Fifteen of the largest banks in the U.S. have written a letter to the chairmen of the Financial Accounting Standards Board and the International Accounting Standards Board encouraging them to resolve their differences over the standards for credit losses in their financial instruments convergence project.
After initially agreeing on the proposed changes they wanted to make in the standards, FASB and the IASB parted ways during a contentious meeting last year (see FASB Splits with IASB on Impairment Standards). The two boards have since issued differing exposure drafts on their proposed changes in the accounting standards for loan loss provisioning and expected credit losses for loan impairment, but said they hope to come together after receiving feedback from their stakeholders (see FASB Proposes More Timely Recognition of Expected Credit Losses and IASB Diverges from FASB in Revised Loan Loss Proposals).
The banks weighed in with a joint letter last Friday addressed to FASB chair Leslie Seidman and IASB chairman Hans Hoogervorst. “In an increasingly global financial marketplace, market participants, users and prudential regulators all recognize the need for a common set of high quality accounting standards related to credit impairment,” the banks wrote. While we acknowledge the difficulty inherent in reconciling disparate points of view, we strongly encourage the boards to achieve convergence on what we believe is the most important [memorandum of understanding] project.
Although we continue to support an event-driven accounting framework for recognizing credit losses consistent with the proposal previously provided by members of the U.S. banking industry, we acknowledge the need for a balanced approach that will broadly appeal to numerous constituents. While a converged standard may not necessarily lead to fully comparable results in practice, the differences between the models proposed by the boards are far too great and will generate vastly different results. Ultimately, we believe compromise will be necessary by both boards in order to achieve a converged credit impairment standard. We strongly encourage the boards to renew their cooperation on this critically important matter.”
The banks, which include Bank of America, Capital One, Citigroup, JPMorgan Chase, Morgan Stanley, PNC Financial, SunTrust and Wells Fargo, noted that there are several differences between the models proposed by FASB and the IASB and said they have several concerns with each of the proposed models. However, they said they believe a single fundamental change would help facilitate a compromise between the two boards while simultaneously addressing many of the core concerns with the proposed models.
“Rather than measuring expected losses over the next 12 months or over the remaining contractual life, we recommend that the boards amend the expected loss measurement period to the greater of 12 months or the period that is reliably estimable and predictable,” they wrote.
They said such a compromise would offer the following benefits: The conceptual basis for the recognition of credit losses would be maintained as credit quality would be more fully considered in the determination of the estimate of credit losses. Credit quality could be evaluated with commonly used credit quality indicators and portfolio and product characteristics, combined with appropriate loss estimation periods that contemplate expectations regarding current and future economic conditions. “As a result, performing assets would not require immediate recognition of a less reliable estimate of expected lifetime credit loss content,” said the comment letter. “The period that is reliably estimable and predictable would capture a substantial portion of expected credit losses for performing assets and all of the expected credit losses for non-performing assets as loss content tends to materialize earlier rather than later in the life of a financial asset. Moreover, during stressed economic environments, allowance levels would not be adversely impacted as expected credit losses should emerge more quickly and would already be reflected in the allowance, supplemented by oversight provided by internal risk management and prudential regulators.”
The banks also predicted that reliability of expected credit loss estimates would be improved, particularly for “long tenor” and evergreen assets.
“Credit risk managers would be better able to validate and back test estimates to their satisfaction and to the satisfaction of banking regulators and auditors,” they predicted. “Loss forecasting models must satisfy rigorous internal and regulatory modeling standards, including demonstrated accuracy in backtesting to historical results. Accordingly, reliable and predictable credit loss estimates would be measured in a well-controlled environment with a reasonable level of confidence.”
They also said that the compromise would allow existing loss estimation techniques to be leveraged. “Many existing loss estimation methodologies are not suitable for long-term loss estimates and would not satisfy prudent model risk and validation requirements,” said the banks. “Limiting the loss estimation period to the period of time which is reliably estimable and predictable would retain the ability of financial institutions to utilize many existing methodologies and retain the ability to capture all or a substantial portion of the expected loss content. Moreover, this would allow smaller, or less sophisticated, institutions to develop and implement loss estimation techniques that meet the standard.”
The banks warned that measurement of losses over the remaining contractual life could adversely impact lending and directly inhibit long-term investment, adding that this factor will be critically evaluated by the Financial Stability Board in their assessment of the implications of accounting standard setting.
“We believe this compromise will resolve this potential unintended consequence,” they wrote. “The reliably estimable period will allay concerns that limiting measurement of expected credit losses to just 12 months would perpetuate the ‘too-little-too-late’ concerns associated with the incurred loss model.”
An explicit transfer principle, as proposed by the IASB, would not be necessary, they added, as the period that is reliably estimable and predictable would consider the credit quality of financial assets at the reporting date, along with reasonable and supportable assumptions regarding current and forecasted events and conditions.
They said the ability to scrutinize the judgments of management through transparent disclosure of the assumptions used to measure expected losses, including the period that is reliably estimable and predictable, by asset class, would allay any “concerns related to earnings management and ultimately promote and improve comparability and consistency among preparers.”
The banks urged the standard-setting board to unite around a common set of standards for credit impairment.
“All parties agree that convergence on credit impairment is critically important,” they wrote. “Accordingly, we encourage the boards to renew their cooperation and consider incorporating our recommendation into their respective proposals. We acknowledge that no accounting model will completely resolve procyclical reserving concerns and loss estimates and estimation periods may vary, by product and across organizations. However, we believe that consistent practice will develop quickly through robust disclosure, coupled with the existence of proper risk governance and regulatory oversight. We believe our recommendation has a solid foundation in existing credit risk management practices in our industry, will more reliably reflect credit losses expected in the portfolio, better align recognition of credit losses to those periods where credit losses are expected in the portfolio, and provide more decision useful information about expected credit losses for investors and other users.”