IFRS Disclosures from banks
Justin Bisseker, Equity Analyst at Schroders
Tony Clifford, Partner, Ernst & Young LLP
An analyst’s view
The Investment management industry needs to consider taking a closer look at which banks they would like to have exposure to. When assessing the performance of banks over the last 20 years against the overall market, it is surprising to see that there is zero total return over the period and a 64% underperformance relative to the UK stock market.
However credit losses are currently at a cyclical low, despite the impact of IFRS 9 and despite Price-to-Earnings (P/E) multiples being in line with the rest of the UK market. As such analysts are interested in understanding two key aspects:
- the sensitivity of EPS to credit losses
- comparability of Probability-of-Default (PD) disclosures
Sensitivity of EPS to credit losses
UK domiciled banks have reported additional information to support staging and sensitivity disclosures compared to their European counterparts, however as the reporting is not uniformly consistent, the level of comparability between banks is challenging. This has resulted in reduced comparability and consistency between banks, which is challenging as investors use these disclosures to better understand risk and future expected losses sensitivity.
In particular, a grey area for consideration is how banks differentiate between exposure to Stage 2 assets which are high-risk versus those assets which are still considered to be of better quality but have experienced a Significant Increase in Credit Risk (SICR) since origination. Critical questions to consider in response to the above are:
Comparability of PD disclosures
When assessing disclosures for sensitivity and scenario analysis the following is noted:
- The scenarios are not based on the same up or downside swing, nor probability of occurring, which limits comparability
- It is not always clear whether the PDs generated in the scenarios are for a Point-in-Time (PiT) estimate on a high or a low point in the business cycle
- There is a wide range of Stage 2 assets as a percentage of book even though some are considered good quality assets
- Long term capital impacts are not clear
- Given the current business cycle and economic conditions are positive there appears to be less emphasis from the industry on Stage 2 and 3 assets and their sensitivity to shocks
The following are areas to consider as enhancements to existing disclosures:
- Understanding of Loan-to-Values (LTVs) and collateral held against Stage 2 and Stage 3 loans
- Cured assets and where they have impacted the ECL numbers is not yet clear
- An analysis of the time of assets spent in Stage 3
IFRS 9 Disclosures 2018
An analysis of how the ratios compared across the banks including stage allocation, coverage ratio, and the overall proportion of loans is presented here.
It is noted from this graph that there is a large variability around the disclosed figures. Overall, UK domiciled banks are leading their European counterparts in the quality of their disclosures.
Key observations from the 2018 disclosures are included below.
Movement of loans between stages
When assessing the migration of loan exposures between the stages and most notably the movement from Stage 2 back to Stage 1 it can be seen that this is higher in the UK at a cure rate of 63% in 2018, whereas banks domiciled in the EU had far lower churn between stages and therefore only 39.3% healed from Stage 2 into Stage 1.
Observations regarding the variability in the coverage ratios
Tony performed a more detailed analysis of the retail portfolios of four UK banks, most of the variability for the coverage ratios arises from non-mortgage lending and other retail lending. However disclosures for non-mortgage lending portfolios include less information. One bank disclosed the ECLs as if they had Purchased or Originated Credit Impaired loans (POCI), which helped to allow comparison between the banks.
The level of Stage 3 ECLs also depends on how long such loans are held before they are written off.This is also affected by the level of forbearance, which is particularly high for Irish portfolios.
Disclosure on staging
The analysis on staging is better in some banks than others, where they include useful information such as how the staging is performed and details around which assets are in included in each respective stage.
It would be useful to understand from these disclosures the approximate proportion of ‘good loans’ in Stage 2, given that whilst the banks do disclose the PDs by stage, the disclosure bands differ and so the information is not easily comparable. Additionally,whilst the banks do disclose Loan-to-value (LTV), this factor in particular has a large impact on ECLs and especially on the mortgage book and highlights how secured lending is overall much less likely to require a high ECL provision than unsecured facilities.
Multiple Economic Scenarios (MES)
When considering the impact of using forward looking Multiple Economic Scenarios (MES) on the book, users are able to examine the composition of the provisions, although to varying extents. A positive takeaway from the MES analysis is that management are exercising their discretion to ensure sufficient coverage by applying suitable overlays.