Chapter 4 — Measurement of Expected Credit Losses

4.4 Measurement Methods and Techniques

ASC 326-20

30-3 The allowance for credit losses may be determined using various methods. For example, an entity may use discounted cash flow methods, loss-rate methods, roll-rate methods, probability-of-default methods, or methods that utilize an aging schedule. An entity is not required to utilize a discounted cash flow method to estimate expected credit losses. Similarly, an entity is not required to reconcile the estimation technique it uses with a discounted cash flow method.

30-4 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. If the financial asset’s contractual interest rate varies based on subsequent changes in an independent factor, such as an index or rate, for example, the prime rate, the London Interbank Offered Rate (LIBOR), or the U.S. Treasury bill weekly average, that financial asset’s effective interest rate (used to discount expected cash flows as described in this paragraph) shall be calculated based on the factor as it changes over the life of the financial asset. An entity is not required to project changes in the factor for purposes of estimating expected future cash flows. If the entity projects changes in the factor for the purposes of estimating expected future cash flows, it shall use the same projections in determining the effective interest rate used to discount those cash flows. In addition, if the entity projects changes in the factor for the purposes of estimating expected future cash flows, it shall adjust the effective interest rate used to discount expected cash flows to consider the timing (and changes in the timing) of expected cash flows resulting from expected prepayments in accordance with paragraph 326-20-30-4A. Subtopic 310-20 on receivables — nonrefundable fees and other costs provides guidance on the calculation of interest income for variable rate instruments.

Pending Content (Transition Guidance: ASC 326-10-65-5)

30-4 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. If a financial asset is modified and is considered to be a continuation of the original asset, an entity shall use the post-modification contractual interest rate to derive the effective interest rate when using a discounted cash flow method. See paragraph 815-25-35-10 for guidance on the treatment of a basis adjustment related to an existing portfolio layer method hedge. If the financial asset’s contractual interest rate varies based on subsequent changes in an independent factor, such as an index or rate, for example, the prime rate, the London Interbank Offered Rate (LIBOR), or the U.S. Treasury bill weekly average, that financial asset’s effective interest rate (used to discount expected cash flows as described in this paragraph) shall be calculated based on the factor as it changes over the life of the financial asset. An entity is not required to project changes in the factor for purposes of estimating expected future cash flows. If the entity projects changes in the factor for the purposes of estimating expected future cash flows, it shall use the same projections in determining the effective interest rate used to discount those cash flows. In addition, if the entity projects changes in the factor for the purposes of estimating expected future cash flows, it shall adjust the effective interest rate used to discount expected cash flows to consider the timing (and changes in the timing) of expected cash flows resulting from expected prepayments in accordance with paragraph 326-20-30-4A. Subtopic 310-20 on receivables — nonrefundable fees and other costs provides guidance on the calculation of interest income for variable rate instruments.

30-4A As an accounting policy election for each class of financing receivable or major security type, an entity may adjust the effective interest rate used to discount expected cash flows to consider the timing (and changes in timing) of expected cash flows resulting from expected prepayments. However, if the asset is restructured in a troubled debt restructuring, the effective interest rate used to discount expected cash flows shall not be adjusted because of subsequent changes in expected timing of cash flows.

Pending Content (Transition Guidance: ASC 326-10-65-5)

30-5A An entity may make an accounting policy election, at the class of financing receivable or the major security-type level, not to measure an allowance for credit losses for accrued interest receivables if the entity writes off the uncollectible accrued interest receivable balance in a timely manner. This accounting policy election should be considered separately from the accounting policy election in paragraph 326-20-35-8A. An entity may not analogize this guidance to components of amortized cost basis other than accrued interest.

The objectives of ASU 2016-13 are to eliminate the incurred loss approach and the probability threshold for recognition of credit losses. In accordance with ASC 326-20-30-3 (see above), various methods can be used to meet these objectives. In paragraph BC50 of ASU 2016-13, the FASB indicates that its reasoning behind allowing entities to use various methods in estimating credit losses is that “entities manage credit risk differently and should have flexibility to best report their expectations.” The Board also acknowledges that “different methods may result in a range of acceptable outcomes” and do not inherently cause a particular estimate to be unreasonable.

Accordingly, an entity can select from a number of measurement approaches to determine the allowance for expected credit losses. Some approaches project future principal and interest cash flows (i.e., a DCF method), while others project only future principal losses. If an entity chooses to estimate credit losses by using a method other than a DCF method, it has the option of estimating such losses on the asset’s amortized cost basis in the aggregate or by separately measuring the components of the amortized cost basis (e.g., premiums and discounts), as described in ASC 326-20-30-5.

While an entity is not required to change its current method(s) of estimating credit losses upon adopting ASU 2016-13, it should consider, for each type of financial asset, whether it is more appropriate to use a different method to meet the ASU’s objectives. For example, the entity may determine that certain methods are not suitable for measuring expected credit losses on prepayable financial assets (see Section 4.4.5 for discussion of how estimated prepayments may affect the estimate of expected credit losses if a method other than a DCF approach is used). In addition, as required by ASC 326-20-30-10 and discussed in paragraph BC63 of ASU 2016-13, regardless of the method used, an entity’s estimate of expected credit losses should take into account “the expected risk of loss, even if that risk is remote.”

