UK Finance has responded to the Prudential Regulation Authority's (PRA) CP 17/18 which proposes changes to the Definition of Default used in credit risk modelling. These are in line with recent Technical Standards and Guidelines released by the European Banking Authority (EBA). Designed to introduce comparability and consistency to the supervisory oversight and use of Internal Ratings Based (IRB) credit risk modelling in order to restore confidence in global risk models, this is an objective UK Finance strongly supports.

The CP proposes a materiality threshold against which a past due loan should be assessed for the purposes of identifying a default, the consequence of which is that a bank must hold extra regulatory capital. More significantly, the PRA has proposed the replacement of the 180 days past due (dpd) component of the Definition of Default with a 90dpd test, removing the current national discretion permitted by European legislation - a change that has been recommended by the EBA.

Our response pointed out that 90dpd is not a strong indicator of default and that a shorter period to default may limit the ability of banks to work with customers to support them if they experience payment difficulties. I should add that this will have no direct impact on customers who will continue be treated fairly under the FCA rules if they fall into arrears.

It also raised concerns about the pace of transition from the 180dpd, which is currently used by most large mortgage lenders. There is no doubt that the transition to 90dpd will require significant modifications to modelling practices at a time when many other modelling change initiatives are also being introduced. These include IFRS9, stress testing, the wide ranging ?not Basel IV? changes, as well as modifications to regulatory reporting requirements. All these will be implemented to a similar but not coincident timeline. With more than 1,000 individual models needing attention this will be a mammoth task for banks and the PRA, not least because of the current uncertainty about the exact nature and timings of these different but complex and inter-related initiatives.

In my view there is a significant risk of volatility in prudential capital ratios if all these changes are introduced sequentially over the course of a number of years, with a consequent and possibly detrimental impact on investor perception and the pricing of bank capital.

A much better solution would be to introduce these changes, including the 180dpd Definition of Default, at the same time as, for instance, the ?not Basel IV? changes by the beginning of 2022. Not only would this reduce the model approval and implementation workload for banks and their supervisors alike, but it would also avoid sending investors the wrong signals if regulatory capital ratios prove to be as volatile as I suspect they will be.