Financial instruments — Impairment
Application of the three bucket model to debt securities
The Boards first discussed whether the guidance around grouping of financial assets was appropriate for debt securities. The Boards felt the guidance would work for debt securities and should not change existing practice of evaluating debt securities on an instrument-by-instrument basis.
The Boards then considered whether the list of indicators decided upon during the previous day were sufficient for both loans and debt securities or whether debt securities, given they often had fair values quoted in active markets, required some rebuttable presumption indicating recognition of lifetime expected losses was appropriate when the fair value was less than the cost basis for a predefined term (either percentage decrease or percentage decrease for a period of time).
The staffs noted that when the fair value of the debt security exceeded the cost basis that in most cases an entity could conclude that recognition of lifetime losses was not required. One IASB member questioned when the fair value exceeding the cost basis could result in recognition of lifetime expected losses. The staff responded that, while remote, there could be a possibility of an instrument with a high coupon in a low interest rate environment where the issuer was experiencing credit deterioration such that the instrument would trade at a premium as a result of the yield above market rates even after taking in to account the credit deterioration of the issuer.
One IASB member raised an issue with the 'change in market indicators of credit risk' indicator decided upon the previous day noting the guidance for securities discusses 'the length of time and the extent to which the fair value of the debt security has been less than the amortised cost'. His concern focused on the trigger established by the Boards for movements to bucket two did not include a time based concept and he feared establishments of brightlines in application. The staffs acknowledged his concern and said they would consider how to clarify the indicator so as not to establish brightlines.
One IASB member then asked the staffs whether debt securities in buckets two or three would be eligible to return to bucket one. The staffs responded that their intention was to keep the impairment model the same for loans and debt securities and that since loans could return to bucket one after being moved to bucket two they felt that debt securities should do the same. This raised the issue with other Board members on whether the impairment model should be symmetrical or not. The FASB Chair raised the issue of a restructured debt and how that would be treated when considering whether to move back from bucket three in to bucket one. Several Board members said the Boards should examine how the indicator approaches when applied to improving credit conditions.
The Boards agreed with the staff that the indicators were sufficient to apply to both loans and debt securities and that no rebuttable presumption (e.g. brightline) should be applied for evaluating debt securities for credit deterioration when fair value is less than the cost basis.
The Boards then discussed the application of an expected value approach in estimating expected losses for debt securities. The Boards had previously decided that expected losses should be estimated with the objective of an 'expected value' approach. However, constituents have raised concerns over application of an 'expected value' approach to assets assessed for impairment on an individual basis as it results in an expected loss that is not one of the possible alternatives and does not consider the fair value of the collateral for secured lending arrangements.
During the March 2011 Board meeting, the Boards had agreed that the final standard would clarify that:
"in practice, a concrete estimate of an expected value would not require the use of every single possible outcome. Rather, in the case where there are many possible outcomes, a representative sample of the complete distribution can be used for determining the expected value of the credit loss. In identifying that sample, the entity would need to take into account only the information that is available about the outcomes. It would not have to (and should not) make up anything else."
The staffs asked the Boards whether additional application guidance should be provided for individually evaluated financial assets. The FASB Chair recommended that the existing guidance in US GAAP with respect to consideration of collateral values be carried forward in to the impairment model. She acknowledged this was not an expected value but would simplify the approach in certain instances. One IASB member expressed concern with inserting US guidance in to IFRS but acknowledged that application guidance should be provided and it should be clarified that a most likely outcome approach is not consistent with an expected value. The FASB Chair asked the IASB member if a loss rate approach were utilised which dropped the high and low end possibilities from consideration if that was viewed as an expected value approach. The IASB member said he would view it as a good proxy for an expected value approach. A FASB member said he agreed with the IASB member but thought the Boards should include some practical expedients are proxy approaches to expected values.
The staffs said they would work on developing further guidance around methods that could reasonably be used to achieve the objective of an expected value approach. The Boards suggested the staff look to the decisions made in the revenue recognition and insurance projects regarding expected value approaches.
Application of the three bucket model to loans
As noted above, the Boards' discussions around development of the three bucket approach have to date focused around commercial loans. The Boards took this opportunity to consider the impairment model under development to both commercial and consumer loans. Based on the analysis provided by the staffs, the Boards felt that the model under development should work for both consumer and commercial loans.
The Boards also decided that the impairment model should not include a presumption (e.g., brightline) that meaningful credit deterioration has occurred based on predetermined factors such as number of days delinquent or reaching a particular credit risk rating.
The staffs also raised the issue of whether concentration risk of loans should be captured within an expected value measurement. Concentration risk of loans is a similar concept to a blockage factor for equity securities in that holding a significant amount of loans in a concentrated area, should the underlying collateral need to be seized and sold, the marketing of all properties in the same area would put downward pressure on the sales price. The Boards were of the view that similar to the blockage factor in a fair value measurement, the concentration risk would not impact the measurement of expected losses but is something that could warrant consideration for disclosure purposes.