Skip to main content

The IFRS 9 Staging dilemma

One requirement, different interpretations, multiple solutions

The IFRS 9 standard is not prescriptive on what ‘Significant Increase’ means or how to identify a SICR event, with the consequence that setting triggers for SICR is a critical judgment that needs to be made by each lender.  The forward-looking sensitivity (and volatility) inherent in IFRS 9 ECL arises principally from choices around SICR implementation and the use of forward-looking information.  The decisions lenders make are important for loan loss reserving and also for stress testing, capital, and prudential soundness under a real-world stress should one occur.We observe that, even after almost six years since the implementation of IFRS 9, a consistent “best practice” for SICR has not yet emerged and scant literature has been published on the topic. This has led to considerable variation in the proportion of loans in Stage 2 for apparently similar portfolios. 

As an example, the chart below shows the range of proportion of loans in Stage 2 for the largest UK banks at 30 December 2023 for the three main asset classes of mortgages, consumer lending and wholesale lending.  Indeed, this lack of consistency has been a key point in the annual IFRS 9 thematic feedback letter from the UK’s Prudential Regulation Authority for the last couple of years, where there was a call for wider use of industry standard metrics.

Whilst there is certainly a lack of consistency within the market in the approach to setting SICR thresholds, we do observe a shift towards the use of the confusion matrix and associated metrics from this to achieve a statistically justifiable quantitative threshold position. It is important to note that even with the confusion matrix, there is no “one size fits all” metric that solves the SICR challenge perfectly, because when improving a metric focussed on one side of the confusion matrix other metrics generally deteriorate. As a result, there is no single answer and an arsenal of metrics should be used to find a range in which the threshold could justifiably sit, with the final threshold based on an optimisation across these metrics based on the specific portfolio. 

In our report, we look to shed some light on the key challenges that lenders are facing in setting SICR thresholds. We cover items across data availability, compliant SICR approaches and how to find the “right” threshold, providing key considerations to effectively overcome each of the issues identified.

Meet the author

Richard Tedder


Richard leads the credit risk team in Deloitte’s Audit and Assurance practice. He has deep expertise in end-to-end credit risk modelling, measurement, data, controls, and governance. He leads a team of c. 60 credit professionals delivering audit, assurance and advisory projects to more than 40 UK lenders ranging from the largest Tier 1 banks to growing non-bank lenders.