To faithfully estimate the collectibility of financial assets within the scope of ASC 326-20, an entity should use judgment in developing estimation techniques and apply those techniques consistently over time. ASC 326-20-55-7 emphasizes that an entity should use methods that are “practical and relevant” given the specific facts and circumstances and that “[t]he method(s) used to estimate expected credit losses may vary on the basis of the type of financial asset, the entity’s ability to predict the timing of cash flows, and the information available to the entity.”

Regardless of whether an entity changes its current method(s) of estimating credit losses upon adoption of ASU 2016-13, the method(s) used must be disclosed by portfolio segment and major security type in accordance with ASC 326-20-50-11. In changing the estimation technique or assumptions used to calculate expected credit losses after initially adopting ASU 2016-13, an entity should consider whether the change reflects a change in accounting principle3 or a change in accounting estimate4 in accordance with ASC 250. ASC 250 also discusses a change in accounting estimate effected by a change in accounting principle.5 The table below discusses (1) when it is appropriate to effect a change in accounting principle or a change in accounting estimate and (2) how such changes should be accounted for in an entity’s financial statements.

Type of Change	When a Change Is Appropriate	Accounting for the Change
Change in accounting principle	May be made upon the initial adoption of a newly issued Codification update. If it occurs under any other circumstances, it must be justified on the basis that the alternative accounting principle is preferable (see ASC 250-10-45-2). SEC registrants are required to file a preferability letter to effect a change in accounting principle.	Reported “through retrospective application of the new accounting principle to all prior periods, unless it is impracticable to do so” (see ASC 250-10-45-5).
Change in accounting estimate	Generally results from “the continuing process of obtaining additional information and revising estimates” related to the present status and expected future benefits and obligations associated with assets and liabilities (see ASC 250-10-45-18 and the definition of a change in accounting estimate). The objective of measuring an asset or liability in other Codification topics will affect the assessment of when a change in accounting estimate is justified given the particular facts and circumstances.	Should not be accounted for by restating or retrospectively adjusting amounts reported in prior-period financial statements. Rather, such a change should “be accounted for in the period of change if the change affects that period only or in the period of change and future periods if the change affects both” (see ASC 250-10-45-17).
Change in accounting estimate effected by a change in accounting principle	May only be made “if the new accounting principle is justifiable on the basis that it is preferable” (see ASC 250-10-45-19). An SEC registrant is not required to file a preferability letter to make a change in accounting estimate effected by a change in accounting principle (see paragraph 4230.2(c)(4) of the SEC Financial Reporting Manual).	Treated as a change in accounting estimate (see ASC 250-10-45-18).

The determination of the allowance for credit losses on financial assets reflects an accounting estimate. Making such an estimate involves the determination of both the estimation technique (e.g., DCFs, loss-rate method) and the assumptions inherent in that technique. ASC 326-20-55-6 notes that “[e]stimating expected credit losses is highly judgmental” and includes the following nonexhaustive list of judgments that may be involved in such estimation:

The definition of default for default-based statistics
The approach to measuring the historical loss amount for loss-rate statistics, including whether the amount is simply based on the amortized cost amount written off and whether there should be adjustments to historical credit losses (if any) to reflect the entity’s policies for recognizing accrued interest
The approach to determine the appropriate historical period for estimating expected credit loss statistics
The approach to adjusting historical credit loss information to reflect current conditions and reasonable and supportable forecasts that are different from conditions existing in the historical period
The methods of utilizing historical experience
The method of adjusting loss statistics for recoveries
How expected prepayments affect the estimate of expected credit losses
How the entity plans to revert to historical credit loss information for periods beyond which the entity is able to make or obtain reasonable and supportable forecasts of expected credit losses
The assessment of whether a financial asset exhibits risk characteristics similar to other financial assets.

As part of its processes for estimating expected credit losses, an entity will need to continually obtain and evaluate new information to determine how it affects the presentation of the net amounts expected to be collected on financial assets. Such processes could include changing the estimation technique for specific types of financial assets, changing the assumptions inherent in the application of a particular estimation technique, or both.

4.4.1 Changes in Estimation Techniques

ASC 326-20-35-1 and ASC 326-20-55-7 address considerations related to determining the estimation technique used to calculate expected credit losses and state:

35-1 At each reporting date, an entity shall record an allowance for credit losses on financial assets (including purchased financial assets with credit deterioration) within the scope of this Subtopic. An entity shall compare its current estimate of expected credit losses with the estimate of expected credit losses previously recorded. An entity shall report in net income (as a credit loss expense or a reversal of credit loss expense) the amount necessary to adjust the allowance for credit losses for management’s current estimate of expected credit losses on financial asset(s). The method applied to initially measure expected credit losses for the assets included in paragraph 326-20-30-14 generally would be applied consistently over time and shall faithfully estimate expected credit losses for financial asset(s). [Emphasis added]

55-7 Because of the subjective nature of the estimate, this Subtopic does not require specific approaches when developing the estimate of expected credit losses. Rather, an entity should use judgment to develop estimation techniques that are applied consistently over time and should faithfully estimate the collectibility of the financial assets by applying the principles in this Subtopic. An entity should utilize estimation techniques that are practical and relevant to the circumstance. The method(s) used to estimate expected credit losses may vary on the basis of the type of financial asset, the entity’s ability to predict the timing of cash flows, and the information available to the entity. [Emphasis added]

