Practical insights on implementing IFRS 9 and CECL
ASU 2016-13 and opportunities for implementation efficiencies
The standard setters—the Financial Accounting Standards Board (FASB) and the International Accounting Standards Board (IASB)—have overhauled the accounting models for credit impairment. Institutions with dual filing requirements should formulate a plan that considers where the requirements between FASB’s and IASB’s credit impairment models intersect. They should also identify potential opportunities for gaining implementation efficiencies.
- Synopsis of the expected standards
- Significant credit deterioration
- Data requirements and credit modeling
- Unit of account
- Measuring expected credit losses
Synopsis of the expected standards
Both the impairment model in International Financial Reporting Standards (IFRS) 9 and the FASB’s current expected credit loss (CECL) model are based on expected credit losses. The IASB, however, differs from FASB in that IFRS 9 uses a three-stage approach. Under IFRS 9, debt instruments, excluding purchased or originated credit impaired financial instruments, move through three stages as credit quality changes.
Consequently, a financial institution would measure expected credit losses and recognize interest income depending upon the following stages:
- Stage 1: Assets that are performing
- Stage 2: Assets that have significant increase in default risk
- Stage 3: Credit impaired
In contrast, the FASB’s CECL model requires entities to recognize lifetime expected credit losses for all assets, not just those that have had a significant increase in credit risk since initial recognition. Stated differently, CECL follows a single credit-loss measurement approach, whereas IFRS 9 follows a dual credit-loss measurement approach in which expected credit losses are measured in stages to reflect deterioration over a period of time. Additional differences and similarities in the FASB’s and IASB’s credit impairment models are discussed below.
Significant credit deterioration
A major point of divergence between the FASB’s and IASB’s impairment models is the fact that credit deterioration affects the amount of loss allowance an entity would recognize under IFRS 9. According to IFRS 9, debt instruments are transferred between stages as credit quality changes. Therefore, a critical decision point in implementing IFRS 9 is determining whether there has been a significant deterioration in credit risk since origination. Depending upon whether a financial asset is in stage 1 or stage 2/stage 3, expected credit losses will be measured as 12 months or lifetime.
Data requirements and credit modeling
For all financial institutions, IFRS 9 and CECL will bring a fundamental change in how impairment of debt instruments is measured. In addition to the requirement to model lifetime expected losses, issues around data quality, availability, and collection will likely be at the forefront of implementation efforts.
The following data will likely be necessary to measure expected credit losses under both IFRS 9 and CECL:
- Historical defaults, attrition, and recovery data
- Risk grades
- Delinquency data
- Internal indicators of the likelihood to pay
- Collateral information
- Forward-looking economic scenario
- Macroeconomic variables
- Origination lifetime probability of default
- Loss given default estimates
- Exposure at default estimates
- Full repayment
Unit of account
Under the CECL model, entities are required to evaluate debt instrument assets on a collective (i.e., pool) basis when similar risk characteristics are shared. If the risk characteristics of a given debt instrument are not similar to the risk characteristics of any of the entity’s other debt instruments, the entity would evaluate the financial asset individually. If the debt instrument is individually evaluated for expected credit losses, the entity would not be allowed to ignore available external information such as credit ratings and other credit loss statistics. Under IFRS 9, the expectation is the same. That is, expected credit losses should be measured on a collective basis if the debt instruments share similar credit risk characteristics. This collective assessment is also applicable for determining whether significant increase in credit risk has occurred as well.
Grouping based on similar or shared credit risk characteristics is an area where banks can align their methodology for pooling debt instruments. Consistent practices can be used to group exposures to assess credit risk (such as by product type, product terms and conditions, industry/market segment, geographical location, or vintages) for both US GAAP and IFRS.
Measuring expected credit losses
Both IFRS 9 and the FASB’s CECL model provide latitude in how expected credit losses are estimated—an entity can use a number of measurement approaches to determine the impairment allowance. Under IFRS 9 and the CECL model, information about past events, current conditions, and reasonable and supportable forecasts of future economic conditions should be considered when measuring expected credit losses. The models differ in terms of how the time value of money should be reflected in the estimate of expected credit losses.
Under IFRS 9, for non-purchased or originated credit impaired debt instruments, expected losses must be discounted to the reporting date using the effective interest rate of the asset (or an approximation thereof) that was determined at initial recognition (i.e., time value of money is required to be incorporated explicitly). Under the CECL model, estimates of expected credit losses must reflect the time value of money explicitly only when a discounted cash flow approach is used to estimate expected credit losses.
Building an integrated framework and approach
Adopting an integrated approach for end-to-end process design and implementing standardized processes and controls can significantly alleviate implementation and operational burden. Furthermore, a well-thought-out integrated approach should also lend itself to a consistent framework and is more likely to be accepted by auditors and regulators.