7.0      7.0       7.0       7.0      7.0        7.0       7.0      7.0      7.0       7.0
        Distribution                54.0     37.0      83.0      94.0    147.0       94.0      150.0   154.0     99.0      61.0
        Total cost                 113.3     80.3     130.8     150.3    196.8      145.8      234.6   213.1    140.6     109.5


        Source: CRA analysis


        The results of the model for bonds

        The results of this data analysis allows us to compare the cost of funds across member
        states. Below we present the results for bond funds.




79      However, given the differences in the sizes of funds that are prevalent in different countries, the coefficients on
        the country dummies should not be used as the estimate of the actual difference in custody costs between the
        member states.



                                                                                                                        Page 132
Annex A: Technical Details on the model

September 2006                                                                                                                                                           CRA International




        Figure 38: Comparison of total production costs for bond funds by member state based on
        current average fund sizes


                                                      Regulator compliance of the fund     Audit             Fund Accounting   Custody         Transfer Agency   Asset management    Overheads

                                               90.0

                                               80.0
          Cost of production in basis points




                                               70.0

                                               60.0

                                               50.0

                                               40.0

                                               30.0

                                               20.0

                                               10.0

                                                0.0
                                                                                 Germany




                                                                                                                      Italy




                                                                                                                                                                                         UK
                                                       Belgium




                                                                                                   Ireland




                                                                                                                                  Luxembourg



                                                                                                                                                    Poland



                                                                                                                                                                 Spain
                                                                    France




                                                                                                                                                                            Sweden
        Source: CRA analysis


        Comparing equity to bonds funds we find that the main difference is that at the current
        fund size, bond funds in Italy are found to have the lowest cost of production. However,
        when countries are compared on the basis of the same average fund size (€212 million),
        the same rankings are found between the countries as seen in Figure 38 above. This
        mainly reflects the fact that Italy has by far the largest average bond fund size at €467
        million and hence is currently gaining considerably from economies of scale in
        comparison to other countries.




                                                                                                                                                                                       Page 133
Annex A: Technical Details on the model

September 2006                                                                                                                                                                 CRA International




        Figure 39: Comparison of total production costs for bond funds by member state based on
        current average fund size


                                                                 Regulatory compliance of the fund and overheads              Administration           Asset management        Distribution

                                            160.0

                                            140.0
            Cost of funds in basis points




                                            120.0

                                            100.0

                                             80.0

                                             60.0

                                             40.0

                                             20.0

                                              0.0
                                                                                 Germany




                                                                                                                   Italy




                                                                                                                                                                                                   UK
                                                     Belgium




                                                                                               Ireland




                                                                                                                                   Luxembourg



                                                                                                                                                       Poland



                                                                                                                                                                     Spain
                                                                    France




                                                                                                                                                                                     Sweden
        Source: CRA analysis


        Figure 40: Comparison of total costs by member state based on current average fund sizes
        for bond funds

                                                               Belgium       France        Germany       Ireland           Italy                Luxembourg      Poland       Spain        Sweden        UK
        Regulator compliance                                     2.4           0.8           4.0            1.2             0.8                      0.5          4.1          2.1          2.7          2.0
        Audit                                                    0.7           0.5           0.5            0.5             0.2                      0.7          0.9          0.5          0.8          0.4
        Fund Accounting                                         11.5          10.8           9.6           10.9             8.2                     10.1          12.8        11.8         11.6         10.1
        Custody                                                  6.8           0.9           9.0            7.4             9.0                      7.5          7.4         11.0          4.9          4.3
        Transfer Agency                                          0.6           0.3           0.6            5.8             0.6                      5.8          24.0         1.0          0.8         10.9
        Asset management                                        16.8          14.6           10.9          14.8             7.5                     12.6          21.7        17.9         17.1         12.4
        Overheads                                                7.0           7.0           7.0            7.0             7.0                      7.0          7.0          7.0          7.0          7.0
        Distribution                                            40.0          18.3           15.6          39.0             73.6                    39.0          26.3        90.0         40.0         36.8
        Total cost                                              85.9          53.1           57.2          86.6            106.9                    83.2         104.2       141.2         84.8         83.9


        Source: CRA analysis


        Given the responses to the quantitative survey it was not possible to develop a similar
        analysis for money market, index tracking or capital protected funds.


