Impairment (IASB only)
Following from Monday’s education session, the IASB held a decision making session to discuss:
- Criteria for recognition of lifetime expected losses
- Methods and information to assess expected losses and transfer criteria
- Disclosures applicable to entities applying the simplified approach for trade and lease receivables.
The staff also provided papers related to advantages and disadvantages of the three-bucket impairment model and the approach in the 2011 Supplementary Document Financial Instruments: Impairment (SD), and a summary of comments received to the SD and a preliminary analysis of concerns raised. However, these papers were not specifically discussed.
The Board considered various alternatives to clarify the criteria for recognition of lifetime expected losses under the three-bucket model.
The IASB previously tentatively decided that an entity should measure the impairment allowance for an asset at lifetime expected losses if, at the reporting date, the probability of not collecting all contractual cash flows has increased more than insignificantly since initial recognition (the credit quality criterion) and is at least reasonably possible (the deterioration criterion). However, the Board has since received feedback from constituents requesting clarification on the application of these criteria.
As such, the staff considered multiple alternatives to clarify the criteria for recognition of lifetime expected losses. All such alternatives considered the interaction between deterioration and credit quality; some providing more subjectivity while others offering more prescriptive criteria.
Based on its analysis, the staff recommended that the three-bucket model be clarified by requiring recognition of lifetime expected losses if there has been a deterioration in credit quality since initial recognition that is significant (when considering the term of the asset and the original credit quality) (the deterioration criterion); and the credit quality of the asset would not be considered investment grade (the credit quality criterion). The staff noted that the term ‘investment grade’ would not be used in the final standard. Instead, this term was being used as an anchor to discuss the principle of the criteria and more descriptive wording would be developed at a later date.
The staff believed its recommendation better reflected the underlying economic deterioration in credit quality and thus was more consistent with the objective to reflect that deterioration by ensuring the timely recognition of lifetime expected losses. They also believed the alternative balanced the benefits (the deterioration in credit risk for both low risk and high risk assets would be relevant) and costs (tracking would not be required for low risk assets) of making the distinction. However, the staff acknowledged subjectivity in this alternative.
Summarising the staff’s analysis, the IASB Chair noted that feedback to the Board’s previous tentative decision suggested many constituents believed the three-bucket model resulted in an allowance balance of a similar amount compared to a day 1 lifetime loss model if any deterioration is expected – which was not a correct assessment in his view. He believed the staff’s recommendation appropriately captured credit deterioration (and more specifically, credit deterioration not captured in the original pricing of the contract) as the primary criteria for recognising lifetime expected losses.
One Board member supported the staff recommendation, but questioned why the staff did not plan to use the term ‘investment grade’ in the assessment criteria. He believed ‘investment grade’ was a generally understood term, and therefore, suggested the term ‘investment grade’ be included in the criteria, along with a definition of the term. However, another Board member was concerned with this view. He believed that using the term ‘investment grade’ would place high reliance on outside rating agencies, and he did not believe that was an appropriate outcome. Instead, he suggested that the term ‘investment grade’ be replaced with a definition more akin to the risk-based definition used by external agencies.
A few Board members believed the staff recommendation placed too much reliance on credit quality deterioration (e.g., considering when credit quality is below ‘investment grade’). Instead, they believed the articulation of the staff recommendation should follow the credit deterioration objective, whereby the credit quality criterion effectively serves as a practical expedient/application guidance in applying the underlying deterioration criterion. Many Board members supported this revision to the staff recommendation.
When put to a vote, the Board tentatively supported the staff recommendation, subject to revising the criteria such that the deterioration criterion acts as the primary criterion, with the credit quality criterion acting as, in essence, a practical expedient.
Following from the discussions on the criteria for recognition of lifetime expected losses (above), the Board discussed the information that should be used for the purpose of assessing the above criteria (i.e., application guidance to assist in the determination of the best available information and assessment of the criteria for recognition of lifetime losses).
The IASB previously noted that an entity should use the best available information to apply the expected loss model. An entity would not be required to undertake an exhaustive search for information. Instead, information should be used that is available without undue cost or effort.
