The Financial Accounting Standards Board (FASB) recently released a proposal that would change the way financial institutions set aside funds to cover losses on loans, debt securities and other assets. Under current accounting rules, the allowance for loan and lease losses (ALLL) is based on incurred losses; the new model, if adopted, would require the allowance to be established for losses expected over the life of the loan based on current and future economic conditions, historical losses, and other factors. The change was prompted by the global financial crisis, when stakeholders were blindsided by the tremendous credit risk that had built up in many institutions’ loan portfolios.
Financial institutions follow generally accepted accounting principles (GAAP) when reporting financial information. Under GAAP, funding for the ALLL is determined by what an institution thinks it will lose on loans based on events that have already occurred; this method is referred to as the “incurred loss” method. There are a number of problems with this method, the most significant of which is that it is not forward-looking. Credit losses aren’t recognized until they are probable or have already occurred, thus making it difficult for investors to be forewarned of imbedded losses, as in the most recent financial crisis. Many bankers knew that their portfolios of subprime loans, for example, were in trouble long before homeowners started defaulting, but their financial statements did not (and could not, under GAAP) communicate that to investors.
The push to revamp accounting for credit losses began in 2009 when FASB and the International Accounting Standards Board (IASB) launched a joint project to improve loss accounting and mitigate differences in U.S. (GAAP) and international (IFRS, or international financial reporting standards) accounting systems. By the summer of 2012, the project had broken down as the two accounting bodies could not agree on some significant matters. FASB’s concerns were that the agencies’ joint proposed impairment method was too complicated and only allowed for a one-year projection period. Bankers had also argued that the joint proposal didn’t take into account the diverse nature of banking institutions in the U.S.
Under FASB’s new Current Expected Credit Loss (CECL) model, unveiled in December, nonaccrual loans would be treated as before. For unimpaired loans and other credit instruments, the institution would estimate expected credit losses at every reporting period. That estimate would capture all contractual cash flows that the institution does not expect to collect and would be based on “past events, current conditions, and reasonable and supportable forecasts about the future.” Importantly, the estimated losses would not be limited to those expected over a specific period of time, leaving most observers to conclude that estimated losses should be based on the life of the credit instrument.
Reaction to the FASB proposal has been mixed. On the positive side, an accounting expert at the American Bankers Association says the CECL model would be “operationally simpler as well as making the ALLL balance easier to understand and to explain to investors and management.” The model would allow ALLL balances to increase during periods of economic growth, if information suggests banks are taking on greater credit risk or future economic conditions are expected to deteriorate. This practice has historically been discouraged by the Securities and Exchange Commission, which is concerned about earnings smoothing.
A recent report by the Government Accountability Office (GAO) highlighting the work of the U.S. Treasury’s Financial Stability Working Group on Loss Provisioning is also supportive. The working group asserts that earlier recognition of potential loan losses could have lessened the impact of the financial crisis since banks ultimately had to recognize their credit loss exposures pro-cyclically through a sudden series of provisions to the loan loss reserve, thus depleting their earnings and regulatory capital.
Bankers do have some concerns about the proposal though. The biggest objection is that FASB has not specified a time period over which losses are to be estimated, leaving most to interpret the time period to be the life of a loan or debt security. Most observers objected to an arbitrary one-year time horizon that the IASB has floated, but they argue that the FASB proposal goes to the other extreme. A “life of the loan” loss projection would require information and forecasts that most bankers currently do not have the expertise to generate. Estimating losses on debt securities that are not government-guaranteed, like municipal bonds, would be difficult. Still others worry that the CECL model moves bank accounting more toward a market-value framework. It is also unclear how—if at all—the new model would affect capital requirements and the regulatory treatment of both the ALLL and bank capital.
The FASB proposal is out for comment until May 31. FASB chair Leslie Seidman has said that a final standard will likely be in place by early 2015.