An entity may consider changing its estimation technique for a particular type of financial asset for various reasons. For example, in practice, it may be common for a financial asset’s risk characteristics to no longer be similar to those of the original pool of financial assets. In such situations, an entity may need to measure expected credit losses on the asset by using a method different from that used for the original pool. Accordingly, it may be appropriate for the entity to change the estimation technique for the financial asset and, depending on the facts and circumstances, the new estimation technique may be applied to the financial asset individually or to a different pool of financial assets whose risk characteristics are now similar to those of that asset. For instance, an entity may determine that the allowance for credit losses on a credit-deteriorated financial asset (including, but not limited to, a modified financial asset) should be estimated by using a DCF approach or by applying the practical expedient related to collateral-dependent financial assets, but the entity may also determine that the allowance for credit losses on the original pool of financial assets should continue to be estimated by using a loss-rate approach.

If there is no change in risk characteristics, an entity would generally be expected to apply its estimation technique consistently over time. However, when a change in an estimation technique is justified given the particular facts and circumstances, an entity should treat the change as a change in accounting estimate rather than as a change in accounting estimate effected by a change in accounting principle. Although the definition of a change in accounting principle in ASC 250 indicates that “[a] change in the method of applying an accounting principle also is considered a change in accounting principle” (emphasis added), the definition of a change in accounting estimate cites uncollectible receivables as an item for which estimates are necessary. In prior registrant comment letters and informal conversations, the SEC staff has expressed its view that changes in the method of estimating the allowance for credit losses represent changes in accounting estimates. We do not expect a change in this view upon adoption of ASU 2016-13.

4.4.2 Applying Different Measurement Methods to Similar Pools of Assets

An entity can select from a number of measurement approaches to estimate expected credit losses. Some approaches project future principal and interest cash flows (i.e., a DCF method), while others project only future principal losses. An entity is not necessarily required to apply the same measurement approach when measuring expected credit losses related to similar pools of assets. While it is clear that an entity could use different approaches to measure expected credit losses related to individual financial assets that do not share similar risk characteristics, it is less clear whether that same conclusion applies to similar pools of financial assets. As discussed in Chapter 3, an entity must evaluate financial assets on a collective (i.e., pool) basis if they share similar risk characteristics. Once the entity determines a pool of financial assets, it can apply any of the measurement methods discussed in ASC 326-20 to that pool. As a result, an entity could potentially apply a different measurement approach to different pools of similar financial assets.

Example 4-1

A commercial real estate lender has historically provided financing in the northeastern and southeastern United States and has determined its pools of assets for measuring expected credit losses on the basis of internal risk rating (1–5). The internal risk rating takes into account borrower-specific factors such as size, geographical location, historical payment defaults, and loan term (note that the lender does not believe that the borrower-specific factors cause the loans not to share similar risk characteristics). Given the amount of historical information gathered on these commercial pools, the lender uses a DCF method to estimate expected credit losses related to the pools.

In the current year, the lender begins to finance commercial loans in the southwestern United States. As the lender has historically done, it assigns an internal risk rating to the loans issued in the southwestern United States and determines that those loans should be pooled together. However, the lender does not have sufficient information to apply a DCF method to these loans (as noted above, the lender did have such information for the loans issued in the northeastern and southeastern United States). Therefore, on the basis of the guidance in ASC 326-20-55-7, which emphasizes that an entity should use methods that are “practical and relevant” given the specific facts and circumstances and that “[t]he method(s) used to estimate expected credit losses may vary on the basis of . . . the entity’s ability to predict the timing of cash flows, and the information available to the entity,” the lender chooses to apply a loss-rate method to the pool of commercial real estate loans issued in the southwestern United States and continues to apply a DCF method to the pools of commercial real estate loans issued in the northeastern and southeastern United States.

4.4.3 Changes in Assumptions

Changes to the assumptions an entity uses in applying a particular estimation technique are appropriate when the entity is making those changes because it has obtained and evaluated new information that affects its process for faithfully estimating the collectibility of financial assets in accordance with ASC 326. Any such changes, which should be justifiable given the particular facts and circumstances, should be reflected as a change in accounting estimate. Such changes would include the nonexhaustive list of judgments listed in ASC 326-20-55-6.

While changes in the estimation techniques and assumptions used to measure expected credit losses are considered changes in accounting estimates, changes in certain items, which are described in ASC 326 as “significant accounting policies,” could indirectly affect the estimation of credit losses and would need to be reflected as changes in accounting principles. ASC 326-20-50-17 states that an entity’s significant accounting policies would include the following:

Nonaccrual policies, including the policies for discontinuing accrual of interest, recording payments received on nonaccrual assets (including the cost recovery method, cash basis method, or some combination of those methods), and resuming accrual of interest, if applicable
The policy for determining past-due or delinquency status
The policy for recognizing writeoffs within the allowance for credit losses.

