Financial instruments – Impairment
The Boards held an educational session on the feedback the FASB received from constituents regarding the proposed impairment and interest recognition guidance. The Boards also discussed the way forward in the project by examining various models that could be pursued. This was educational session only, no decisions have been made.
The FASB staff noted that the comment letters urged the Boards to work together and to develop a converged impairment model for financial assets. Constituents also strongly supported development of a single model for both loans and securities.
The outreach showed a strong support for elimination of probability threshold for recognition of impairment. Nonetheless most constituents disagreed with limiting the information considered for estimation of credit loss to current events (i.e. no forecasting of future situation) and noted that such limitation would be inconsistent with pricing of the loans. The FASB noted that it received mixed feedback on timing of recognition of credit losses — some constituents supported upfront recognition of credit losses, others supported spreading the losses over the life of the instruments.
Finally, most constituents urged the FASB to simplify the guidance for accounting for purchased assets and retained guidance on income recognition. In general, constituents urged the Board to provide practical expedients and address operational issue in its model.
The FASB noted that the feedback on its proposal received on public roundtables was consistent with the summary of comment letters.
The FASB staff provided an overview of the four significant questions that need to be addressed by the Boards and permutation of impairment models that flow from answers to these questions:
- Information considered for credit loss estimates (no forecasts, forecasts for foreseeable future, forecasts for the remaining life of the instrument)
- Time horizon for the loss estimates (expected loss over the life of the instrument, expected loss over a shorter period)
- Timing of recognition of credit losses (upfront recognition, spread recognition, distinction between the good book and bad book), and,
- Presentation of yield on the face of the performance statement.
The Boards discussed these four issues and their permutations. They noted that presentation of yield would be probably the least controversial issue and can be addressed separately. Some Board members noted that the presentation of yield would be important as decoupling could potentially lead to different objective of amortised cost measurement than effective yield.
The Boards discussed the information that should be considered for loss estimates. The Boards concluded that in practice the options of considering foreseeable future or the life of the instrument would be similar as in practice estimate of losses over the life of the instrument would be based on forecasts over foreseeable future period (as expectations over one to two years out are fairly unreliable).
The Boards then considered the timing of recognition of credit losses and acknowledged that this would be their biggest hurdle in reaching a converged standard. The preliminary views of the Boards diverged on this issue. Each Board leaned to their respective approaches in their exposure drafts with the FASB members gravitating towards upfront recognition of credit losses, whereas the IASB members preferred spreading the losses over the life of the instrument. When defining the objective of the model, the FASB focused on the principal amount not collected, whereas the IASB focussed on the present value of the future cash flows. The FASB noted that the IASB model, despite proposed practical expedients would be very difficult to apply and noted the operation difficulties in distinguishing between the good book and bad book on a portfolio level.
The Boards noted that these issues should be deliberated jointly and asked the staff of both Boards to consider the issues as well as ways forward. The Boards will discuss the way forward as well as good book — bad book distinction at the November joint meeting. The IASB Chairman noted that the Boards should try to develop a common model for the impairment model or in case of divergence in opinions present alternatives that could be submitted to constituents for comments.