FRTB - coalescing of the regulatory storm?

15 October 2015

A few years ago I foolishly tried to sail from Falmouth across the Bay of Biscay in December by nipping between two slow moving deep low pressure systems. Naturally halfway across they coalesced into a Force 10 storm and we had to run for shelter at Gijon on the Spanish coast.

In the world of financial regulation we are facing a similar scenario but with three low pressure systems...! 

The first "low" was the Global LEI initiative (ensure everybody is uniquely identifiable) which meant you could be sure who your counter-party was. Then there was “BCBS 239 - Risk Data Aggregation and Reporting" so you were confident you were managing your risk to a level of quality and detail agreed with your regulator. Now on the horizon is a the third low - "BCBS 265 - Fundamental Review of the Trading Book” which will attempt to normalise the risk models so that the dollar of risk you report is equivalent to the dollar of risk I report.

The final draft of the FRTB regulation is due out any day now after the responses have been collated from the three consultation papers. Most people are not yet aware though of how extensive the changes are going to be to the banks. Unlike BCBS 239 which had a staggered compliance date based on bank categorisation ( 1 Jan 2016 for G-SIB's, 2-3 years from when a D-SIB is nominated) all banks currently are expected to have to comply by the same deadline for FRTB, anticipated to be 2018.

So that’s ok then with 2.5-3 years to get prepared? Well - have you looked at how fundamental the changes are going to be? There are 5 key areas which individually have a significant impact on existing risk management processes and platforms but combine for a perfect storm of ongoing risk change that needs to be planned for now:

1. Expected Shortfall vs. VaR
Under FRTB the regulatory capital charge is to be calculated using Stressed Expected Shortfall (SES) instead of VaR. SES will be based on the average tail beyond the 97.5th percentile using 5 liquidity horizons (10, 20, 60, 120, and 250 days) with various risk weights applied depending on which of the 5 asset classes the instrument falls into. The SES data will be calibrated to a period of market stress and will replace VaR and SVaR to eliminate double counting of the risk. Diversification will be limited by regulators through the use of a correlation diversification parameter

2. Desk based run instead of Legal Entity
Model approval will be granted on an individual desk basis and SES will be run daily by business unit (desk). As a expected a monthly report has to be sent to the bank's Supervisor but the data also has to be available on an ad-hoc basis. 

3. Mandatory use of Standard Approach in addition to IMM Approved Models
Whilst an approved internal model can be used all banks must also run the standard model. A Sensitivity Based Approach (SBA) is currently under review but the actual approach will be confirmed in the Final Draft. Securitisation will only run under the Standard Approach.

4. Restriction on transfers between Trading Book and Banking Book
Assets can only be moved between the Banking and Trading book with Supervisory approval, and if they are then the most conservative capital charge has to be maintained via the Standard Approach

5. Removal of CVA hedges from the Trading Book (under discussion in d325 - Review of CVA Risk Framework)
Due to the complexity and model risk of the proposed changes to the CVA calculations the respective CVA hedges will probably have to be removed from the Trading Book and managed in a specific CVA Book

Conclusion and Impacts

As a result of FRTB there will have to be significant changes to existing risk engines to support both standard and internal models on a daily basis (many are legacy platforms with limited upgrade capability). Model approval will be necessary at a desk level (which potentially is a 9-18 month programme).
There will have to be extensive infrastructure changes to support the storage of risk model runs for ad-hoc reporting requirements. 
Processes and procedures will need to be updated to support the restrictions on the Trading and Banking books, and governance processes reinforced to ensure adequate controls are in place. 
So hopefully your BCBS 239 programmes to improve data quality and lineage will be winding down in time to provide the underlying data needed to support the work required for FRTB?

Either way, you’ll still need to batten down the hatches and lock in your good resources to continue to weather the ongoing regulatory storm!

Courtesy of James Hope-Lang