In the determination of the best information available, the staff believed that it should be emphasised that an entity should use information that is as forward looking as possible. They believed this was best achieved through observable inputs such as prices. However, because credit risk is a component of a price and cannot be directly observed, the measurement of credit risk is inherently subjective regardless of the use of such inputs. Therefore information about default expectations and credit quality is more persuasive when several pieces of information corroborate each other.
The staff also noted that if historical inputs are used, those inputs must be adjusted to reflect expected future events. The use of a statistical method alone is not enough to conclude that the information is forward looking or that the measurement of credit risk is appropriate. In that sense, a qualitative assessment based on forward looking information might better reflect the measurement of credit risk than a statistical method using historical data. The staff analysed many methods that might be used to measure credit risk and changes in credit risk which could be considered as inputs to determine expected losses or as information to assess the criteria for lifetime losses in the three-bucket model (i.e., to determine whether the asset is investment grade and whether there has been a significant deterioration in credit quality from initial recognition).
Based on its analysis, the staff recommended:
- the information used to determine expected losses and to assess the need to recognise lifetime expected losses should include probabilities of default, pricing information, credit ratings and other qualitative inputs. The staff provided illustrative examples of application of these methods.
- an entity should use the best available information to apply the expected loss model. However, delinquency information could be used to determine expected losses and to assess the need to recognise lifetime expected losses. This represented a slight revision to the recommendation provided in the staff paper.
- the inclusion of a rebuttable presumption that both lifetime loss criteria (as established above) are met if the asset is 30 days past due, and a required disclosure of how the entity has rebutted the presumption that assets 30 days past due have not met the criteria for lifetime loss recognition.
- that entities be permitted to use 12 month probability of default rather than lifetime probabilities of default to assess the lifetime loss criteria if there is no evidence that the risk curve is abnormal.
In response to the staff recommendations, one Board member, considering the staff recommendation regarding the inclusion of a rebuttable presumption that both lifetime loss criteria are met if the asset is 30 days past due, questioned the appropriateness of 30 days - believing that 30 days past due was not necessarily indicative of an impairment indicator for many entities. The staff noted that delinquency is a lagging indicator, and in the staff’s view, would only be relied on in isolation to assess the criteria when that information is the basis of the credit risk management for the relevant assets and only where more forward looking information is not available. Therefore, the staff believed it would be a relevant consideration in very few circumstances. Further, they noted that 30 days was a number they heard from their outreach activities, but acknowledged that a higher number could be used.
Another Board member asked that in drafting, the staff emphasise, in relation to its recommendation that entities be permitted to use 12 month probability of default rather than lifetime probabilities of default to assess the lifetime loss criteria if there is no evidence that the risk curve is abnormal, that this election should not yield a significantly different impairment allowance result.
With little additional feedback, the Board tentatively agreed with the staff recommendation.
The IASB previously asked the staff to consider the application of the disclosures in the proposed impairment model to entities applying the simplified approach for trade receivables and lease receivables (the ‘simplified approach’). Under the simplified approach, for trade receivables with a significant financing component and lease receivables within the scope of the proposed leases model and IAS 17 Leases, entities would be permitted a policy election to use an impairment allowance measurement objective of lifetime expected losses at initial recognition and throughout the life of the asset instead of applying the three-bucket model. For trade receivables without a significant financing component, an entity would have an impairment allowance measurement objective of lifetime expected losses on initial recognition and throughout the life of the asset.
The staff acknowledged that simplified disclosures would be consistent with the simplified impairment approach. Therefore, the staff recommended that the disclosures previously tentatively decided for general impairment accounting be applicable to entities applying the simplified approach to the extent they apply to the measurement of lifetime expected losses. The staff did, however, recommend the following amendments to existing general disclosure requirements:
- to meet the requirements for the disclosure of their risk profile, entities may use provision matrices as a basis for the disclosure;
- assets accounted for under the simplified approach should disclose the effect of modifications for assets that are 30 days past due; and
- lease receivables, due to their treatment under the Leases project, are additionally excluded from the qualitative description of the disclosure of collateral.
No differentiation would be made between trade receivables with and without a significant financing component.
All IASB members tentatively supported the staff recommendations.
The staff noted that this meeting concluded the technical discussions on the project, absent any necessary sweep issue discussions. The staff intends to discuss the re-exposure period, due process requirements and present a balloting request at a future meeting. An exposure draft is expected to be published during the first quarter of 2013.