In addition, an entity’s change in its interest income recognition practices may reflect a change in accounting principle or a change in accounting estimate effected by a change in accounting principle. Interest income recognition guidance is largely addressed in other Codification topics. However, ASC 326 does address an entity’s treatment of the change in the allowance for credit losses that results from the passage of time when a DCF approach is used to estimate expected credit losses. ASC 326-20-45-3 allows an entity to either (1) report the entire change in present value as credit loss expense or (2) report the change in present value attributable to the passage of time as interest income. The alternative chosen would reflect an accounting policy; thus, any change in the alternative used would be considered a change in accounting principle.

See ASC 250-10-50 for information about the disclosures that must accompany a change in accounting principle or a change in accounting estimate.

4.4.4 Considerations Related to Estimating Credit Losses by Using a DCF Method

4.4.4.1 Effect of Prepayments on an Entity Using a DCF Method

An entity that uses a DCF method to estimate expected credit losses must “discount expected cash flows at the financial asset’s effective interest rate” (emphasis added) in accordance with ASC 326-20-30-4. This paragraph further states that “[w]hen 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.” ASC 326-20-20 defines the EIR, in part, as “[t]he 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.”6

Consequently, ASC 326 is unclear on whether an entity that applies a DCF method to discount the expected cash flows should use the same EIR it applied to recognize interest income in accordance with ASC 310-20. With few exceptions, it is assumed under ASC 310-20 that for interest income recognition purposes, the loan will remain outstanding until its contractual maturity (and that, therefore, expected prepayments are not considered). However, the CECL model requires an entity to consider prepayments when applying the DCF method. As a result, the loan term used for recognizing interest income is inconsistent with that used for estimating expected credit losses.

Further, this inconsistency could result in certain anomalies. For example, when an entity uses a DCF method to estimate expected credit losses for a loan that includes a premium to par, any expected prepayments would accelerate the recognition of premiums that are related solely to the use of a discount rate under which, for CECL model purposes, prepayments are assumed to a set of cash flows for which prepayments are not assumed for interest income recognition purposes. The acceleration of the premium recognition results in an increase to the credit allowance because the amortized cost basis on day 1 would be greater than the present value of the expected cash flows. The opposite is true for a loan that includes a discount to par under which the use of a DCF method that includes expected prepayments would accelerate the recognition of the discount. This acceleration would, in turn, artificially lower the allowance because the amortized cost basis on day 1 would be less than the present value of the expected cash flows.

This issue, among others discussed at the June 2017 TRG meeting, led the FASB to issue ASU 2019-04, under which an entity can make an accounting policy election (at the “class of financing receivable” level) to use a prepayment-adjusted EIR when applying a DCF method under the CECL model, even though the EIR used for interest income recognition is not adjusted for prepayments. The ASU also states that an entity that has elected an accounting policy of adjusting the EIR for prepayments should update the adjusted EIR periodically to match any changes in expected prepayments. Moreover, the ASU clarifies that an entity should not adjust the EIR used to discount expected cash flows for subsequent changes in expected prepayments if the financial asset is restructured in a TDR.7

Changing Lanes

FASB ASU on Troubled Debt Restructurings and Vintage Disclosures

ASU 2022-02 states that an entity that uses a DCF method to calculate the allowance for credit losses will be required to use a postmodification-derived EIR as part of its calculation in accordance with ASC 326-20-30-4.

4.4.4.2 Variable-Rate Instruments

ASU 2016-13 originally stated that if a financial asset’s contractual interest rate varies on the basis of an independent factor, such as an index or rate, “[p]rojections of changes in the factor shall not be made for purposes of determining the effective interest rate or estimating expected future cash flows” (emphasis added). After the ASU was issued, however, stakeholders questioned whether it was inconsistent for the guidance to prohibit entities from projecting changes in the factor that leads to changes in the financial asset’s contractual interest rate while requiring them to consider projections when estimating expected cash flows.

As a result, in ASU 2019-04, the FASB clarifies that an entity is permitted to consider projections of changes in the factor as long as such projections are the same as those used to estimate expected future cash flows. For example, the reasonable and supportable forecast period over which an entity chooses to project the contractual variable interest rate should be consistent for both estimating future cash flows and determining the EIR.

Specifically, ASC 326-20-30-4 states, in part:

If the financial asset’s contractual interest rate varies based on subsequent changes in an independent factor, such as an index or rate, for example, the prime rate, the London Interbank Offered Rate (LIBOR), or the U.S. Treasury bill weekly average, that financial asset’s effective interest rate (used to discount expected cash flows as described in this paragraph) shall be calculated based on the factor as it changes over the life of the financial asset. An entity is not required to project changes in the factor for purposes of estimating expected future cash flows. If the entity projects changes in the factor for the purposes of estimating expected future cash flows, it shall use the same projections in determining the effective interest rate used to discount those cash flows. In addition, if the entity projects changes in the factor for the purposes of estimating expected future cash flows, it shall adjust the effective interest rate used to discount expected cash flows to consider the timing (and changes in the timing) of expected cash flows resulting from expected prepayments in accordance with paragraph 326-20-30-4A. Subtopic 310-20 on receivables — nonrefundable fees and other costs provides guidance on the calculation of interest income for variable rate instruments.

