The Fundamental Review of the Trading Book (FRTB) is the biggest and most significant market risk regulatory change since the introduction of Value at Risk almost two decades ago. On the surface, the changes required under the new framework appear quite
clear. However, the waters get murkier as one ventures deeper. The impact of FRTB goes far beyond the move from Value at Risk (VaR) to Expected Shortfall (ES) and cannot be overstated. It will necessitate significant changes in business model, governance,
processes, methodology, data management and technology infrastructure.
The latest regulation tries to address some of the important lessons learnt in the aftermath of the financial crisis. It was observed that the existing VaR models (unsurprisingly) failed to capture the magnitude of the losses once the VaR was breached (i.e.
the tail region), rendering the market risk capital inadequate. Liquidity in the markets evaporated during stress, resulting in loss crystallisation as banks were forced to hedge as the prices were falling, further contributing to the negative feedback loop.
In addition, the risk and capital numbers, which were highly dependent on the banks’ internal model in the absence of industry wide homogeneity, were not comparable among banks depriving the regulators from a reliable picture of banks’ health. The different
capital treatment of the trading and banking books also left the door for regulatory arbitrage open.
The three main areas of impact of the FRTB are as follows:
Revised Trading Book/Banking Book boundary
FRTB introduces detailed guidance on allocation of activities to the trading book and banking book. It also restricts any movement across the trading/banking book boundary. In the exceptional cases where the movement is allowed, the capital treatment is
uniform, thus removing any possibility of arbitrage. The trading desks are now subject to the supervisory approval.
The banks will need to review their activities and book/desk structure in order to determine the classification of activity under the new framework. This would also necessitate migration of positions across the trading/banking book boundary with potential
implications on capital requirements.
Mandatory Standardised Approach (SA) calculation
To improve comparability and provide a fall back that is model independent, the regulation introduces mandatory calculation and disclosure of SA both at the desk level and bank wide. The SA linear risk charge is calculated using sensitivities to prescribed
risk factors and regulator prescribed risk weights and correlations (purportedly calibrated to the stress period). An additional charge is calculated for the non-linear risk using stress scenarios. A default risk charge and residual risk add-on is added to
the sum of above two charges.
SA based capital charges are higher compared to the existing SA and are expected to be higher than IMA based capital charges due to multiple factors including limited risk factor sensitivity and constrains on correlations among various risk factors. In a
significant move, the capital charges for all securitised products will be based on SA.
At present it’s not clear if the regulators would use the resulting SA capital charge as a floor or add-on to the IMA based capital requirements.
Changes to the internal models-based approach
Prior to being able to use internal models for market risk capital calculations, each trading desk needs to gain separate approval for its models. FRTB introduces rigorous PnL Attribution requirements in addition to VaR based backtesting as a hurdle to IMA
Use of ES as the primary risk measure should improve the capture of tail risk compared to VaR measure. In addition, FRTB introduces varying liquidity horizon in ES calculations, based on the risk factor liquidity varying from 10 to 250 days. The regulation
states full revaluation method for calculating ES, while allowing for grid based methods as alternative. The correlations between risk factor classes are constrained in the bank wide ES model.
In order to be a part of the ES model, the risk factors need to fulfil strict modellability criteria. For the risk factors deemed unmodellable, the capital is calculated based on stress scenarios (worsening to worst loss based capital if regulator doesn’t
accept the proposed stress scenario).
In an apparent relief for the banks, the implicit double counting due to summing VaR and SVaR is now replaced with single stress calibrated ES measure and the regulatory multiplier is reduced from 3.x to 1.x.
This article only scratches the surface of the complex behemoth that is FRTB. Although still in consultation, the main points are nearly crystallised, allowing banks the opportunity to plan and formalise the implementation strategy.
The effects of FRTB are profound and far reaching at all organisational and technical levels of a bank. Across the spectrum, changes are mandated in terms of additional and more stringent regulatory requirements on data, models, reporting, governance and
management. Faced with the changes of regulatory methods and capital charges, banks will also see a need and opportunity to amend internal risk measures and to integrate further the management of regulatory charges, risk and return.
The necessary adaptations of the business model can only be achieved with a flexible calculation architecture that empowers a bank to clearly identify drivers of both the capital charge changes and the total new capital charge. Integration with Front Office
and pre-trade risk, capital and cost transparency will be crucial to ensure competitiveness.
FRTB arrives close on the heels of other sweeping regulatory change programmes, such as BCBS239, which have already highlighted the importance of strategic versus tactical choices. These lessons must now be carefully applied to FRTB, in order to establish
flexible and efficient infrastructure and processes.