        Estimation of potential costs savings

        To calculate the cost savings we need to make a number of simplifying assumptions:

        •                                   Our data analysis is based on the results of the quantitative survey. So the results
                                            were as comparable as possible we focused on particular types of fund. To calculate
                                            cost savings we need to assume that the cost savings relating to the category of
                                            funds investigated in the quantitative survey apply to all similar funds, i.e. that cost
                                            savings relating to broad equity funds also applies to those focusing on industry
                                            specialisms;




                                                                                                                                                                                               Page 134
Annex A: Technical Details on the model

September 2006                                                                      CRA International




        •   We assume that the reduction in costs happens instantly and calculate the resulting
            annual cost saving; and

        •   The model on looks only at on-going costs and does not take into account initial costs
            associated with new business.

        In reality of course any change will take time to come into affect, therefore the transition to
        the new regime and the time taken would need to be taken into account in any regulatory
        analysis (cost-benefit analysis of potential regulatory changes for example).

        An important caution regarding the outputs of project is that the model is based on a
        series of relationships that can only accurately be estimated for marginal or small
        changes. Therefore where we have assumed large changes, i.e. changing the average
        fund size to the US average, this is likely to result in estimates with significant error
        margins and these should be used with caution.

        To calculate aggregate cost savings we also need to scale the results. The output of the
        model is a euro value of the potential cost savings based on the member states that have
        been included in our study. This figure is calculated through scaling the basis point
        estimates of potential cost savings according to the market size of the different fund types
        in each of these member states. This gives us the potential cost saving for the 10
        member states examined in this study.

        A further issue with respect to Ireland was the lack of data regarding the fund size of
        different asset categories, in this case we have assumed the distribution of fund is the
        same as Luxembourg.


        ADDITIONAL SCENARIOS USED TO TEST THE IMPORTANCE OF SCALE

        To look at the impact of scale we applied the scenarios discussed in chapter 5 after
        adjusting the fund size. We did this in a number of different ways:

        •   Increasing fund size to the European average;

        •   Increasing fund size to the European maximum;

        •   Increasing fund size to the average of the US.

        The results from these simulations are set out below.

        Comparisons with EU average fund sizes

        In chapter 5 we considered the impact of adjusting all fund to the European average, a
        slight variation on this would be to use the average EU fund size for those countries which
        have a below average fund size and to retain the existing fund size for those countries
        which have an above average fund size. These results are provided in Table 17 below.




                                                                                              Page 135
Annex A: Technical Details on the model

September 2006                                                                                                                                                               CRA International




        Table 17: Cost savings in production based on at least average EU fund size for equity funds

                                                                            Cost of production                                Cost savings in €                         Cost savings in bp

         At least average EU                                                         €6,129 million                                           -                                            -
         fund size

         Average cost                                                                €5,855 million                                 €275 million                                      2 bp

         Cross-border cost                                                           €6,118 million                                      €12 million                                <1 bp

         Lowest cost                                                                 €4,946 million                              €1,183 million                                       9 bp

         US cost                                                                     €4,635 million                              €1,494 million                                     12 bp

        Source: CRA analysis


        Comparisons with EU maximum fund sizes

        The comparison in Table 10 was done on the basis of moving from the current fund size
        in each member state to the average fund size across member states and then observing
        what cost differences could be achieved by reducing the cost of production to different
        scenarios. However, this involved some countries seeing an increase in the cost of
        production compared to today since the assets under management fell in the baseline for
        those countries that currently have above average fund sizes. For this reason in this set
        of scenarios we consider the impact of first moving to the maximum EU fund size (€253
        observed in the UK) and then observing the potential cost savings that could be made
        between different scenarios.

        Figure 41: Comparison of total production costs for equity funds by country based on
        maximum EU fund sizes


                                                      Regulator compliance of the fund         Audit   Fund Accounting        Custody        Transfer Agency       Asset management            Overheads

                                               70.0
          Cost of production in basis points




                                               60.0

                                               50.0

                                               40.0

                                               30.0

                                               20.0

                                               10.0

                                                0.0
                                                                           Germany




                                                                                                       Italy




                                                                                                                                                                        UK


                                                                                                                                                                                Weighted-avg
                                                      Belgium




                                                                                         Ireland




                                                                                                                 Luxembourg


                                                                                                                                Poland


                                                                                                                                                  Spain




                                                                                                                                                                                                    US
                                                                France




                                                                                                                                                          Sweden




                                                                                                                                                                                    fund




        Source: CRA analysis



                                                                                                                                                                                                 Page 136
Annex A: Technical Details on the model

September 2006                                                                     CRA International




        Note that the US comparator uses existing US fund size (€886 million) rather than the EU
        average fund size. It is also important to observe from Figure 41 above that in this
        scenario, the US does not in fact have the lowest average cost, but rather France is now
        lower than the US.

