PC Findings _amp; Prudent Valuation

Document Sample
PC Findings _amp; Prudent Valuation Powered By Docstoc
					Product Control Findings and Prudent Valuation

      Valuations & Product Control Team

                Ragveer Brar

                  November 2010

    Product Control Findings & Prudent Valuation

•    Introduction                                     3
•    Executive Summary                                5
•    Project Findings
      – Culture and Governance                        8
      – Staffing                                      9
      – Front Office                                 10
      – Daily P&L                                    11
      – Post Trade Controls                          13
      – Model Risk and Validation                    15
      – Valuation Framework                          16
      – Reserves                                     19
•    Bid Offer Reserve Thematic Review               21
•    Prudential Valuation Framework and Next Steps   27


• In August 2008, the FSA issued a Dear CEO Letter to the Industry
  outlining our key concerns in the area of Valuations and Product
  Control. The letter outlined concerns around control failings that had
  led to several mismarking incidents such as:

- Senior FO staff did not exercise adequate oversight of traders’
- PC staff were unable to adequately challenge FO; and
- Inadequate valuation policies and procedures.

• The Dear CEO Letter also stated that ‘The complexity of firms’ trading
  book operations have been continually increasing over the last ten
  years. Firms must be prepared to take extra steps to deal with the
  incremental risks associated with less liquid, complex and/or
  modelled products. The increased valuation uncertainty that can be
  associated with these products is likely to precipitate the need for
  increased focus on the regulatory principles of prudent valuation,
  which are an important component of the overall regulatory capital
  framework. Adjustments to Tier One capital are required where
  prudent value is assessed as being materially below fair value’.


• We have visited a sample of banks with material trading operations to
  assess their Product Control (PC) functions in their widest context,
  and have also begun to assess banks’ approach to applying
  prudential valuation principles (Dear CEO Valuation Project - Phase A
  and B).
• This presentation sets out our findings and observations for Product
  Control, across the firms included, at the time of the review. We have
  fed back bilateral observations to each of the firms involved for Phase
  A in both meetings and through written communication. There remain
  many areas where we will need to conduct further work and
  investigation with firms and this work will progress on a bilateral
• A more detailed document outlining the findings, associated risks and
  proposed solutions will be sent to all relevant firms shortly. Feedback
  from firms will be welcome as this document will form a key
  component in defining our expectations of PC functions.

                   Executive Summary

The Dear CEO Review of Valuations and Product Control sought to
establish a clear picture of industry practice.
Although there are some areas that most firms do relatively well (e.g. the
independence of P&L production & the NPA sign-off process), there are
areas of dispersion across the Industry where some firms are performing
well, whilst others are not. Many of these were a surprise as we would
have expected all firms to have good controls (e.g. consistency of
market data within firms and Model Val restriction monitoring).
In addition, there are some areas where we feel that the industry as a
whole has some way to go. Understanding and reporting valuation
uncertainty is an example of this that we will return to.
On a positive note, we have been encouraged that many firms appear to
have already started to make significant progress over the last year on
key issues such as P&L attribution, P&L commentary and IPV reporting.

                   Executive Summary

However, there are still some major areas of concern:
  Several PC functions do not appear to have sufficient authority / ability
to provide effective challenge to Front Office (FO). Examples of positive
indicators would be timely IPV adjustments, the burden of proof resting
with Front Office, PC having the ability to force sales to prove marking
and effective management review of the IPV process.
   Insufficient investment in PC infrastructure has left processes manual
(e.g. IPV), incomplete (e.g. P&L explain) and often inconsistent across
business groups.
  Firms are not effectively quantifying, reporting and limiting valuation
uncertainty or differentiating between prudent valuations and accounting
fair value. Comprehensive model risk frameworks are similarly lacking.
  The level of off-shoring is significantly higher than we had anticipated,
thus increasing inherent control risk. The elements of PC off-shored are
the vanilla areas, which have traditionally been where PC staff have
gained experience. This raises long-term concerns about developing the
London PC management of the future.

