The Current Expected Credit Loss (CECL), or loan-loss rule, is expected to go into effect in 2020 with a serious conceptual flaw: discounting expected cash collections by the contractual rate, as prescribed in the Financial Accounting Standards Board (FASB) guidance, which would result in accounting losses where no economic losses exist. Fortunately, there is a “cure” to CECL’s deficiency: use as discount rate the internal rate of return (IRR).
But first, let’s examine the new FASB guidance, which requires entities to project lifetime losses—albeit only taking into consideration information that is reasonably available—upon the origination of loans. Projected cash collections are then to be discounted by the contractual rate. If the resulting present value is below the amount of loan extended, the difference would be reported as a loss, decreasing regulatory capital.
The guidance fails to consider lenders’ behavior. Once a prudent lender has projected cash collections as required by the guidance, he/she would either refuse to extend the loan if projections suggest a loss, or charge a higher contractual interest rate, if and when feasible, to compensate for the expected loss. Discounting cash collections by such higher contractual rates would result in accounting losses, decreases in regulatory capital, and consequent decisions to suppress lending. Consider this against the recent market drops and the expectation of an economic slowdown following the possible failure of a trade deal and other uncertainties—it is almost certain to exacerbate the backsliding of the U.S. economy.
In both bullish and bearish markets, the importance of loans cannot be overstated: Many billions of mortgages alone (not considering other types of loans) are underwritten every quarter (for example, $457 billion in mortgages originated in the country in the third quarter of 2018). In light of this, the potential for depressed lending resulting from the CECL rule, especially during downturns when the reported accounting loss is likely to be particularly large, can spell disastrous economic consequences.
The cure for CECL is to discount expected cash collections using the IRR, the rate that yields as present value the amount of the loan originally extended without incurring an economic loss. For the prudent lender who fixes the terms of the loan to avoid economic losses, this discounting will result in zero accounting losses, properly reflecting zero economic losses. If an imprudent lender extends a loan and knowingly incurs an economic loss given his/her projected collections, properly discounting using the IRR will result in accounting losses that are equivalent to the economic losses. As a result, the loan would be properly judged to be imprudent, and the loan-extending individual would be justifiably blamed.
One way to more generally implement the cure is to use discounted cash flow (DCF) values to quantify loans. FASB Member Hal Schroeder on May 7, 2019 said CECL wasn’t his first choice to better align accounting rules with the credit risk banks take as part of their routine lending. He would have preferred a fair value approach. Unfortunately, the “fair value” concept Mr. Schroeder probably has in mind is the FASB-prescribed exit price—the consideration expected to be received if the assets were to be hypothetically sold at the financial report date. Banks and other lenders would be right to reject this concept. During downturns, such as during the most recent financial crisis, exit prices would be pummeled by market uncertainty and illiquidity, severely impacting capital and recovering only when markets stabilize and liquidity is restored. Similarly, credit risk would be confounded by extraneous market forces, as would loan loss provisions. Volatility would unduly surge, magnifying business uncertainty.
Such higher volatility is suggested by research I co-authored with Alex Dontoh, Fayez A. Elayan, and Tavy Ronen. The research reveals declines in fair values (exit prices) of financial instruments in U.S. and foreign banks during the financial crisis (from the third quarter of 2007 to the third quarter of 2009) of $507 billion. These fair values largely regained the losses when the markets recovered; the mean decrease in fair value during 2008 of $1.421billion contrasts with the mean increases in fair value over the years 2009 through 2012 of $1217 billion, $538 million, $386 million, and $825 million respectively. During the crisis, relative to what would normally be expected, equity prices of financial institutions announcing the fair value declines plunged by a whopping 8.22% on average over the three days surrounding the date of the announcement. Similarly, the premium on credit default swaps, relative to what one would normally expect, increased by a very substantial 5.09%. Reactions to whatever fair value declines were announced after the markets recovered were far more muted.
The appropriate concept of value-to-use in determining loan impairment would be the discounted cash flow (DCF) value—the present value of expected cash collections discounted at the IRR. Applied to originated loans, this is equivalent to the present value suggested in the above “cure.” Changes in DCF values would reflect economic losses or gains properly shown as accounting losses or gains. Prudent lenders would show no accounting loss upon origination, consistent with the absence of economic loss. Imprudent lenders would report accounting losses that are equivalent to economic losses. Importantly, once the methodology of estimating cash collections has been established, a decrease in default risk would automatically give rise to increases in DCF values and corresponding gains that increase capital.
Under the FASB’s current guidance, lenders will find it difficult to convince overly conservative auditors that expected losses decreased sufficiently to merit reporting gains. Whereas, my “cure” for CECL, which reports DCF values of loans and other financial instruments in balance sheets and reflects changes therein as gains or losses, would enhance transparency and clarity in comparison with the reporting of fair values in footnotes, and thus help mitigate auditors’ conservative bias.
An added advantage of using the IRR to discount expected cash collections in determining DCF values, beyond the proper reflection of economic losses or gains as accounting losses or gains, is its consistency with how managers evaluate projects and make investment decisions. Specifically, in normal circumstances the IRR would be equivalent to the cost of capital used as hurdle rate to evaluate the desirability of investing in projects—including loans. That is, the cost of capital is used as the rate to discount cash flows expected on alternative projects and the resulting present value is then compared with the required outlay to decide on whether to invest.
Using DCF values for loans and recording changes therein as gains and losses, which essentially amounts to a corrected CECL, yields accounting gains and losses that are equivalent to economic gains and losses. Upon origination, prudent lenders who fix the loan terms in order to avoid economic losses will not show accounting losses. Transparency and clarity would be enhanced in that the DCF values appear on the balance sheets rather than in footnotes that investors may ignore. Furthermore, DCF values are superior to exit prices that are used to quantify fair values under the current FASB guidance. The latter are more volatile and are especially vulnerable to market uncertainty and illiquidity during economic downturns.
In sum, before implementation, CECL needs to be corrected to prevent the reporting of accounting losses in the absence of economic losses. This is not to suggest that a cured CECL, fixed by applying a proper discount rate, would be easy to implement. CECL is complex and projections are difficult to make. And the lenders’ mode of making decisions on how to project future cash collections, how to decide whether to make loans, and how to determine loan terms (interest rate, collateral, etc.) will be continuously changing as CECL is implemented.
While the ultimate effects of CECL are subject to a great deal of uncertainty, the implementation will depend on how different banks will apply the guidance, how projections methodologies will vary across the banks, how such variation will affect comparability and how investors will react to the different ways banks approach the implementation of CECL. Let there be a period of testing in which the banks submit, but not yet report publicly, the CECL simulated results to regulators who can then assess the potential impact on capital and lending. It is far better to be aware of potential consequences before they are thrust upon markets.