        Table 18: Cost savings in production based on maximum EU fund size for equity funds

                                  Cost of production      Cost savings in €      Cost savings in bp

         Maximum EU fund size        €5,614 million                -                      -

         Average cost                €5,323 million           €291 million               2 bp

         Cross-border cost           €5,538 million           €76 million               <1 bp

         Lowest cost                 €4,325 million          €1,289 million             10 bp

         US cost                     €4,575 million          €1,039 million              8 bp

        Source: CRA analysis


        As observed above, the cost of production in France is below that of the US. Hence in
        Table 18 above, the cost savings of moving to the lower cost (France) is greater than that
        of moving to the US cost. Furthermore, the cost of production in the US scenario is
        different to that in Table 9 and Table 10 because France does not move to the US
        production cost since it is already below it.

        Looking at the impact of the moving to the average European fund versus the cheapest
        European fund, unexpectedly the cost savings from the assuming the cheapest fund are
        significantly higher. However, it is surprising that after adjusting for scale, the impact of
        moving to the cheapest fund is similar to assuming the US cost structure. That is even
        allowing capturing the differences in productive efficiency in Europe results in similar
        saving to moving to the US average fund (that assumes much bigger fund size).

        Comparisons with US fund sizes

        To test whether the savings resulted from the US comparator result from scale effects or
        efficiency, we first assume that all the member states have funds equal to the US
        average. We then observe what cost differences could be achieved by reducing the cost
        of production in different scenarios.




                                                                                              Page 137
Annex A: Technical Details on the model

September 2006                                                                                                                                                               CRA International




        Figure 42: Comparison of total production costs for equity funds by country based on
        average US fund sizes


                                                      Regulator compliance of the fund         Audit   Fund Accounting        Custody        Transfer Agency       Asset management            Overheads

                                               60.0
          Cost of production in basis points




                                               50.0


                                               40.0


                                               30.0


                                               20.0


                                               10.0


                                                0.0
                                                                           Germany




                                                                                                       Italy
                                                      Belgium




                                                                                                                                                                        UK


                                                                                                                                                                                Weighted-avg



                                                                                                                                                                                                    US
                                                                France




                                                                                         Ireland




                                                                                                                 Luxembourg


                                                                                                                                Poland


                                                                                                                                                  Spain


                                                                                                                                                          Sweden




                                                                                                                                                                                    fund
        Source: CRA analysis


        What is striking about Figure 42 above is that with the exception of Poland and the UK, all
        countries have a lower production cost than does the US with the result that the average
        production cost for the EU is also below that for the US.

        Table 19: Cost savings in production based on average US fund size for equity funds

                                                                            Cost of production                                Cost savings in €                         Cost savings in bp

         Average US fund size                                                        €4,183 million                                           -                                            -

         Average cost                                                                €3,942 million                                 €241 million                                      2 bp

         Cross-border cost                                                           €4,119 million                                      €64 million                                <1 bp

         Lowest cost                                                                 €3,156 million                              €1,027 million                                       8 bp

         US cost                                                                     €4,153 million                                      €30 million                                <1 bp

        Source: CRA analysis


        Table 19 shows that once the average fund size has moved to that observed in the US,
        moving to the US cost does not then reduce the cost of production by very much since
        only the UK and Poland are affected by this.

        It is, however, important to note that considerable caution should be placed on
        interpreting this result since it involves forecasting the cost functions for the various
        components of the value chain far outside of the current range in which the modelling was
        undertaken. For this reason the uncertainty surrounding the likely cost of production


                                                                                                                                                                                                 Page 138
September 2006                                                                 CRA International




       would be expected to be quite considerable and hence these estimates may not be
       robust. Taken at face value, however, this suggests that the cost of European fund differ
       from US because of scale rather than efficiency of production.




                                                                                        Page 139
Annex B: Bibliography

September 2006                                                                   CRA International




       ANNEX B: BIBLIOGRAPHY

       A Harmonised, Simplified Approach to UCITS Registration, EFAMA and IMA, April 2005.

       AFG comments to the Green Paper on UCITS, November 2005, available at
       http://www.afg.asso.fr/upload/3/Fichier411.pdf.

       Ahorro Financiero de las Familias: Informe 2004 y Perspectivas 2005, Inverco, February
       2005 available at www.inverco.es.

       Analiza wybranych kategorii opłat w funduszach Inwestycyjnych akcji cz 2, Piotr Kozak,
       avaiable at www.fundi.pl/artykuly/fundusze_akcji_oplaty2

       An assessment of the extent of an identified need for simplified, standard products, A
       report for the European Commission, CRA International, December 2004

       Annual Report      of   the   AMF,   2004   http://www.amf-france.org/documents/general/
       6339_1.pdf.