Connecting the Dots

Measuring Expected Credit Losses When an EIR on a Fixed-Rate Loan Is Lower Than a Current Market Rate

ASU 2019-04 clarifies that an entity is permitted to consider projections of changes in factors that lead to changes in a variable-rate instrument’s contractual interest rate when determining the asset’s EIR. However, when an entity is determining the EIR on a fixed-rate loan, even in situations in which interest rates have increased since the origination of the loan, the entity would not consider the current market rate when discounting the expected future cash flows of the loan. Keep in mind that discounting the future cash flows at a rate lower than the current market rate will result in a lower amount of expected credit losses. An entity should change its estimate of expected credit losses solely on the basis of changes in credit quality (i.e., a change in future cash flows that is attributable to credit). The FASB did not intend the measurement of expected credit losses to reflect changes in interest rates in general or changes specific to a borrower.

However, if the creditor uses a practical expedient in measuring expected credit losses (e.g., fair value of collateral on a collateral-dependent loan), the measurement may reflect changes in interest rates because fair value incorporates current market conditions as of the measurement date.

4.4.5 Estimating Credit Losses by Using Methods Other Than a DCF Method

If an entity chooses to estimate credit losses by using a method other than a DCF method, it has the option of estimating credit losses on the asset’s amortized cost basis in the aggregate or by separately measuring the components of the amortized cost basis (e.g., premiums and discounts), as described in ASC 326-20-30-5. Depending on which of these alternatives the entity chooses, it may be required to adjust its historical loss information. For example, if an entity’s historical loss rate reflects losses of only principal amounts, it may need to adjust that information if it chooses to estimate credit losses on the entire amortized cost basis of an asset to which premiums, discounts, or other basis adjustments apply.

4.4.5.1 Accrued Interest

ASU 2016-13 defines “amortized cost basis” as “the amount at which a financing receivable or investment is originated or acquired, adjusted for applicable accrued interest, accretion or amortization of premium, discount, and net deferred fees or costs, collection of cash, writeoffs, foreign exchange, and fair value hedge accounting adjustments” (emphasis added). The ASU’s inclusion of accrued interest in the definition of amortized cost basis has three significant implications for financial statements with respect to the measurement, presentation, and disclosure of the amortized cost basis and the allowance for credit losses of financial assets:

To measure an allowance for credit losses on the amortized cost basis of a financial asset, entities will be required to include an allowance for the applicable accrued interest of that asset.
Entities will have to present the accrued interest amount in the amortized cost basis of the financial assets in the same line item on the balance sheet.
Entities will be required to include accrued interest in their disclosures about the amortized cost basis by class of financing receivable and vintage in accordance with ASC 326-20-50-5 and 50-6, respectively.

Further, because accrued interest is included in the definition, the reversal of such interest will need to be written off in the same manner as the principal or other components of the amortized cost basis (see Section 4.5 for a discussion of write-offs). ASC 326-20-35-8 states that “[w]riteoffs of financial assets, which may be full or partial writeoffs, shall be deducted from the allowance” (emphasis added). In other words, all components of the amortized cost basis, including the accrued interest, must be written off through the allowance for credit losses.

After the issuance of ASU 2016-13, stakeholders raised concerns that the inclusion of accrued interest in the definition of amortized cost basis could be operationally burdensome because many loan systems are not able to track accrued interest on an individual loan level. Stakeholders have also expressed concerns about the conflict between existing nonaccrual policies, which generally follow regulatory instructions requiring the reversal of accrued interest as a debit to the interest income line item (at least in part), and the write-off guidance in ASC 326-20-35-8. Many stakeholders, primarily financial institutions, indicated that existing nonaccrual policies present a more accurate reflection of the earning potential of a loan and interest income than does the write-off guidance in ASC 326-20-35-8. Those stakeholders further maintained that any change from the existing nonaccrual policies would reduce the consistency and comparability of current-period financial statements, regulatory reports, and important interest-income-based metrics (e.g., net interest margin) with those of prior periods.

ASU 2019-04 addresses these concerns (which were originally discussed at the June 2018 TRG meeting). Specifically, ASU 2019-04 states that an entity would be allowed to:

Measure the allowance for credit losses on accrued interest receivable balances separately from other components of the amortized cost basis of associated financial assets.
Make an accounting policy election not to measure an allowance for credit losses on accrued interest receivable amounts if an entity writes off the uncollectible accrued interest receivable balance in a timely manner and makes certain disclosures.
Make an accounting policy election to write off accrued interest amounts by [either] reversing interest income or recognizing credit loss expense, or a combination of both. The entity also is required to make certain disclosures.
Make an accounting policy election to present accrued interest receivable balances and the related allowance for credit losses for those accrued interest receivable balances separately from the associated financial assets on the balance sheet. If the accrued interest receivable balances and the related allowance for credit losses are not presented as a separate line item on the balance sheet, an entity should disclose the amount of accrued interest receivable balances and the related allowance for credit losses and where the balance is presented.
Elect a practical expedient to disclose separately the total amount of accrued interest included in the amortized cost basis as a single balance to meet certain disclosure requirements.

4.4.5.2 Nonaccrual Loans

ASC 310-20-35-17 states that the amortization of net deferred fees and costs should be discontinued when an entity is not accruing interest on a loan “because of concerns about the realization of loan principal or interest.” The guidance in ASC 310-20-35-17 applies even if an entity has net deferred costs or a premium associated with the loan.

