Understanding ECL-based Loan Classification and Provisioning under IFRS 9 and its Pros and Cons

In the evolving landscape of global banking regulation, one of the most significant changes has been the shift from the traditional "incurred loss" model to the Expected Credit Loss (ECL) model under IFRS 9 – Financial Instruments. This forward-looking approach marks a paradigm shift in how banks assess and account for credit risk in their loan portfolios. Instead of waiting for a credit loss event to occur before recognizing it in the books, IFRS 9 requires banks to anticipate potential credit losses based on past events, current conditions, and future forecasts.

At the heart of this model is a three-stage classification system. In Stage 1, banks recognize a 12-month ECL on performing loans where credit risk hasn’t significantly increased. If the risk increases substantially but the asset is not yet impaired, it moves to Stage 2, where a lifetime ECL is recognized. Loans that are credit-impaired fall under Stage 3, also requiring lifetime ECL provisioning, but with more conservative assumptions and a focus on recovery values. This tiered structure ensures that the financial statements reflect the evolving credit quality of assets more realistically.

The ECL model under IFRS 9 relies heavily on data — historical credit performance, forward-looking macroeconomic indicators (such as GDP growth, inflation, or unemployment), and scenario analysis. Consequently, banks need robust IT infrastructure and analytics capabilities to implement and operate such systems effectively. This includes developing credit risk models, enhancing data collection frameworks, and training personnel.

Formula for Expected Credit Loss (ECL):

Where:

• EAD (Exposure at Default): The total value a bank is exposed to when the borrower defaults.

• PD (Probability of Default): The likelihood that the borrower will default within a given time horizon.

• LGD (Loss Given Default): The portion of the exposure that is expected to be lost, after accounting for recoveries.

Example:

Imagine a bank in Dhaka, Bangladesh—let’s call it “Unity Bank Ltd.”—which granted a term loan of $1,000,000 to a mid-sized garment exporter, “Textura Apparels Ltd.”. The company had always been a reliable client, but in 2025, the global textile market started showing signs of slowdown due to declining demand from Europe. Unity Bank’s risk team, following the ECL-based provisioning under IFRS 9, reevaluated Textura's credit risk.

Based on internal ratings and updated macroeconomic forecasts, the Probability of Default (PD) for Textura was estimated at 5% over the next 12 months. The Loss Given Default (LGD), based on collateral value and recovery history in similar cases, was estimated at 40%. The Exposure at Default (EAD) remained at $1,000,000.

Then:

$$\text{ECL}=1,000,000\times 0.05\times 0.40=20,000$$

The bank should provision $20,000 for this loan as per the ECL model under IFRS 9.

Below is a table summarizing the pros and cons of implementing ECL-based loan classification and provisioning under IFRS 9:

Pros	Cons
Forward-looking risk assessment enhances predictive credit loss accuracy	High implementation cost (systems, training, and data requirements)
Improves financial transparency and global reporting standards	Operational complexity due to advanced modeling needs
Strengthens credit risk management per Basel Committee recommendations	Requires detailed historical and forward-looking data
Aligns with international regulatory compliance (e.g., IFRS 9)	Transition from rule-based to ECL model is challenging and gradual
Enables proactive provisioning to reduce economic procyclicality	Dependence on macroeconomic forecasts adds model risk
Promotes capacity building and IT infrastructure upgrades	May necessitate external expertise and continuous training