Financial instruments — Impairment
The Boards continued their discussions on development of the three bucket impairment model for financial instruments.
Direction of movements between impairment buckets
The first issue the boards discussed was whether financial assets should be permitted to move back to bucket one once they have been moved to buckets two or three.
Purchased credit-impaired assets
For purchased credit-impaired assets, the Boards have previously agreed that those assets would be initially classified in buckets two or three and would recognise interest based on the initial expectations of cash flows rather than contractual cash flows. The effective interest rate based on those initial cash flow estimates will be locked at initial recognition and not revised for subsequent changes in estimates, rather those changes in estimates will be recognised immediately in profit or loss.
The staff raised concerns with permitting purchased credit-impaired assets to be moved back to bucket one upon an improvement in the estimated cash flows. They questioned whether subsequent transfers back to buckets two or three would occur based on any deterioration or only if the assets have deteriorated back to the level at original acquisition.
One IASB member mentioned her request that specific disclosures be developed for purchased credit-impaired assets because of the difficulty in understanding these assets given their separate income recognition model. Both the IASB and FASB tentatively agreed (unanimous votes by both Boards) not to move purchased credit-impaired assets back to bucket one after their initial classifications in either buckets two or three.
Originated and purchased non-credit-impaired assets
Originated and purchased non-credit-impaired assets are initially classified in bucket one and moved to buckets two or three when there has been more than an insignificant deterioration in credit quality since initial recognition and the likelihood of default is such that it is at least reasonably possible that all or some of the contractual cash flows may not be collected.
The Boards discussed whether, and when, assets should be moved back in to bucket one after previously experiencing deterioration such that they had moved into either buckets two or three. The staff presented the Boards with the following alternatives:
- permitting movements back to bucket one with the upward transfer based on the same notion used for downward transfers
- permitting movements back to bucket one with the upward transfer based on different criteria than that used for downward transfers, or
- not permitting assets to move back to bucket one after previously deteriorating to bucket two or bucket three.
The FASB Chair clarified that the Boards were not discussing whether reversals of impairment provisions would be permitted as regardless of which alternative the Boards selected, there would be reversals of impairment provisions. For those assets in bucket two or three, entities would continue to update their estimates of expected losses when recognising full lifetime expected losses and any improvements would be recognised in the change of lifetime expected losses. The real issue the Boards were considering is whether the measurement attribute for impairment provisions should change for those assets where credit deterioration has occurred. She raised concerns over introducing additional complexity to the impairment model by requiring entities consider not only downward transfers but also upward transfer criteria. Another FASB member raised similar concerns citing the complexity of the three bucket impairment model and felt that including upward transfers would increase that complexity. However, another FASB member retorted that he had concerns that not permitting upward transfers may put additional pressures on the downward transfer threshold as entities may be more hesitant to make the initial downward transfer if they are never permitted to transfer back to bucket one.
The staff highlighted that outreach with the expert advisory panel noted that permitting movement back to bucket one and creating a symmetrical model would be preferable as it would align closer to how they manage credit whereas requiring assets that have recovered to be kept in buckets two or three would require separate tracking.
An IASB member noted that with the shift to an expected loss model, he felt that movements across buckets would be much more prevalent than they would using an incurred loss model. He also agreed with the concerns raised earlier that not permitting reclassifications back to bucket one would put strain on the decision to originally transfer an asset down to buckets two or three.
The Boards tentatively agreed (13 IASB/4 FASB members in favour) that originated and purchased non-credit-impaired assets would be moved back to bucket one when the downward transfer notion is no longer satisfied.
Application of the impairment model to trade receivables
Use of an ‘incurred loss’ model or an ‘expected loss’ model
The Boards discussed the potential impacts of a shift from an incurred loss model to an expected loss model for trade receivables. Under the incurred loss model today, the threshold recognising an impairment provision for IFRS is ‘objective evidence of impairment’ while under US GAAP the threshold is probable.