While ASC 310-20 does not explicitly refer to premiums and discounts, if a loan is on nonaccrual status, presumably no interest should be recorded, including the amortization of premiums and discounts. Note, however, that when an entity estimates expected credit losses on a nonaccrual loan, it must consider the entire amortized cost basis of the loan, including unamortized net deferred costs and unamortized premiums (irrespective of whether the entity chooses to measure expected credit losses on the asset’s amortized cost basis in the aggregate or by separately measuring the components of the amortized cost basis, as permitted by ASC 326-20-30-5).

4.4.5.3 Discounting Inputs When a Method Other Than a DCF Method Is Used

An entity that uses a method other than a DCF method (e.g., a probability-of-default or a loss-given-default credit loss method) in estimating credit losses is not allowed to discount only certain inputs; partial discounting is prohibited. If an entity wants to discount the inputs used to measure the allowance for credit losses, it should discount all the inputs used in the measurement. Note that this question was addressed at the November 2018 TRG meeting, and it was determined that ASU 2016-13’s guidance on discounting is clear. As a result, the FASB is not expected to amend the guidance in ASC 326 to reflect the TRG discussion.

In addition, it would not be acceptable for an entity to discount future losses when using a loss-rate method (i.e., amortized cost × expected loss rate). Discounting is allowed only under a DCF approach in which an entity uses the EIR at inception.

4.4.5.4 Consideration of Capitalized Interest When a Method Other Than a DCF Method Is Used

When an entity uses a method other than a DCF method to estimate credit losses, its allowance for credit losses should not take into account future capitalized interest, since such interest is not included in the asset’s amortized cost basis on which the entity is estimating expected credit losses. For a financial asset issued at a discount, the amount due upon default — provided that there are no terms or conditions that only require the repayment of “accreted value” at any point in the asset’s term — would be the par amount and accrued interest to date (which is greater than the asset’s amortized cost basis). For a financial asset issued at par with expected future capitalized interest, the amount due upon default is also the par amount and accrued interest to date, which would equal the amortized cost basis at the time of default (as long as there are no other adjustments to the amortized cost basis). Therefore, an entity would not consider any unearned interest, regardless of the expectation that it will be capitalized in future periods before a default occurs, in the calculation of an allowance for credit losses because the legal amount owed upon default would not include that future expected accrued interest.

To illustrate this point, at its June 2018 meeting, the TRG discussed an example in which an entity issues a student loan for $60,000. The student is not required to make any payments on the loan until after four years (i.e., the loan will be in deferment for four years). However, interest accrues during the deferment period in such a way that at the end of year 4, the student will owe the lender $100,000, comprising a principal amount of $60,000 and interest of $40,000. In this scenario, although the student will owe the lender $100,000 after the deferment period ends, the lender will calculate its expected credit losses during the deferment period by using an amortized cost amount that does not include future capitalized interest (i.e., the principal amount of $60,000).

4.4.6 Effect of Timing on Expected Credit Losses

The timing of defaults and prepayments (e.g., repayment of the financial asset, either partially or entirely, before its stated maturity according to its contractual terms) affects an entity’s calculation of expected credit losses differently depending on the method used:

DCF method — When a DCF method is applied to a pool of financial assets, the time value of money is explicitly incorporated (i.e., both the amount and timing of cash flows matter). For example, the timing of prepayments affects the timing of the recognition of discounts and premiums and the number of interest coupons to be received. These factors could increase or offset credit losses when a DCF method is applied to a pool of assets by using an overall EIR.
Measuring expected credit losses on the separate components of amortized cost (e.g., premiums, discounts) or on the asset’s combined amortized cost basis — Regardless of whether an entity estimates expected credit losses on the separate components of amortized cost or the combined amortized cost basis of the asset, an entity is permitted but not required to consider the timing of when credit losses will occur.8 However, if an entity chooses to consider timing when estimating expected credit losses, the timing will affect the amount of amortized premiums, discounts, deferred fees and costs, etc. In addition, the entity would also need to estimate the acceleration of the amortization of the amounts resulting from prepayments.

4.4.7 Inclusion of Taxes, Insurance, and Other Costs

A lender’s expectations of future losses on payments of tax, insurance premiums, and other “costs” (i.e., payments made by the lender that may not be recovered from borrowers) should not be included in the lender’s day 1 estimate of expected credit losses. Those amounts should only be included in its estimate of expected credit losses when the funds are advanced. Some would argue that because the lender is not obligated to make those payments on the borrower’s behalf, not establishing an allowance for these amounts before such advances is consistent with the treatment of unfunded loan commitments discussed in Section 2.1. Others would argue that even if the lender was effectively compelled to make the advances to protect its collateral, not including estimates of future advances in the expected credit losses is consistent with the treatment of future capitalized interest discussed in Section 4.4.5.4 since those amounts do not represent the amount owed by the borrower on the balance sheet date.

4.4.8 Weighted-Average Remaining Maturity Method

In January 2019, the FASB staff issued a Q&A document that addresses whether and, if so, how the WARM method could be used to estimate expected credit losses. The document states that the “WARM method uses an average annual charge-off rate [that] contains loss content over several vintages and is used as a foundation for estimating the credit loss content for the remaining balances of financial assets in a pool at the balance sheet date. The average annual charge-off rate is applied to the contractual term, further adjusted for estimated prepayments to determine the unadjusted historical charge-off rate for the remaining balance of the financial assets.” The staff indicated that it is acceptable to use the WARM method to estimate an allowance for credit losses, particularly for less complex financial asset pools, and that an entity needs to consider whether qualitative adjustments should be made.

