Abstract
The Current Expected Credit Loss (CECL) Financial Accounting Standards Board (FASB) standard that goes into effect for major banks in 2020 contains a serious conceptual error. Using the contractual rate rather than the hurdle rate (the competitive rate on a loan for which there is no expected loss) as the rate to discount expected cash collections gives rise to accounting losses where no economic losses exist. This can have a profound effect on required capital and hence lending, especially in economically depressed episodes.
A new model for estimating and reporting credit losses came into effect beginning in 2020. 1 During the financial crisis, banks were following the so-called “incurred loss” methodology for credit losses, under which a bank does not recognize losses in financial statements until it is probable (based on available information) that loans are impaired and the amount of loss can be reasonably estimated. Throughout the crisis, however, banking agencies complained about how slowly banks were provisioning for losses on loans. As a result, in April 2009, the G20 and the Financial Stability Board (FSB) recommended that the Financial Accounting Standards Board (FASB) and the International Accounting Standards Board (IASB) reconsider how banks account for losses. 2 The purpose of this article is to show that a conceptual error in the computation of the expected loss gives rise to accounting losses upon origination where no economic losses exist.
The Current Expected Credit Loss (CECL) model allegedly improves the decision usefulness of the reporting of credit losses by removing constraints to the timely recognition of credit losses and requiring reporting entities to consider a broader information set. In a significant change from the previous impairment model, CECL requires that estimates of credit losses be based on reasonable and supportable forecasts. Under the CECL model, banks must provide for all losses expected over the entire life of the loan when they first originate the loan. For example, if a bank projects the loss rate on 5-year home equity loans to be 2% per year, it will book an immediate loss equal to 10% of the loan amount when it originates the loan. For each subsequent period, the bank would update its projections of remaining lifetime loan losses based on fresh information about the performance of the loan and changes in economic forecasts, and adjust its loan loss provision accordingly by either increasing or decreasing it. 3
There is general agreement that lending standards deteriorated in the years preceding the crisis, and that loans originated in those years performed poorly. Consequently, a requirement that banks take losses based on a more forward-looking perspective would allegedly contribute to a more salutary transparency: Financial statements of banks would become less opaque and would duly warn investors about impending losses. Furthermore, increasing provisions during the boom periods would supposedly incentivize against originating bad loans and hence better equip banks to weather a future recession.
The American Bankers Association (2019) has called CECL the “most sweeping change to bank accounting ever.” Indeed, this change to the accounting rules has big implications for the way institutions operate and the amount of credit they provide. It most likely will not stop there.
As the availability and cost of credit are critical to the economy’s performance, CECL will likely also have an impact on the business cycle.
The Conceptual Error in Estimating the Loss
ASC 326 requires a reporting entity to determine the allowance for credit losses for an instrument based on the amortized cost of the financial asset, including premiums, discounts, deferred origination fees or costs, foreign exchange adjustments, and fair value hedge accounting adjustments.
The computation of the expected credit loss, as stipulated in the standard, creates accounting losses where no economic losses exist. I illustrate this by analyzing how the standard prescribes the estimation of the credit loss under the discounted cash flow method. 4
Discounted Cash Flow Method
When a reporting entity measures the allowance for credit losses using a discounted cash flow approach, the allowance will reflect the difference between the amortized cost of the financial asset and the present value of the expected cash flows of the financial asset. Under this methodology, the standard prescribes that the original effective interest rate of the instrument be used as the discount rate. This method is described in ASC 326-20-30-4 as follows: If an entity estimates expected credit losses using methods that project future principal and interest cash flows (that is, a discounted cash flow method), the entity shall discount expected cash flows at the financial asset’s effective interest rate. When a discounted cash flow method is applied, the allowance for credit losses shall reflect the difference between the amortized cost basis and the present value of the expected cash flows. (p. 109)
The effective interest rate is defined in the standard as “the rate of return implicit in the financial asset, that is, the contractual interest rate adjusted for any net deferred fees or costs, premium, or discount existing at the origination or acquisition of the financial asset.” 5
Simply articulated, the use of the “effective interest rate,” as defined, ignores that lenders would rationally increase interest rates to compensate for whatever default risk and consequent nonpayment of principal and/or interest they anticipate over the lifetime of the loan. Hence, they would expect not to incur any economic loss upon loan origination. Furthermore, if over time they forecast a heightened default risk, they would increase interest rates again (in many cases of loans with no fixed interest rate) to make themselves whole. The purpose of those rational interest rate adjustments is to prevent economic loss.
