{{vm.title}}
25{{vm.includesText}}
{{item}}
{{vm.selectText}}
{{ vm.welcomeLabel.trim() }},
{{ vm.userName }}
{{ vm.text }}
{{vm.newUser1}} {{vm.newUser2}} {{vm.newUser3}} {{vm.newUser4}}
{{ vm.unknownUser }}
{{ vm.title }}
{{ vm.title }}
{{ vm.title }}
- {{ vm.siteSelectorList.flyout.cell1.heading }} {{ site.countryLocale }} {{ vm.siteSelectorList.flyout.cell1.global.countryLocale }}
{{ vm.closeTabLabel }}
- {{ vm.flyout.cell2.title }} {{ vm.flyout.cell2.newTabAllow }} {{ vm.flyout.cell2.title }}
- {{ vm.flyout.cell3.title }} {{ vm.flyout.cell3.newTabAllow }} {{ vm.flyout.cell3.title }}
- {{ vm.flyout.cell1.viewAll.label }} {{ vm.flyout.cell1.viewAll.newTabAllow }} {{ vm.flyout.cell1.viewAll.label }}
ECL models: A deeper dive | Real-time IFRS 9
Real-time IFRS 9
Ewa Bialkowska
Ewa specialises in accounting for the banking and financial services sectors, with over two decades of experience in auditing and delivering accounting advisory services to financial institutions in the UK, continental Europe and other jurisdictions.
IFRS 9 bank disclosures
A deeper dive into what we know so far
28 March 2018
Since our last blog, we have taken a deeper dive into the detail that banks have disclosed to date. We have looked at information published by 14 large international banks from Canada and Europe in their 2017 annual reports, IFRS 9 transition reports and first quarter 2018 reports.
We have focused on disclosures on IFRS 9’s expected credit loss (ECL) model, as this is where banks have spent most of their effort when implementing the new standard.
Assessing significant increase in credit risk
In our last blog, we looked at the way several banks allocated their portfolios to different stages of credit deterioration under the ECL model. Here, we look in more detail at how they have done it.
Some banks have provided quite a bit of detail when describing their approach to assessing a significant increase in credit risk (SICR) – the trigger for moving a financial asset from Stage 1 to Stage 2 in the impairment model. Most of them intend to apply a mix of quantitative and qualitative criteria, and a backstop.
The most common quantitative measure is probability of default (PD) during the remaining life of loans and other exposures – known as lifetime PD. Some explained that they intend to use internal credit ratings for wholesale exposures, and 12-month PD (i.e. probability of default in the next 12 months) specifically for retail.
Some gave more detail of what relative (and absolute) movement in PD they will use. For example, some banks defined SICR as a doubling of the PD estimated on origination (further refined for individual portfolios), although others indicated that they will use a smaller increase.
As regards qualitative information, a number of banks mentioned three possible indicators of a SICR, including: moving an exposure to a ‘watch list’ where credit monitoring is more intensive; initiating forbearance measures on an exposure; or receiving adverse information on a borrower from credit bureaux.
A number of banks emphasised the importance of expert credit risk judgement in making the assessment.
Some other areas of interest are as follows.
- 30dpd backstop: Almost all banks in our sample stated that they will apply the 30dpd (days past due) backstop for their SICR assessment. This means that for those banks, all financial assets that are more than 30 days past due will move to Stage 2.
- Low credit risk exception: As we expected, and in line with the Basel Committee’s guidance, this is not widely discussed. The exception allows a bank to allocate an asset (or other exposure) to Stage 1 without assessing whether credit risk has increased significantly, as long as the asset is of sufficiently high quality (commonly understood to mean investment-grade status with the major credit ratings agencies). A few banks that discussed this topic intend to apply the exception to debt securities where external credit ratings are available and/or to some other commercial exposures.
Incorporating forward-looking information
IFRS 9’s ECL impairment model requires forward-looking information to be incorporated for the remaining life of an exposure – one of the key changes from the old standard, IAS 39, with its incurred loss approach.
The new impairment model is a probability-based measurement and includes the impact of different possible outcomes. Operationally, this can be difficult to implement and selecting relevant economic scenarios can be challenging.
All banks in our sample discussed their approach and disclosed how many different economic scenarios (i.e. differing assumptions about key drivers of credit risk, such as GDP growth) they will use. As illustrated in the chart below, it looks like the most popular option is to use three scenarios covering a base or central outcome, as well as the possibility of one better and one worse outcome.
Click to enlarge chart (JPG, 31 KB)
Using macro-economic variables
Building economic scenarios involves determining the key economic variables that drive credit risk for loans and other exposures. Almost all banks disclosed what they consider these key drivers to be. Several banks disclosed this information on an overall basis only, but some provided separate disclosures for wholesale and retail.
Click to enlarge chart (JPG, 53 KB)
What next?
All eyes will be on the publication of first quarter results by European banks in around a month’s time. We can expect more detail on how policies have changed due to the new requirements and what the quantitative impact on transition is. It will also be interesting to see how the first quarter disclosures of banks in Europe compare to those of their Canadian counterparts who have already published their first quarter results.
We will publish an update once the first quarter results have been released, so stay tuned to Real-time IFRS 9 as the story continues to unfold.
