PwC on Basel Committee’s Views of its Fundamental Review of the Trading Book

On January 26 the Basel Committee on Banking Supervision (BCBS) released its first set of Frequently Asked Questions (FAQs) on the Fundamental Review of the Trading Book (FRTB). The BCBS published the FRTB in January 2016 with the intent to harmonize (i.e., reduce variability) the treatment of market risk across national jurisdictions.[1] It will generally result in higher global capital requirements.

The BCBS calls for each jurisdiction to finalize implementation of the FRTB before January 2019 and for compliance to begin by December 2019. We do not expect US regulators to adopt the standard until 2018 at the earliest as they are likely to wait for any changes to the framework that may result from the BCBS’ recalibration efforts and other evolving global standards. In addition, uncertainty arising from the Trump Administration might lead to further delays or modifications to the FRTB.

  1. The FAQs left the most important questions unanswered. Although the FAQs shed light on a number of issues pertaining to both the Standardized Approach (SA) and the Internal Models Approach (IMA), they failed to clarify several areas where the industry had been seeking guidance. For example, the FAQs did not address some important issues surrounding P&L attribution, Non-Modellable Risk Factors (NMRF), and a potential floor for regulatory capital calculations using the IMA.[2] This capital floor is a big issue as regulators have grown skeptical of the capital determinations banks estimate using internal models. Further, some of the FAQs posed new interpretive questions. For instance, the FAQs clarified that banks could align their P&Ls for different time zones, but did not provide guidance on potential approaches.
  1. Hedging benefits for trading activities are not generally available. The FAQs clarified that hedging benefits for market risk are only available if the hedge matches the hedged position exactly (e.g., back-to-back trades).[3] The increased cost of hedging resulting from this clarification will reduce motivation to hedge market risk, thereby reducing trades in illiquid products. This cost could, however, be shifted to the end-user as banks will demand higher prices for hedging customized products, thereby forcing end-users to manage their risk themselves. This would not be a desirable outcome for end-users as they may not be equipped to manage market risk.
  1. Banks can use internal pricing models to calculate SA sensitivities. In a win for the industry, the FAQs clarified that banks can use internal pricing models in order to calculate SA sensitivities, provided the results are close to those prescribed by the FRTB and approved by regulators. This will reduce operational complexity for the industry, particularly because most pricing models already calculate sensitivities for use in market risk management.
  1. Opacity for Hypothetical P&L (HPL) exclusions raise additional questions. The FRTB explicitly requires that credit valuation adjustments and debit valuation adjustments be excluded from the HPL.[4] While the FAQs clarified that bid-ask spreads and valuation adjustments measured less frequently than daily can also be excluded from HPL, they fail to clarify which valuation adjustments are included and which are excluded. As a result, the implementation of P&L attribution (which requires each desk to measure its HPL) will be dependent on additional guidance by the regulators on the correct definition of HPL.[5]
  1. Stricter requirements for risk factors to be deemed modellable. Under the FRTB, only risks that meet strict data availability and quality requirements are deemed modellable. All other risks (i.e., NMRFs) must be accounted for by a catch-all capital charge which is calculated for each NMRF based on a risk-specific stress scenario. The FAQs emphasize that banks have to prove real prices (e.g., by presenting proof of the transaction) for the risk factor in order for it to be considered modellable. In addition, nominal prices are no longer permitted.[6] Further, the FAQs did not provide guidance to the industry as to whether banks can combine data in order to prove the modellability of the risk factors, or if they must rely on their own data. This provides a challenge for banks who are not able to prove the required data points (i.e., real price data) on their own.
  1. Trading desks will likely segregate securitizations. The FAQs clarified that if a desk’s portfolio contains securitized products, the entire trading desk must be capitalized under the SA. As a result, banks will likely separate their securitizations from other products to avoid having to capitalize other products under the SA.
  1. The treatment of backtesting exceptions at the desk-level are clarified. The FRTB allows banks to disregard firm-wide backtesting exceptions[7] arising from NMRFs, as long as they are approved by the regulator. In a win for the industry, the FAQs clarified that this also applies to desk-level backtesting. As a result, banks can attribute their backtesting exceptions to trading outcomes linked to NMRFs, allowing them to avoid having to put banks’ risk models under unnecessary scrutiny.
  1. Different treatment for equities and fixed income securities. The FAQs clarified the differences in treatment for equities and fixed income securities for the DRC calculation. While the FRTB allows the use of a 60-day liquidity horizon for equities, the FAQs clarified that banks can also apply the 60-day horizon for correlation calculations to portfolios predominantly made up of equities. However, for mixed portfolios, the standard horizon of one year remains. The FAQs, however, provided some relief by allowing the probability of default for equities to be rescaled to 60 days. In addition, the FAQs stated that longer liquidity horizons are permitted when appropriate.
  1. The Internally Modeled Capital Charge (IMCC) remains a challenge. The FAQs clarified that the multiplier used in the IMCC charge calculation may be updated weekly rather than daily. However, banks must also have the ability to update the multiplier daily should a supervisor request it. As a result, banks will still need to implement a daily calculation. Banks also need to have procedures and controls in place to ensure that the weekly calculation does not differ significantly from the daily calculation.
  1. Capital floors will remain a key issue for discussion in March. The BCBS was originally meant to meet on January 8th to finalize its outstanding bank capital rules. However, following the opposition of certain EU member countries, the meeting was postponed until March. As mentioned above in point 1, the capital floors will be one of the main issues discussed, especially since US regulators have supported moving away from the IMA while European and Asian authorities have not.

ENDNOTES

[1] See PwC’s First take, Ten key points from Basel’s Fundamental Review of the Trading Book (January 2016).

[2] The FRTB requires all banks to calculate their market risk capital under the SA, but banks can also apply to use the IMA to calculate such capital for certain trading desks. While the BCBS has stated that a percentage of the SA capital charge will act as a floor banks using the IMA, it has yet to provide clarity on the percentage of the SA that will act as a floor.

[3] Back-to-back trades consists of two trades, one facing the internal desk and one facing external counterparties. The market risk of the trades cancel each other out, leaving the owner of the trade with no market risk exposure.

[4] The HPL is calculated from yesterday’s portfolio using today’s data.

[5] The P&L attribution test is a requirement for each of a bank’s trading desks. Failure bars the desk from using an internal model to calculate its own capital requirements. As a result, the failing desk will have to apply the SA, leading to a higher capital requirement.

[6] Nominal prices are defined as prices of the transactions carried out for the purpose of creating a valid data point, e.g., a minimal volume transaction.

[7] Backtesting is the process of testing a risk measure. A backtesting exception occurs when the trading outcome is not covered by the risk measure.

This post comes to us from PwC and is based on the firm’s First take, “Ten key points from Basel’s FAQs on the Fundamental Review of the Trading Book,” dated February 3, 2017, and available here.