Yet, under the FASB’s guidance, requiring the use of the “effective interest rate” (the discount rate that is the same as the stipulated loan interest rate) rather than the conceptually correct discount rate (the hurdle rate that is otherwise charged on loans where no loss is expected) results in recording accounting losses when the lender does not expect any economic loss due to having taken rational steps to avoid economic loss.
If, however, a lender, imprudently or purposefully for other business reasons (such as when viewing the loan as part of a bundle of business deals), fails to charge an interest rate that compensates for expected losses, then a credit loss that is equivalent to the economic loss would be reported if the conceptually correct hurdle rate is used. Use of the contractual rate as stipulated in the standard would magnify the accounting loss in excess of the economic loss.
Below I illustrate this with a hypothetical example.
Illustration of How Accounting Losses Are Created With Zero Economic Losses
Suppose a lender extended a loan of US$1,000 for a period of 3 years, at the end of which the borrower is supposed to repay exactly US$1,000. However, the lender expects the borrower, while able to make all interest payments, to partially default on the lump sum principal payment, such as by paying only US$900. To cover himself against this shortfall, he charges the borrower a rate that is higher than the rate of 10% he will have charged in the absence of default risk. The lender fully covers himself by charging 13.0213%—recall that the borrower is expected to make full payments of interest rates, which would be based on the contractual rate. As the loan amount he extended was US$1,000 and the face value is also US$1,000, then under the FASB’s definition of effective interest rate, the discount rate would be the contractual rate of 13.0213% applied to the expected cash collections, that is, the US$900 expected cash collection of lump sum at the end of 3 years, implying a shortfall of US$100, plus the full interest payments. 6 This discounting results in the present value of US$930.73, yielding a credit loss to be charged to the income statement of US$69.27. Note, however, in actuality the lender would not suffer any loss: Using the hurdle rate of 10% to discount the expected payment of US$900 and the interest payments (calculated based on the 13.0213% contractual rate as applied to the face value of US$1,000) that cover him for his expected loss of US$100, he obtains a present value of US$1,000, which implies a credit loss of zero!
The hurdle rate, the appropriate interest rate of 10% chosen in this example, would typically be the competitive rate charged on risk-free loans. Thus, in my simple example, discounting by using the hurdle rate of 10%, the lender obtains US$1,000: The added 3.0213% spread, compensating for the expected cash shortfall of US$100, makes him whole; no economic loss would be incurred, and, appropriately, no credit loss would be reported. The reported credit loss of zero would equal the economic loss of zero. Consequently, the 10% hurdle rate—rationally used in selecting investment projects—would be the appropriate discount rate to apply to expected cash collections.
Suppose now the lender did not adjust the interest rate and set the contractual rate at the hurdle rate of 10%, even while expecting a shortfall of US$100 on the principal. The present value using the hurdle rate of 10% (which in this case is also the contractual rate) would be US$924.87. The reported credit loss would be US$75.13, which also equals the economic loss (the present value of the shortfall of US$100). With a contractual rate of 11%—still short of the 13.0213% required to make the lender whole—the present value using the contractual rate would be US$926.88; this yields a credit loss of US$73.12, in excess of the true economic loss of US$50.26 which would be reported as credit loss when the correct hurdle rate of 10% is used as the discount rate.
When reassessing default risk subsequent to the loan’s origination, the hurdle rate implicit in new loan originations (during the later periods) with respect to which no losses are expected would be the appropriate rate to use upon changed estimates of default risk.
Note that this hurdle rate is observable. Once the lender has estimated expected cash collections—which will be necessarily available as the lender has to make the estimate under CECL in any case—it is a simple matter to ascertain the hurdle rate as that charged on loans with respect to which no losses are expected. Hence, this rate is both verifiable and auditable, conditional on expected collections.