Certain constituents raised concerns with the use of an expected loss model for trade receivables. They noted that the project was attempting to address issues with loans and debt securities but the accounting for trade receivables was not considered ‘broken’. They also questioned whether the benefits of utilising an expected loss model would outweigh the associated costs and whether using expected losses would provide useful information given their short lived nature.
Some of the Board members questioned whether commercial entities truly understood the difference between an expected loss model and an incurred loss model. Some Board members felt that current practice for trade receivables was actually closer to an expected loss model than an incurred loss model although the staff’s noted that US GAAP has somewhat more latitude for recognising losses under an IBNR approach for current receivables. Some Board members wondered whether the Boards could clarify that a move to an expected loss model would not necessitate a change in current practice for provisioning of trade receivables.
The Boards tentatively decided that trade receivables with a significant financing element would be subject to an expected loss model. The Boards will further consider the application to trade receivables without a significant financing element during a future meeting.
Application of the expected loss model to trade receivables that have a significant financing component
The Boards then discussed whether to simply the three bucket impairment model for trade receivables that include a significant financing component. The staff presented the Boards with three alternatives to consider: 1) treating trade receivables with significant financing components similar to all other financial assets, 2) requiring entities to apply a simplified approach for trade receivables with a significant financing component, or 3) providing entities with a practical expedient that may be used as a simplified approach.
One IASB member began the discussions stating that receivables with significant financing components should work similar to a loan but acknowledged that commercial entities may not have the system capabilities needed to apply the three bucket impairment model. As a result, he supported providing the choice of applying a practical expedient. A FASB member had similar concerns and also expressed support for a practical expedient.
However, another IASB member mentioned that an entity providing significant financing is in the same business as a bank and therefore should be subject to similar accounting requirements, although he noted he wouldn’t object to the practical expedient approach. A FASB member had similar concerns noting that if entities are providing significant financing then it’s hard to imagine them not having credit systems in place. The staff did note that one advantage banking institutions have is they are already familiar with the 12-month expected loss estimate through regulatory requirements whereas commercial entities not subject to similar regulations would not.
The Boards tentatively decided to provide a practical expedient for trade receivables with significant financing components such that entities would be permitted a policy election to either fully apply the three bucket impairment model or to follow a simplified approach where the allowance measurement objective is ‘lifetime expected losses’ throughout an asset’s life and they would not be required to track credit deterioration for disclosure purposes.
Application of the expected loss model to trade receivables without a significant financing component
The IASB and FASB discussed the application of the expected loss model to trade receivables without a significant financing component separately due to the differences in their respective requirements related to initial measurement (IFRS requiring initial recognition at fair value where US GAAP recognises at transaction price). Specifically, the Boards considered whether trade receivables should be subject to a credit deterioration model given the relative short term nature of these financial assets and the fact that entities with trade receivables would typically not have systems designed to track credit deterioration of trade receivables.
The IASB considered three alternatives for trade receivables. The first alternative would provide an exception from the three bucket impairment model by not requiring an allowance balance to be established for initial estimates of expected losses and requiring that changes in lifetime expected losses always be recognised. The second alternative would measure the trade receivable at the invoice amount on initial recognition and initially allocate those receivables directly to bucket two or three and recognise lifetime expected losses throughout the asset’s life. The third alternative would require the bucket one measurement to be recognised on initial recognition (consistent with all other financial assets). The staff recommended the second alternative because, even though it creates a separate model for trade receivables, it addresses the operational concerns of corporates.
The IASB tentatively decided to create a separate model for trade receivables without a significant financing component so that the receivable is initially measured at the invoice amount and initially classified in to buckets two or three where lifetime losses would be reserved.
The FASB also discussed application of the three bucket impairment model to trade receivables without a significant financing component. The FASB considered two alternatives for trade receivables. The first alternative would require entities to track credit deterioration so the categorisation into buckets can be appropriately disclosed. The second alternative would not require tracking of credit deterioration but instead have a measurement objective of lifetime expected losses throughout the life of the trade receivable.
Similar to the decision reached by the IASB, the FASB tentatively decided that for trade receivables without a significant financing component entities would not be required to track credit deterioration for disclosure purposes and lifetime expected losses would be the measurement objective for impairment.