IFRS9 Implementation Challenge For Low Default Portfolios

Catarina Souza, Director, Risk Services EMEA at S&P Global, Market Intelligence, looks at one possible approach to compute PDs after IFRS9 goes live. Catarina also presented a webinar, "When the Rubber Meets the Road on IFRS9 Implementation", watch this on demand.

This article was originally published July 2017, by S&P Global, Market Intelligence, click here for the original publication. 

As of January 1, 2018, IFRS9 will replace the current IAS 39 across several jurisdictions, including many European countries.

By focusing on expected credit losses, IFRS9 will represent a significant shift from IAS 39 (incurred losses) since the new impairment requirements determine that expected losses will have to be computed not only for non-performing assets, but also for performing assets, with a direct impact on Profit and Loss (P&L).

Financial institutions, particularly banks, will be required to recognize an expected loss allowance of either 12-month expected credit loss for assets classified under stage one (performing) or lifetime expected credit loss for assets classified under stage two (under-performing).

Given the different requirements under IFRS9 compared to the Basel requirements, adjustments to existing Basel-related systems, including data and models, will be a must in order to comply with IFRS9.

A particular challenge will arise regarding the computation of Probability of Default (PD) for IFRS9 purposes for Low Default Portfolios (LDPs).

In order to address this challenge, S&P Global Market Intelligence has developed a tool that converts the PDs based on long-term average default rates (typically used for Basel IRB approach) into forward-looking PDs, taking into account current and future economic conditions. The adjustment factor, called the Z-Factor*, will adjust the average observed long-run PDs to reflect a certain set of systematic scenarios thereby predicting the expected PDs for IFRS9 purposes. The illustration for the computation is provided in the figure below:

The overlay considers the following three-step approach:

1 Estimate Long-Term (LT) PDs

To begin the process, one would need an assessment of the counterparty’s credit quality (i.e. a credit score or a rating), and the associated long-run PD for the score/rating.

S&P Global Market Intelligence scorecards are credit quality assessment tools that  are developed based on S&P Global Ratings criteria and produce a forward-looking assessment, which can be mapped to probabilities of default (PDs) derived from observed default rates of ratings from S&P Global Ratings (over 36 years of default data).

By using our scorecards, it is possible to cover low-data asset classes for all geographies (e.g. corporates, banks, insurers, other FIs, project/asset finance, and sovereigns), and obtain an assessment of the counterparty’s credit quality and respective PD.

2 Determine Z-Factors

The IFRS9 standard would expect one to use PDs which resemble observed default rates to calculate expected losses, therefore the average long-run PDs derived in step one will need to be adjusted to meet IFRS9 requirements.

The adjustment will take into account changes in the systematic environment, while company-specific changes are reflected in the credit assessment, which by definition is forward-looking.

S&P Global Market Intelligence’s tool will compute the expected Z-Factor considering economic indicators, market indicators, and industry indicators, and benchmarking these to multiple Z-Factors. The final Z-Factor is computed as a weighted average of the Z-Factors derived for each indicator.

3 Derive Point-In-Time PDs

Once forward-looking Z-Factors have been established, the final step involves adjusting LT PDs, LT (t), by the predicted Z-Factor to estimate the forward-looking IFRS9 compliant PD for year t.

In order to obtain the lifetime PD term structure, the LT marginal PD for each future year is adjusted with a Z-Factor which is specific for that year. These marginal PDs are then combined to form a PiT cumulative term structure.

*The “Z-Factor” here denotes the adjustment performed using S&P Global Market Intelligence’s tool, and should not be confused with the “Altman’s Z-Score” used for measuring credit risk.

Interested in other regulations impacting the risk management industry, like FRTB and stress testing requirements? Join the leading risk management event for the latest expert insights.