The main areas which were reviewed were:

• Culture and Governance
• Staffing
• Front Office
• Daily P&L
• Post Trade Controls
• Model Risk and Validation
• Valuation Framework
• Reserves
• Prudential Valuation

               Culture and Governance


1.   Global policies are not always translated into documented
     procedures, and procedures are often inconsistent across

2.   In more than half of the firms, systems are inconsistent across
     regions and/or across business lines.

3.   In more than half of the firms, daily P&L and IPV reports are not
     produced at Legal Entity level.

4.   In our opinion, in approximately half of the firms, there are serious
     concerns around the effectiveness and independence of PC, and
     PC’s ability to challenge FO.



1.   A few firms are over-reliant on temporary staff (industry ranges 0%
     to 28%).

2.   There is significant dispersion in the ratio of FO to PC average pay
     (industry ranges from 5:1 to 2:1 for London staff) as well as in the
     absolute pay levels (range from £45k to £95k as average London
     total compensation). Lower pay is likely to be reflected in the
     quality and status of the PC staff.

3.   This, combined with a high staff turnover (range of 12% - 37%
     annual external to the company), can cause decreased
     effectiveness and result in a poor reputation for the PC function.

4.   The use of off-shoring arrangements within the Product Control
     function is more prevalent than thought at the start of this review
     and can represent nearly 50% of the Product Control headcount.
     Firms generally offshore the controlling of vanilla products and
     reconciliation processes.
                       Front Office


1.   In approximately half of the banks, FO responsibilities were not
     formally documented.

2.   Most banks do not have consistent and documented firm-wide
     procedures for FO spot checking pricing.

3.   A few of the banks do not have formalised price maker / price taker

                          Daily P&L


1.   P&L explain / attribution varies by business line (often system
     driven) and where P&L explain does exist, it is not tracked and
     explained over time. Most firms are also not sufficiently
     investigating unexplained P&L through consistent thresholds for
     investigation and escalation.

2.   Most banks do not have the capability to comprehensively
     calculate P&L explain /attribution, using full revaluation or the risk
     based method. Some areas are seeking to move towards solely
     relying upon a sequential shock approach.

3.   P&L commentary is often inconsistent and standards are neither
     clearly documented nor monitored.

4.   Although firms review P&L vs budget and prior year results, only a
     few PC functions are conducting effective analytical review of P&L
     over time compared to the desks’ business strategies.

                         Daily P&L


5.   Approximately half of the banks do not meet the standards
     expected for new deal review and sign-off. Half of the firms do not
     have the means to systematically identify eligible trades.

6.   Most firms have no specific controls in place to prevent P&L
     smoothing intra-month.

7.   More than half of the firms do not report separately the impact of
     cancels and amends and the impact of new trades in the daily P&L.

                  Post Trade Controls


1.   Approximately half of firms restrict the balance sheet
     substantiation process to the Financial Control function (without
     involving PC or Operations) or limit the QA of the process to close

2.   Over half of firms don’t have granular and statistical reporting on
     ledger manual adjustments and breaks.

3.   Half of the firms don’t have effective review and authorisation of
     manual adjustments.

4.   Two thirds of the firms haven’t established any dependency
     between the trade validation and reconciliations processes, and the
     P&L production process.
                  Post Trade Controls


5.   In most firms, Market Risk use different market data sets for VaR
     purposes to Product Control.

6.   Only a few firms have market data repositories feeding across
     business units.

7.   Completeness of reconciliations of internal (including
     intercompany) trades varies widely across firms.

              Model Risk and Validation


1.   Firms do not have a model risk framework that attempts to quantify
     model risk and ensures that senior management are aware of the
     materiality of the uncertainty this creates (GENPRU 1.3.19). As a
     result, firms have not expressed a risk appetite for model risk.

2.   Most banks do not define appropriate risk limits and reserving
     methodologies for unapproved models. Where restrictions on
     model use exist, they are often not monitored consistently across
     business lines.