       Annual Report BEAMA, 2004.

       Annual Report BVI 2004.

       Annual Report CBFA, 2004.

       Annual Report CMNV, 2004.

       Annual Report Consob, 2004.

       Are European Funds Achieving Economies of Scale? Fitzrovia Press Release 22, Many
       2002, London.

       Asset Management Survey, Investment Management Association, July 2006.

       Assogestioni, Fact Book 2004

       Assogestioni, Fact Book 2005

       Banca d’Italia Abridged Report for the Year 2004.

       Benefits of an integrated European Fund Management: Cross border merger of funds a
       quick win? Invesco January 2005.

       BNP Paribas comments to the Green Paper, available under http://forum.europa.eu.int/
       Public/irc/markt/markt_consultations/library?l=/financial_services/asset_management&vm
       =detailed&sb=Title.

       Bulletin 2005, IZFIA, available at http://www.izfa.pl/biuletyn/biuletyn_2005_01.pdf.




                                                                                              Page 140
Annex B: Bibliography

September 2006                                                                 CRA International




       BVI: Grünes Licht für Garantiefonds, 2 December 2005, available under
       http://www.bvi.de/export/sites/internet/de/infocenter/presse/pressemitteilungen/2005/pm0
       21205/pm021205.pdf.

       CESR’s guidelines to simplify the notification procedure of UCITS, 2nd Consultation
       Paper, May 2006.

       Consob Resolution 15268, of 28 December 2005

       Costly Search and Mutual Fund Flows, Journal of Finance, 53 (5), Sirri, E. and P. Tufano,
       1998

       Cross-border marketing of “harmonised” UCTIS in Europe: Current situation, constraints
       and ways forward, Price Waterhouse Coopers November 2001.

       Does Fund Size Erode Mutual Fund Performance? The Role of Liquidity and
       Organization, Chen, Hong, Huang, and Kubik,2004, American Economic Review, 94.

       Economies of Scale and Scope in French Mutual Funds, Jean Dermine and Lars-Hendrik
       Roller, Journal of Financial Intermediation, 1992, 2, pp. 83-93.

       Economies of Scale in Mutual Fund Administration, Latzko, 1999, Journal of Financial
       Research, 22.

       EFAMA Fact Book, 2005.

       Ernst and Young, available from www.ey.com/global/content.nsf/International/Ireland_-
       _Market_Summary

       Ernst and Young Study, Ireland, available from www.ey.com/global/download.nsf/
       International/Investment_Funds_in_Ireland_-_May_2005

       Ernst and Young, Irish Investment.

       European Fund Market Data Digest 2006, FERI Fund Market Information.

       Expense ratios of North American mutual funds, Ruckman, 2003,,Canadian Journal of
       Economics, 36.

       Financial Services Action Plan: Progress and Prospects. Expert Group o.n Insurance and
       Pensions, Final Report May 2004 available at http://europa.eu.int/comm/internal_market/
       finances/index_en.htm.

       Fitzrovia data, www.dfia.ie/public/files/1143808588_copyofindstatsmar06.pdf.

       Fondbolagens Förening, Outlook 2005/2006.

       Fund Management, City Business Series 2003, IFSL, May 2003.

       Gazeta available at http://gospodarka.gazeta.pl/pieniadze/1,49966,2938617.html



                                                                                        Page 141
Annex B: Bibliography

September 2006                                                                   CRA International




       Gazeta available       at   http://gospodarka.gazeta.pl/pieniadze/1,49966,3175789.html?
       as=2&ias=2

       Gestion d’Actifs: Les Economies d’échelle, existent, mais faut-ils les rechercher a tout
       prix? Revue D’Economie Financière, D Negiar and A Hoyos Gomez, 2004.

       Global situation of UCIs, Press           Release,       31   March   2006,   available    at
       http://www.cssf.lu/index.php?&L=1

       http://en.luxembourgforfinance.lu/products-and-services/investment-funds/index.html.

       http://en.wikipedia.org/wiki/Structured_products.

       http://www.abbl.lu/imperia/md/content/abbl/documents/aguide_en.pdf.

       http://www.aberdeen-asset.co.uk/aam.nsf/uk/europeansavingsdirective.

       http://www.afg.asso.fr/afg/uk/french_asset/french_funds_nad.html.

       http://www.alfi.lu/en_index.hcsp.

       http://www.fi.se/upload/90_English/20_Publications/10_Reports/2006/Report2006_1.pdf

       http://www.ftmandate.com/news/fullstory.php/aid/27/Evolution_of_solutions.html

       Interview of Gazeta with Dr. Marcin Dyl available at http://gospodarka.gazeta.pl/
       pieniadze/1,49966,2938617.html

       Interview   with     Dr.       Marcin      Dyl,      http://gospodarka.gazeta.pl/pieniadze/
       1,49966,2938617.html

       Investment Funds in Luxembourg: A Practical Guide, PWC, September 2005.

