The Fundamental Review of the trading book or FRTB regulations had the final paper released by the BIS on the 15th Jan 2016. You can download this paper here for your perusal, but read on to get a short and succint summary. I will be posting more more detailed notes on some of the individual aspects of these regulations, focusing on the challenges to implementation and the points to be aware of in the future. However for now here is the FRTB summary.
FRTB Regulations : Why do we need them?
The FRTB regulations are a response to a pre-crisis framework that is now deemed inadequate and weak in many areas. This is particularly true of the definition of the boundary of the trading book. There is also a feeling that the IMA approach left a lot of gaps that need to be closed down for a better regulatory capital framework.
Tail risk is also something that the current VaR approach does not capture adequately along with liquidity (or illiquidity). Most IMA based approaches also allow for generous diversification effects as they are based on historic parameters which definitely do not hold in a crisis situation (correlation largely become relevant in very stressed markets).
The current standardised approach is highly inadequate in that there is not clear linkage between the IMA and the SA. There is a lack of risk sensitivity in the current standardised approach which needs to be addressed, along with constraining the benefits of diversification and hedging.
FRTB – A high level summary of the main changes
Revised Boundary – This will set a clear boundary between the banking and the trading book, and the aim is to reduce arbitrage of regulatory capital between the 2 books. There are stricter limits along with capital disincentives for transfers between the 2 books. Coupled with strict reporting guidelines and regulatory oversight this should allow for a much better framework that governs the boundary between the 2 books under the FRTB regime.
Revised Internal models approach – The aim here is to capture the effect of tail risk more effectively and also capture liquidity effects. Tail risk is captured moving from a VaR based approach to a approach that is based on Expected shortfall (in theory at least the jury is out on if ES actually captures tail risk adequately. There are also a series of liquidity horizons (10 to 120) based on asset / risk classes which define the liquidity holding period from a capital calculation point of view. Finally there will be a stress calibration period which will do away with he need for a separate stressed VaR calculation.
Under FRTB the approval of internal models now moves to a desk level approach, which will put more onus to comply at a trading desk level and if not be moved onto the standardised approach. Trading desks will have to show that their models comply by showing that the have adequate PnL attribution and backtesting in place. It is important to note that PnL attribution here means model based PnL will be compared to Risk based theoretical PnL in order to measure that the risk models adequately capture the risk from the models.
There are also more constraints on the benefits to risk numbers from hedging and diversification under FRTB – effectively constraining the benefits of hedging and diversification. There will also be a separate charge for non-modellable risk factors (given by a series of guidelines around observable data points).
The revised standardised approach – This has undergone significant change. This will be used for banks who want to remain with a a simple straight forward model but is also the fallback for banks where the IMM does not gain approval. The major methodological change here is that the approach is now based on risk sensitivities across asset classes. The revised SA should provide a consistent way to measure risk across geographies and regions, giving regulatory a better way to compare and aggregate systemic risk. The sensitivities based approach should also allow banks to share a common infrastructure between the IMA approach and the SA approach. Thera are a set of buckets and risk factors that are prescribed by the regulator which instruments can then be mapped to. Aggregation across buckets and risk classes are subject to constraints and will not allow full diversification.
There is also a standardised default risk charge which is added on along with an add-on charge for residual risk which is harder to model.