3.   A few banks do not have a global model library. This may allow
     different models to be used for the same products in different areas
     of the bank.

4.   In more than half of banks, the initial sign-off and re-review of
     models is not being performed on a timely basis.

                 Valuation Framework


1.   Most firms have inadequate IPV policies commonly omitting
     some of the following components; effective stratification of
     results based on quality of IPV testing, aggregated assessment
     of IPV completeness, and effective layers of materiality
     thresholds for IPV adjustments.

2.   IPV coverage data is commonly reported by parameter. Whilst
     some firms have told us that this is an area they are starting to
     focus on, no firm has yet developed an approach for the
     aggregation and the reporting of untested inventory (or quality
     of IPV testing) across parameters and businesses. In addition,
     we have seen proxy tested positions included as ‘tested’.

3.   The IPV process in more than half of the firms is highly manual
     and spreadsheet based, impacting on internal consistency of
     approaches and on the timeliness of adjustments.

4.   Where there is a valuation dispute, more than half of firms do
     not take an adjustment until the dispute has been resolved.
                 Valuation Framework


5.   In most firms, there is no clear policy defining the hierarchy of
     market data to be used.

6.   Approximately half of the firms do not have procedures that
     would enable escalation of material initial IPV variances

7.   There is little evidence of firms making use of exit P&L or long-
     run P&L analysis (e.g. drip-loss on trades with large day 1 P&L)
     to check the net position valuations produced by the IPV
     process and associated reserving methodologies.

8.   Most firms lack adequate processes to systematically identify
     and assess stale prices. Only 1 firm was seen to have a process
     for reviewing the static component for a spread which is marked
     over a reference rate.

                  Valuation Framework


9.   Most firms do not have a process to systematically identify and
     assess aged inventory across the firm.

10. In approximately half of the firms, IPV teams did not have
    sufficiently active involvement in the valuation of collateral or
    use the available information effectively .

11. Firms do not have a valuation uncertainty framework that
    attempts to quantify valuation uncertainty and report this to
    senior management in aggregate. As a result, firms have not
    expressed a risk appetite for valuation risk.



1. There is a wide dispersion in the bid offer spreads applied by
   firms. For example, the most conservative bid offer spread
   applied to the 1 yr point of the EUR yield curve was 15 times
   greater than the most aggressive.

2. Most firms do not review their bid offer spreads on a monthly

3. There is a wide variation in the netting approaches and
   granularity of bucketing used in the bid offer calculation. For
   example, the number of buckets used for the EUR yield curve bid
   offer calculation ranged from 5 to 21.



4.   Only a few firms aggregate risk at a desk level, the remaining
     firms aggregate at a business, entity or group level.

5.   The policies of approximately half of the firms do not require
     explicit liquidity and concentration reserves to be applied.

6.   More than half of firms do not take separate skew bid offer
     adjustments. This is often due to system deficiencies.

       Bid Offer Reserve Thematic Review


   –    A thematic review was conducted for the calculation of
        Yield Curve (& Tenor basis), Equity Index Vega and Crude
        Oil Vega bid offer reserves for positions as at COB 30th
        June 2010. The 9 participating firms were asked to
        recalculate these reserves using a methodology
        incorporating simple netting rules and pre-defined spreads.

   –    The simple FSA approach is itself more aggressive than
        would be preferred as it allows full netting at entity level
        across products. In addition, we did not incorporate any
        liquidity or concentration adjustments, nor did we choose
        the most prudent approach to spreads or bucketing.

       Bid Offer Reserve Thematic Review

Overview (Continued)

   –    The results of the reserve calculations using the simple
        FSA methodology were used to normalise the firms’ results
        so that they could be compared with each other and
        against the benchmark provided by the FSA methodology
        for a hypothetical portfolio.

   –    The normalisation process was chosen to set the most
        conservative reserve in each category to 100 million and
        then scale each firm’s reserve accordingly (using the size
        of the corresponding FSA reserve for each firm as a factor).