To illustrate how an entity might apply the WARM method, the FASB staff included a number of examples in the Q&A, one of which is reproduced below.

FASB Staff Q&A

Topic 326, No. 1: Whether the Weighted-Average Remaining Maturity Method Is an Acceptable Method to Estimate Expected Credit Losses

Question 3 . . .

Fact Pattern

Estimate the allowance for credit losses as of 12/31/2020
Pool of financial assets of similar risk characteristics
- Amortized cost basis of ~$13.98 million
- 5-year financial assets (contractual term adjusted by prepayments)
Management expects the following in 2021 and 2022:
- Rise in unemployment rates
Management cannot reasonably forecast beyond 2022
Assume 0.25% qualitative adjustment to represent both current conditions and reasonable and supportable forecasts

The example illustrates estimating an allowance for credit losses on a pool of financial assets as of December 31, 2020. The pool has an outstanding balance of approximately $13.98 million as of December 31, 2020 and has financial assets with a contractual life of 5 years. The $13.98 million amortized cost is for a pool of financial assets with similar credit risk characteristics.

Management expects a rise in unemployment rates for 2021 and 2022 and cannot reasonably forecast beyond 2022. The example assumes a 0.25% qualitative adjustment for current conditions and reasonable and supportable forecasts discussed further below. It is important to note that this input will be a significant assumption when estimating expected credit losses under Update 2016-13 because it represents amounts for the current conditions and reasonable and supportable forecast. Moreover, because the example is for illustrative purposes, the staff has not assumed a specific type of financial asset pool given the breadth of products that exist in the market place and the specific facts and circumstances that may exist for a particular entity. Rather, the calculations are meant to depict the mechanics of the model in various ways. Therefore, as noted in the example calculations, an entity will need to determine if adjustments need to be made to historical loss data in accordance with paragraph 326-20-30-8 in addition to the reasonable and supportable forecasts.

Step 1: Calculate Annual Charge-Off Rate

In Table 1 above:

Red bolded number of 0.36% is an average of 5 years of annual charge-off rates.
The historical time period used to determine the average annual charge-off rate is a significant judgment that will need to be properly supported and documented in accordance with paragraph 326-20-30-8. For this example, assume the entity compared historical information for similar financial assets with the current and forecasted direction of the economic environment, and believes that its most recent 5-year period is a reasonable period on which to base its expected credit-loss-rate calculation after considering the underwriting standards and contractual terms for loans that existed over the historical period in comparison with the current pool. Additionally, assume the entity considered whether any adjustments to historical loss information in accordance with paragraph 326-20-30-8 were needed before considering adjustments for current conditions and reasonable and supportable forecasts but determined that none were necessary. It should be noted that this is a simplified example using a generic pool. An entity that estimates the allowance for credit losses using the WARM method (or any method) should determine if its historical loss information needs to be adjusted for changes in underwriting standards, portfolio mix, or asset term within the pool at the reporting date.

Step 2: Estimate the Allowance for Credit Losses

In Table 2 above:

First column titled “Year End” displays subsequent years, until 2025, which represents the time anticipated for the pool to be paid off.
Second column titled “Est. Paydown” represents expected payments in the future periods until the pool is expected to fully pay off. Management will need to estimate the future paydowns, which includes the scheduled payments + prepayments.

Note: Do not include the expected credit losses in this column. Paydowns should include scheduled payments and non-credit related prepayments.

Note: Estimated prepayments are also a significant judgment that will need to be properly supported and documented.
Third column titled “Projected Amort Cost”:
1. Begin with $13.98MM outstanding balance as of the balance sheet date of 12/31/2020.
2. Subtract projected paydowns from the “Est. Paydown” column to estimate future projected amortized cost for each of the remaining years of the pool’s life (for example, $13,980M minus $3,700M equals $10,280M).
Fifth column titled “Allowance for Credit Losses”:
1. Take each of the future years’ projected amortized cost and multiply by the average annual charge-off rate, thereby estimating each of the remaining years’ losses and aggregating to estimate the cumulative losses (for example, in the first year, $13.98MM of amortized cost is multiplied by the average annual charge-off rate of 0.36% for a first year’s credit loss estimate of $50K dollars).
2. For the second year, which is 2022, the $10.28MM representing the ending balance as of 2021 and the beginning balance as of 2022 is multiplied by the average annual charge-off rate of 0.36% to estimate the second year’s credit losses of $37K dollars. This process is repeated for each remaining year.
3. Sum the last column to estimate the total expected credit losses of $126K dollars.
  Note: This is not the full allowance for credit losses because the entity has not yet accounted for current conditions and reasonable and supportable forecasts.
4. Convert $126K of expected losses into a loss rate of 0.90% by dividing $126K by the amortized cost of $13.98MM.
Finally, add 0.25% of qualitative adjustments as an assumption established as part of the fact pattern of the example to estimate the allowance for credit losses rate of 1.15%. The 1.15% is multiplied by $13.98MM to estimate the total allowance for credit losses of $161K dollars.
Note: 0.25% is a significant assumption made by management that will need to be adequately documented and supported. For this example, in accordance with paragraph 326-20-55-4, the entity considered significant factors that could affect the expected collectability of the amortized cost basis of the pool and determined that the primary factor is the unemployment rate. As part of this analysis, assume that the entity observed that the unemployment rate has increased as of the current reporting period date. Based on current conditions and reasonable and supportable forecasts, the entity expects that unemployment rates are expected to increase further over the next one to two years. To adjust the historical loss rate to reflect the effects of those differences in current conditions and forecasted changes, the entity estimates a 25-basis-point increase in credit losses incremental to the 0.9 percent historical lifetime loss rate related to the expected deterioration in unemployment rates. Management estimates that the incremental 25-basis-point increase based on its knowledge of historical loss information during past years in which there were similar trends in unemployment rates. Management is unable to support its estimate of expectations for unemployment rates beyond the reasonable and supportable forecast period. Under this loss-rate method, the incremental credit losses for the current conditions and reasonable and supportable forecast (the 25 basis points) is added to the 0.9 percent rate that serves as the basis for the expected credit loss rate. No further reversion adjustments are needed because the entity has applied a 1.15% loss rate where it has immediately reverted into historical losses that reflect the contractual term in accordance with paragraphs 326-20-30-8 through 30-9. This approach reflects an immediate reversion technique for the loss-rate method. It is important to note that the 25-basis-point increase reflects the entity’s estimate of the incremental losses in years 2021 and 2022 from unemployment and assumes no incremental losses for the remaining years. Further, the reversion technique selected by the entity is a significant assumption that will need to be supported by management and is not a policy election or practical expedient.