It is also noteworthy that the interest revenue recorded would not be compromised when no expected credit loss is recorded. It is essentially 10% on the book value of the loan (as amortized by receipts on account of principal over the years; see the appendix for the journal entries).
Further Clarification
The expected credit loss of 69.27 under CECL (100/[1 + 13.0213%] 3 ) is overstated. The alleged but incorrect expected credit loss should be 75.13 (=100/[1 + 10%] 3 ). In other words, the expected shortfall would be discounted using the hurdle rate. The underestimation of the discounted expected cash collections as reflected in the face value of UA$1,000 minus the present value of the shortfall (i.e., 1,000−75.13) is solely due to the failure to recognize the present value of the additional interest charged to make the lender whole, which is exactly 75.13 (i.e., the mismatch between revenue and expense). To see this, consider the following equations:
Equation 1 shows the net present value discounting expected cash collections using the hurdle rate, and Equation 2 shows the present value using the hurdle rate and the face value of the loan (assuming zero default risk) and interest payments that do not need to be increased to compensate for the (assumed to not exist) default risk.
The difference between Equations 1 and 2 can be expressed as follows:
The left-hand side is the expected credit gain, which is the same as the expected credit loss on the right-hand side (so that the lender breaks even). CECL, however, does not allow a lender to recognize the left-hand side gain at the loan origination date. In summary, the correct book value of the loan should be US$1,000, whereas the recorded book value of the loan under CECL is US$930.73. The distortion of US$69.27 is driven by (a) an understatement of the (incorrect) expected credit loss by US$5.86 (US$75.13−US$69.27) and (b) an understatement of the expected credit gain by US$75.13. If we correct these two understatements, the amortized amount of the loan net of the allowance for credit loss (which is zero) would be US$1,000. Equation 3 illustrates how the expected credit loss is exactly offset by the additional interest revenue. 7
Concluding Remarks
CECL was designed to address a potential procyclicality of the incurred loss model. The argument that estimating loans’ lifetime losses and reflecting these in financial statements as they change over time is sound in principle. However, the prescription that expected cash collections, which prudent lenders have adjusted to avoid economic loss, should be discounted using the “effective interest rate,” defined as the contractual rate of return, undermines the transparency objective. Specifically, it gives rise to apparent accounting losses upon origination of loans in the absence of economic losses. This is the opposite of transparency and can have a profound effect on required capital and hence lending, especially in economically depressed episodes.
The cure for this is to use as the discount rate the hurdle rate, which in equilibrium would be the competitive rate of interest charged on loans for which there is no expected loss.
Footnotes
Appendix
Journal Entries for the Example.
| Beginning of Year 1 | Debit | Credit |
| Loan | US$1,000 | |
| Cash | US$1,000 | |
| End of Year 1 | ||
| Cash | US$130.21 | |
| Interest revenue | US$100 | |
| Loan | US$30.21 | |
| Recording receipt of interest and allocation to revenue and loan principal | ||
| End of Year 2 | ||
| Cash | US$130.21 | |
| Interest revenue | US$96.98 | |
| Loan | US$33.23 | |
| Recording receipt of interest, revenue computed at 10% of the loan beginning balance of Year 2, and further amortization of the loan | ||
| End of Year 3 | ||
| Cash | $130.21 | |
| Interest revenue | $93.66 | |
| Loan | $36.55 | |
| Recording receipt of interest, revenue computed at 10% of the loan beginning balance of Year 3, and final amortization of the loan | ||
| Cash | $900 | |
| Loan | $900 | |
Note. Loan balance at the end of Year 3: US$1,000−US$30.21−US$33.23−US$36.56−US$900 = $0. Total interest revenue: US$100 + US$96.979 + US$93.656 = US$290.635. The difference of US$9.365 between US$300 (3 × US$100) and US$290.63, the actual interest revenue recorded, is explained by the fact that part of the principal is paid earlier than the end of the third year. This difference reflects the interest gained on the reinvestment of payments on account of principal. Specifically, US$30.21 is received at the end of Year 1 and it potentially earns an interest of 20% over the next 2 years amounting to US$6.042; adding to this the potential interest earned on the US$33.3 of US$3.323 (10% of US$33.3), we obtain the difference of US$9.365.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Data Availability
Data are available from the public sources cited in the text.