       Bid Offer Reserve Thematic Review


   –      The normalised bid offer reserves for the three market
          parameters are shown in the following table. The
          equivalent result per the FSA methodology is also shown
          for comparison.

   –      In the calculation of the equity index vega and crude oil
          reserves all but 2 of the firms had to apply variations on the
          skew netting approach defined by the FSA due to system
          limitations. The mapping assumptions made will have led
          to some distortion of the results and this has not been fully
          investigated. However, we do not believe that this would
          materially alter the findings.

                Bid Offer Reserve Thematic Review


            – The table below illustrates our findings from the bid offer
              thematic review.

                  Hypothetical                          Hypothetical Equity                       Hypothetical
                  Yield Curve b/o                       Index Vega b/o                            Crude Oil Vega
Firm              reserve (mn)        Firm              reserve (mn)          Firm                b/o reserve (mn)
Most Conservative           100.00    Most Conservative              100.00   Most Conservative              100.00
FSA                          27.10    FSA                             78.13                                    91.58
                             22.81                                    76.87   FSA                              32.57
                             18.34                                    73.78                                    28.67
                             17.01                                    57.38                                    22.21
                             10.52                                    54.30                                    21.04
                               4.77                                   47.41                                     3.34
                               3.63                                   29.32   Most Aggressive                   2.11
                               2.21                                   22.50   Mean (excl FSA)                  29.88
Most Aggressive                1.42   Most Aggressive                 17.13
Mean (excl FSA)              20.08    Mean (excl FSA)                 53.19

    Bid Offer Reserve Thematic Review


  – The results show a concerning level of dispersion that will
    require further regulatory consideration. The ratio of the
    highest reserve to the lowest is 70 for yield curve!
  – For yield curve and equity index vega the vast majority of
    firms are significantly more aggressive than the already
    aggressive FSA methodology.
  – The observed dispersion is a function of both the bid offer
    spreads employed and the implicit netting in the approaches
  – The most conservative reserves in each category are all
    driven in part by wider spreads than used in the FSA

     Bid Offer Reserve Thematic Review

Summary (Continued)

   – The most conservative yield curve reserve incorporates a
     separate concentration reserve. Removing the concentration
     reserve reduces the ratio between the most aggressive and
     conservative reserves from 70 to 28.

   – However, since many firms do not have a policy or process
     to include an additional charge for concentration, this merely
     serves to highlight the potential impact of such an omission.

                   Prudent Valuation


• Trading book positions are subject to prudent valuation rules.
  (GENPRU 1.3.39)
• Trading book positions must be re-valued at least daily. (GENPRU
• Where marking to market is not possible, a firm must use mark to
  model. (GENPRU 1.3.17)
• A firm must ensure that its senior management are aware of the
  positions which are subject to mark to model and understand the
  materiality of the uncertainty this creates in the reporting of the
  performance of the business of the firm and the risks to which it is
  subject. (GENPRU 1.3.19)
• Where independent pricing sources are not available or pricing
  sources are more subjective, prudent measures such as valuation
  adjustments may be appropriate. (GENPRU 1.3.27)

                   Prudent Valuation


• A firm must consider the need for establishing reserves for less liquid
  positions. (GENPRU 1.3.32)
• Factors to be considered when implementing valuation adjustments or
  reserves include; unearned credit spreads, close-out costs,
  operational risks, early termination, investing and funding costs,
  future administrative costs, model risk, hedging costs, availability of
  market quotes, volatility of bid offer quotes, volatility of trading
  volumes, market concentration and ageing of positions. (GENPRU
• If the result of establishing adjustments or reserves is a valuation
  which differs from fair value then a firm must reconcile the two
  valuations. (GENPRU 1.3.34)
• Reconciliation differences under GENPRU 1.3.34 should be disclosed
  to the FSA in prudential returns. (GENPRU 1.3.35)

                   Prudent Valuation
The August 2008 Dear CEO Letter stated that

- Based on our review work, most firms are not in a position to
  rigorously or systematically apply the prudent valuation principles.

