EU latest proposals for “Key Information Document” and Risk Ratings Key findings of ABI/IMA Research 23 September 2009 Julie Patterson Director, Authorised Funds & tax What is the KID? “Key Features” according to CESR: A tool to help retail investors make informed investment decisions Not primarily a marketing document Not an investor education document Not encumbered with info serving only legal or regulatory requirements Need not address investors’ information requirements on an ongoing basis Plain language and enabling comparisons between funds Generally no more than two pages/sides What will it tell you? Key Elements: Names of fund, manager and promoter/group Fund objectives and investment strategy Material risk/reward factors likely to affect fund Past performance Charges (payable directly or indirectly by investors) Treatment of income In two pages! Risk Rating The indicator should be: Applicable to as many funds as possible Based on robust methodology with no room for manipulation Clear that categorisation does not imply any guarantee Simple as possible Easy to implement for managers Allow easy and effective supervision by regulators Be stable and robust against normal changes in the risk of capital markets “The indicator must not be misleading” Methodology proposed by CESR Historical volatility with use of a VaR measure for structured funds Based on investment returns over a 3-year period (or 5 years for funds with returns less frequent than weekly) For new funds, use available data extended by a series “representative of the way in which the fund would have behaved in the past” For funds with a markedly asymmetrical distribution, use a VaR approach Risk buckets A numerical scale of 1-6 Prescribed bucket boundaries - two options: - Split fund universe along risk continuum (possible bunching of funds) - Split fund universe into equal-sized sub-sets (some buckets v large or v small in risk terms) Migration rules to minimise bucket hopping – three options: - No rule - “Observation period” (e.g. change has to have held for last three months) - Different entry/exit thresholds around the bucket boundaries The BIG ISSUE is frequency of bucket hopping ABI/IMA Research The aim of the project: • To identify whether there exists a simple and reliable measure of investment risk • To suggest how measure could then be used to categorise (bucket) funds • To suggest what structure would be needed to implement the process • To consider whether the same measure could/should be applied to structured products and absolute return funds What do Managers currently do? Fund managers that risk rate their funds do not “return rate” them Managers estimate risk inherent in broad asset classes rather than for individual funds. Funds are then fitted into the asset class risk spectrum The way in which managers assess asset class risk varies greatly, but the methodologies are generally based upon measures of historic volatility. The results in terms of risk ordering seem to vary less Managers also use their discretion based upon their professional expertise to adjust the risk order of asset classes that seem to be ‘out of line’ There is no clear consensus on how to group asset classes into risk buckets, or on how many buckets to use. It is perhaps here that there seems to be the greatest diversity of approach to risk rating Structured products and absolute return funds seen to pose particular issues The risk measures tested The research considers: Maximum drawdown Range Standard deviation VaR Sharpe ratio Downside deviation Sortino ratio Beta How reliable are the measures? Correlation coefficients using 36 month pre-assessment and 12 month post assessment periods respectively 100.0% 80.0% Spearman's rank correlation test 60.0% 40.0% 20.0% 0.0% -20.0% -40.0% -60.0% 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 Stdev Max draw Range Down dev Sharpe Sortino Beta • Standard deviation and Range appear to be the most “reliable”, Sharpe and Sortino the worst • Under the assumption of normality, the rankings using VaR and Standard deviation will be observationally the same. However, standard deviation operationally easier to compute and a more commonly used measure. VaR measures very sensitive to assumptions. How to bucket funds? 40.0 Stdev Range Max Draw Down SD Risk level, relative to "cash" (full sample) 30.0 20.0 10.0 0.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Asset class The risk continuum is not smooth. Assets in the top sextile are not six times more risky than those in the bottom. Risk bucket boundaries Standard deviation Range Max drawdown Downside deviation Asset class Bucket Asset class Bucket Asset class Bucket Asset class Bucket Money market 1 Money market 1 Money market 1 Money market 1 UK direct property 1 UK direct property 1 UK Gilts 1 UK direct property 1 UK Gilts 2 UK Gilts 1 Defensive 1 UK Gilts 1 Defensive 2 Defensive 2 UK direct property 1 Defensive 2 £ fixed interest 2 Sterling long bond 2 Global fixed interest 2 £ fixed interest 2 Global high yield 2 Cautious 2 Sterling long bond 2 Cautious 2 Cautious 2 £ fixed interest 2 Cautious 2 Global fixed interest 2 Global fixed interest 3 Global fixed interest 3 £ fixed interest 3 Sterling long bond 2 Sterling long bond 3 Global high yield 3 Balanced 3 Global high yield 3 Balanced 3 Balanced 3 Global high yield 3 Balanced 3 Flexible 4 Global equities 4 Global equities 4 Global equities 5 North America 4 North America 4 Japan 4 Flexible 5 Global equities 4 Europe excl UK 4 North America 4 North America 5 UK all companies 4 Flexible 4 Europe excl UK 5 UK all companies 5 UK equity income 5 Japan 4 Flexible 5 UK equity income 5 Europe excl UK 5 UK equity income 5 UK smaller comps 5 Europe excl UK 5 UK smaller comps 5 UK smaller comps 5 UK all companies 5 Japan 6 Japan 5 UK all companies 5 UK equity income 5 Commodities/energy 6 Private Equity 6 Property securities 6 Property securities 5 UK smaller comps 6 Commodities/energy 6 Commodities/energy 6 Commodities/energy 5 Property securities 6 Property securities 6 Private Equity 6 Private Equity 6 Private Equity 6 Asia Pacific excl Jap 6 Emerging equities 6 Asia Pacific excl Jap 6 Emerging equities 6 Emerging equities 6 Asia Pacific excl Jap 6 Emerging equities 6 Asia Pacific excl Jap 6 The bands are the sextiles The ordering is similar but differs, and in one case there are no 4s, so the choice of metric matters # of switches De fe 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 n C a s iv u e B a t io u UK la s nc U K a ll F le e d c x s m o m ib l UK a pa e l e q le r n ie s u co G ity mp l o in s E u ba l co m ro e q e p u A s N e e it i e i a o rt x c s Pa h A l U cif m K ic er Em ex ica cl er Ja gi n g J ap p eq a n ui £ t S t f i x U K ie s e e G G r l i n d i n il t s lo g b a lo t e r l n es G f ix e g b t lo d o U K b a i nt n d l h er eighteen year period is high P r d i r e ig h e s t op ct y i Number of switches e r p r e ld ty o C o M se p er m o n c u ty m e y r it od ie Risk bucket boundaries i t ie m a s P r s / e rk e iv a n t t e e rg Eq y ui ty Number of switches greater than 2 categories Using 3-year calculation period, the number of one category switches over an Risk bucket boundaries 16.0 3yr 5yr 10yr 14.0 12.0 s 10.0 e h c t i w 8.0 s f o r 6.0 e b m u 4.0 N 2.0 0.0 Increasing the calculation period increases stability Using 5-year or 10-year data, no switches of two or more categories Research recommendations The risk standards should focus on the risks inherent in asset classes Slight preference for downside deviation as the standard measure of risk, since it focuses on the part of the return distribution that preoccupies investors in practice. Of the other measures, standard deviation does best. The risk metric would be used to bucket funds based on their underlying asset classes This calculated bucket would represent the minimum acceptable rating for a product focusing on that asset class. Managers can use intuition to increase risk rankings but not reduce them. A long data span should be used to calculate risk metrics – ideally at least 10- years. This will help improve the stability of the risk bucketing and ensure the data accounts for at least one business cycle and facilitate independent verification. The bucket boundaries should be set using as long a data span as possible. Research recommendations Multi-asset class funds can be “slotted in” to the risk continuum by analysing the risk characteristics of their asset mix relative to the single asset classes. Capital guarantee funds are effectively a combination of a zero-coupon bond and a call option. They can be ranked using a combination of the guarantee and the asset the investor is exposed to. Intuitively its ranking will be above the underlying zero coupon bond, but below the underlying asset. Income-enhanced structured products generally involve the selling of insurance, which is complex to value and the value is not independent of the model used. They should be included in a separate “complex” risk category, where it is made clear that the investor is effectively selling an insurance product, rather than on the main scale. Ranking absolute return funds should probably be based on the asset mix. An adjustment to the risk calculations should be made to reflect the leverage allowance in the fund. Timetable Responses submitted to CESR 10 September ABI/IMA Research to be published CESR to report to Commission by 30 October Implementation July 2011?
Pages to are hidden for
"Consumer Panel - Risk Ratings"Please download to view full document