Over a year has passed since the IFRS 9 implementation deadline of 1 January 2018, but many firms are still facing challenges today. A year ago most firms were in the developing stage. For many, IFRS 9 represented transformational change to their existing infrastructure but, since then, some firms have already begun using IFRS 9 models for forecasting, stress testing, and IRB profitability. However, for most, the main focus is simply on learning how to run IFRS 9 models effectively.

In a recent webinar, Jaywing covered the four key areas currently surrounding impairment forecasting under IFRS 9, discussed in detail below. These are especially pertinent given the recent PRA review, which provided an update and the main thematic findings from written auditor reporting.

IFRS 9 - beyond implementation

Now that the models are live, firms are continuing to embed IFRS 9 as a BAU process - this includes the development of monitoring frameworks, improved governance, and ongoing enhancements to address weaknesses in the first-generation models.

Poll question: How would you describe the current state of IFRS 9 implementation in your organisation?

Unsurprisingly, most firms have redevelopments underway or in plan to enhance their IFRS 9 solution. While the majority consider themselves compliant, in reality it is a moving feast as firms evolve and establish new strategies in line with regulatory steer.

Monitoring performance of forward-looking models

It is important to understand performance on an ongoing basis, in order to provide confidence to stakeholders around the accuracy of impairment estimates and satisfy auditors that the models remain fit for purpose. The forward-looking nature of IFRS 9 imposes an extra time dimension over standard monitoring practices which, coupled with the need to isolate errors from modelling and economic forecasting, creates additional complexities around effective model monitoring.

Poll question: How advanced is IFRS 9 model performance monitoring in your organisation?

There is a mix in levels of advancement, which may reflect variations in organisation size, model complexity, and timing of impairment disclosures. The expectation is that more and more firms will develop and enhance automated monitoring frameworks as the benefits become clear, amid increased expectation across the industry.

Best practice in significant increase in credit risk

There is a variety of approaches in place to allocate accounts into stages under IFRS 9, particularly in terms of quantitative measures of significant increase in credit risk (SICR). Our experience suggests that a combination of relative and absolute thresholds is often appropriate. There is further variation around the specific metrics in place to derive the thresholds, although as noted in the recent PRA review, some industry-standard approaches are emerging in this regard.

Poll question: Which of these bests describes your current approach to SICR?

More than half of firms regard their SICR framework as simple, with use of quantitative thresholds. The level of complexity should typically account for the size and nuances of a portfolio, and it may well be that a simple framework can be suitably effective and easier to maintain. Again, comprehensive monitoring helps to demonstrate that the thresholds in place remain appropriate, via a continual assessment of the accounts allocated into Stage 2 against suitable performance measures.

Limitations of using historical data in macroeconomic modelling

Arguably the biggest challenges faced by firms reporting under IFRS 9 surround the development and application of macroeconomic models. These include a lack of representative data covering a downturn period, concerns over whether past relationships between macroeconomic factors and risk are likely to hold in the future, and the difficulty of ensuring forecast scenarios and their weightings are realistic. These are not easy problems to solve, although there is a recognition that the industry needs to evolve  its approaches, which may include consideration of more sophisticated modelling techniques going forward.

Poll question: How advanced is your approach to economic modelling under IFRS 9?

The majority of firms indicated a need for some level of manual overlay. This is likely to change over time given the PRA's recent focus on improvements to core models and accordingly less reliance on Post-Model Adjustments (PMAs).

Join our upcoming webinar on 11 June to find out more about overcoming the challenges of stress testing under IFRS 9.