- As the prudent valuation principles are associated with valuation risk
  and uncertainty, we believe there could be synergies from
  incorporating risk management skills into the frameworks.

- Firm’s IPV processes are generally focussed on deriving a point-
  estimate of price that satisfies fair value accounting standards. This
  does not involve explicit assessment of the range of uncertainty.
  Analysis undertaken could be enhanced to facilitate the systematic
  quantification of valuation uncertainty.

- By developing better frameworks for the measurement of residual
  valuation and model risks after the application of IPV would be better
  placed to deliver a control framework consistent with the prudent
  valuation principles.

  Prudent Valuation Framework & Next Steps

• Based on our recent review, firms are not disclosing valuation
  differences or making the prudent valuation adjustments required. In
  addition, firms are not defining a prudent valuation framework. This is
  very disappointing given the explicit messages that were delivered in
  the Dear CEO Letter more than two years ago.
• It is important to understand that there is a difference between fair
  value valuation and prudent valuation. Whilst many firms state that
  they believe they comply with prudent valuation standards we noted
  that one of the firms commented that the bid offer spreads and the
  methodology provided by the FSA for the bid offer review seemed
  overly conservative, and that the impact calculated seems to relate to
  prudence rather than fair value.
• Prudent valuation adjustments which are not allowed under
  accounting standards include valuation uncertainty adjustments
  (model, parameter) and level 1 concentration reserves.
  Prudent Valuation Framework & Next Steps

• All of the components of prudent value previously mentioned (e.g.
  liquidity, concentration, model risk) combine to create valuation
  uncertainty. It can be difficult to precisely attribute uncertainty, even
  in instances where valuation uncertainty clearly exists.
• The FSA is implementing a new assessment framework for prudent
  valuation to promote a more systematic approach for assessing
  valuation uncertainty / prudent valuation.
• The first assessment date will be 31 December 2010, with firms
  expected to report initial prudent valuation adjustments to the FSA in
  Feb 2011, along with policies and procedures defining the firms’
  valuation uncertainty approach and framework.
• Thereafter, there will be quarterly reporting by the end of the month
  following quarter end.

  Prudent Valuation Framework & Next Steps

Approach to assessing valuation uncertainty

• Determine the range of plausible valuations for all fair valued
  positions by position or parameter. This would require determining
  the appropriate holding period (e.g. firms may choose a 10 day
  horizon) and degree of certainty for prudent valuation (e.g. firms may
  choose a confidence interval of 90%). This would yield the difference
  between accounting value and prudent value by position / parameter.
• Aggregate results to arrive at a total impact of the prudent valuations
  (e.g. firms may choose to use a simple sum approach within asset
  class and a correlation of 0.5 between asset classes).

  Firms must construct an approach that they believe is appropriate and
  compliant. FSA expects to open bilateral discussions with all firms to
  discuss methodologies between November 2010 and February 2011.

  Prudent Valuation Framework & Next Steps

• Where the application of prudent valuation would lead to a lower
  carrying value than actually recognised in the accounting, the value of
  the difference should be deducted from Core Tier 1 capital.

• The recent Trading Book DP is consulting on whether there should be
  an explicit Pillar I capital charge for valuation uncertainty (as is
  already the case for market, credit and operational risks) as opposed
  to a capital deduction.

• We expect to review and formally assess each firm’s prudent valuation
  framework and methodology through comprehensive PC reviews and
  close & continuous meetings with firms around IPV and P&L. We will
  also be assessing the granular judgemental analysis by firms at a
  parameter or product level on a rolling thematic basis.

  The FSA’s views on prudent valuation may be refined during 2011 and
  we expect further formal interaction with the industry. This is
  consistent with the Trading Book DP recommendation of more robust
  guidelines to ensure that firms adopt prudent valuations.

Shared By: