The revised market risk framework, known as the "FRTB" (Fundamental Review of the Trading Book) is scheduled to come into effect on January 1st, 2020. The final document has been released by the BCBS (Basel Committee of Banking Supervisors) on January 16th, 2016. Quantitative impact studies (QIS) have shown a significant increase of the regulatory capital charge for trading activities. This does not come as a surprise as it was clearly the main objective of the reform, following the financial crisis of 2008-2009.
Source: Rod Waddington
As banks are getting ready for this new market risk measurement framework, the impact on risk management governance, processes and the underlying IT infrastructure begin to be revealed as well. IT impacts in particular are related to some significant new features, which entail increased data management and data analysis requirements and increased computational burden.
The significant new features that will need to be implemented obviously include the complete overhaul of the Standardized Approach (SA), now based on a “Sensitivities Based Approach” (SBA) and the new requirement for Internal Model Approach (IMA) to be based on Expected Shortfall (ES) measurement instead of VaR (Value at Risk). Both are based on a more granular and layered calculation approach, with the introduction of variable liquidity horizons per asset class. Then there is the introduction of DRC (Default Risk Charge), which replaces the IRC (Incremental Risk Charge), and has to be calculated for the equity asset class as well as interest rate and credit.
Both approaches require large volumes of historical data, with reliable data analysis and data cleansing features, as the Expected Shortfall in particular is very sensitive to data outliers. Moreover, the IMA requires the identification of modellable and non-modellable risk factors based on stringent data quality requirements. As failure to comply with these requirements would entail a supervisor decision to revert to the Standardized Approach with greater capital charge, there is clearly a strong need for large, reliable datasets of market data.
Whether a bank decides to qualify for IMA and compute the Expected Shortfall or stick to the standard Sensitivities Based Approach, it has to implement and apply the SBA anyway. Indeed, this approach is mandatory for all securitization products, it is required to provide a credible fallback in case of failure to qualify for the IMA, and moreover, it has to be calculated and disclosed for all trading desks. The revised SA, now based on the SBA, is more computational intensive than the previous SA. Regarding the Expected Shortfall, it will have to be calculated both at the bank-wide level and at each trading desk level, as approval for the IMA will be granted for each trading desk individually. In order to obtain this approval, banks will need to perform daily P&L attribution and backtesting, which are computational intensive exercises as well.
All in all, the new framework seems to be intended to provide an incentive for banks to use the market risk capital charge calculation as a daily monitoring tool and not just yet another mandatory regulatory reporting burden. However, this objective will become a reality only if banks succeed in incorporating the new requirements in their daily P&L and risk calculations processes.