       Investment Management in Italy 2005, Assogestioni, October 2005.

       Key dynamics affecting the outsourcing market, Mc Cumiskey, E. European Fund
       Administration,       2003.        The       article is   available     under
       http://www.efa.lu/EN/research/articles.html.

       Newsletter, ALFI, 16 February 2006, available at http://www.alfi.lu/en_index.hcsp.

       Out of Sight, Out of Mind: The Effects of Expenses on Mutual Fund Flows, forthcoming
       Journal of Business., Barber, Brad M., Odean, Terrance and Zheng, Lu, 2003. Available
       at SSRN: http://ssrn.com/abstract=496315.

       Parkiet, Polacy zainwestowali, za granico, April 2006.

       Performance persistence in UK equity funds – A literature review, CRA International,
       January 2002 available at www.crai.co.uk




                                                                                            Page 142
Annex B: Bibliography

September 2006                                                                     CRA International




       Pioneer Pekao Website, http://www.pioneer.com.pl/pioneer/arts.tfi_kim_jestesmy.

       PKO Credit Suisse Website, http://www.pko-cs.pl.

       Polish Securities    and   Exchange     Commission     website    http://www.kpwig.gov.pl/
       f1b3_ang.htm

       Presentation given by Rudolf Siebel at the Euroforum conference on 30th May 2006
       provided by the BVI

       Qualitative study among cross-border buyers of financial services in the European Union
       Final Report, Optem, for the European Commission, December 2003.

       Report 2004, IZFIA, available at http://www.izfa.pl/dane/raport_2004.pdf.

       Report on Mutual Fund Fees and Expenses, SEC - Division of Investment Management
       (2000), available at http://www.sec.gov/news/studies/feestudy.htm.

       Report of the Expert Group on Investment Fund Market Efficiency, European
       Commission, July 2006

       Report on the marketing of financial products, J Delmas-Marselet, AMF, November 2005,
       http://www.amf-france.org/documents/general/6383_1.pdf.

       Sandler Review “Medium and Long-Term Retail Savings in the UK” available at
       http://www.hm-treasury.gov.uk/media/606/7A/sandler_ch1to2.pdf.

       Standardization of Funds Processing in Europe, February 2005, EFAMA

       Study of intermediary remuneration for the Association of British Insurers, CRA
       International, February 2005.

       Summary of the Investment Funds market in Poland in 2005, IZFIA, http://www.izfa.pl/
       pdf/podsumowanie_izfa_2005.pdf

       Swedish Mutual Fund Distribution, Cerulli Edge–Global Edition, 1st quarter 2001.
       available  at    http://www.fpanet.org/global/planners/publications/april2001.cfm#CP
       _JUMP_5603.

       Tax discrimination against foreign funds: Light at the end of the tunnel, PWC and EFAMA,
       available         at         http://www.pwcglobal.com/extweb/pwcpublications.nsf/docid/
       2ADB8A098BFDCA47852570B20055B387

       The effect of commission based remuneration on financial advice for the Financial
       Services Authority, CRA International 2002.

       The Future of UK Asset Management: Competitive Position and Location Choice, Oxera,
       May 2005.




                                                                                            Page 143
Annex B: Bibliography

September 2006                                                            CRA International




       The “Luxembourg Model”: a success story, Luxembourg Wort, Investment Fund
       Supplement, B Lockwood, March 2005, available at http://www.efa.lu/EN/research/
       articles.html.

       The Role of Custody in European Asset Management A report for EAMA, OXERA
       (November 2002).

       UCITS III: More Europe in the funds sector, Vander Stuyft, 2001, available under
       http://www.beama.be/content/publicaties/index.php?page=documentatie.

       UCITS III in België, Financeel Forum, De Keyser, E. and Vander Stuyft, J., 2005,
       http://www.beama.be/content/publicaties/index.php?page=documentatie.

       Wertschöpfungskette im Fondsgeschäft in Deutschland verankern – Durch Deregulierung
       den Standort stärken und das Potenzial besser nutzen, Seip, S, Vertragsbeilage
       Investmentfonds, Frankfurter Allgemeine Zeitung.




                                                                                   Page 144

								
To top