Amongst mortgage lenders, variations in banks' Risk Weighted Assets (RWAs) are observed in the market. This information is published in firms' Pillar 3 disclosures, as well as by the Prudential Regulation Authority (PRA). As an example, in January 2015 the PRA reported performing risk weights ranging from 10.8% to 14.6% for performing residential mortgages with 70%-80% Loan To Value (LTV) amongst firms using the Internal Ratings Based (IRB) approach to calculating RWAs. The variations lend credence to a distrust of banks' internal models by regulators, shareholders and the general public.

The PRA has issued Consultation Paper CP29/16, in which updates to Supervisory Statement 11/13 (SS11/13) are proposed. CP29/16 expands the PRA's expectations of compliant approaches, with key points including:

  1. The PRA believes that Point In Time (PIT) Probability of Default (PD) models do not adequately reflect the long term nature of mortgage lending, and so proposes that banks should not use PIT models for mortgages.
  2. The PRA believes that Through the Cycle (TTC) PD models are not able to differentiate between cyclical and non-cyclical (e.g. internal policy and process changes) changes in default risk, and so proposes that banks should not use TTC PD models for mortgages.
  3. The PRA proposes that firms incorporate the 1990s recession into their mortgage PD model calibration and cyclicality assessments.
  4. The PRA proposes a 30% cap on the level of cyclicality assumed in mortgage PD models.
  5. The PRA proposes a 25% floor on peak-to-trough house price falls assumed in mortgage LGD models.

The consultation comes at a time when many firms' mortgages IRB models are reaching the end of their service life, but are also used for a wide variety of internal applications including debt pricing and estimation of credit losses for IFRS 9. If incorporated into the next update of SS11/13, action is likely to be required by all IRB mortgage lenders, lenders with IRB aspirations, and standardised lenders with proxy-IRB Pillar 2 models. Although IRB firms are likely to take the opportunity to critically revalidate their mortgages RWA models, the potential downstream impacts are likely to complicate the scope and scale of remediation activity, as well as influence firms' solution designs.

The PRA has requested responses by 31 October 2016.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.