Connecting the Dots

Applicability of the WARM Method

The FASB staff indicated that it is acceptable to use the WARM method to estimate expected credit losses. Under that method, entities use qualitative adjustments to alleviate the operational challenges they may face when applying other loss methods. For example, in its response to Question 2 of the Q&A document, the staff suggests that such qualitative adjustments may be used to overcome “situations involving minimal loss history, losses that are sporadic with no predictive patterns, low numbers of loans in each pool, data that is only available for a short historical period, a composition that varies significantly from historical pools of financial assets, or changes in the economic environment.”

However, the FASB staff also cautions preparers that although the WARM method can be used effectively in some situations, certain “challenges will be more significant, and an entity may find that the WARM method is inappropriate for its situation.” Because the WARM method relies heavily on qualitative adjustments for which significant judgment is required, we believe that entities may discover that the costs of applying it outweigh its benefits. As a result, we do not expect many entities to apply the WARM method in practice.

4.4.9 Practical Expedients Related to Measuring Expected Credit Losses

ASC 326 permits entities to use practical expedients to measure expected credit losses for the following two types of financial assets:

Collateral-dependent financial assets — In a manner consistent with its practice under existing U.S. GAAP, an entity is permitted to measure its estimate of expected credit losses for collateral-dependent financial assets as the difference between the financial asset’s amortized cost and the collateral’s fair value (adjusted for selling costs, when applicable).
Financial assets for which the borrower must continually adjust the amount of securing collateral (e.g., certain repurchase agreements and securities-lending arrangements) — An entity is permitted to measure its estimate of expected credit losses on these financial assets as the difference between the amortized cost basis of the asset and the collateral’s fair value.

4.4.9.1 Collateral-Dependent Financial Assets

ASU 2016-13 does not change the U.S. GAAP guidance on determining when a financial asset is a collateral-dependent financial asset. As is consistent with previous U.S. GAAP, a financial asset is considered collateral-dependent if “repayment is expected to be provided substantially through the operation or sale of the collateral.” Furthermore, under ASC 326-20-35-4, and in a manner consistent with previous U.S. GAAP, an entity is required to measure the allowance for credit losses on the basis of the fair value of the collateral when it determines that foreclosure is probable.

However, ASU 2016-13 does change an entity’s ability to measure the allowance for credit losses on a collateral-dependent financial asset when foreclosure is not probable. Under previous U.S. GAAP, an entity could elect, as a practical expedient, to measure the allowance for credit losses on a collateral-dependent financial asset that is impaired on the basis of the collateral’s fair value in any circumstance. By contrast, ASC 326-20-35-5 indicates that if foreclosure is not probable, an entity can elect to use such a practical expedient for collateral-dependent financial assets only “when the borrower is experiencing financial difficulty based on the entity’s assessment as of the reporting date.” The phrase “when the borrower is experiencing financial difficulty” is consistent with the guidance that an entity considers to determine whether a modification of a financial asset (1) reflects a TDR before the adoption of ASU 2022-02 or (2) is subject to the disclosure requirements in ASC 310-10-50-42 through 50-44 (added by ASU 2022-02). Therefore, the guidance in ASC 310-40 (before the adoption of ASU 2022-02) and 310-10-50 (after the adoption of ASU 2022-02) that is used to determine whether a debtor is experiencing financial difficulties is relevant to the determination of an entity’s ability to use the practical expedient for a collateral-dependent financial asset. Before the adoption of ASU 2022-02, ASC 310-40-15-20 indicates that the following indicators9 should be considered in the determination of when a debtor may be experiencing financial difficulties (this list is not all-inclusive):

The debtor is currently in payment default on any of its debt. In addition, a creditor shall evaluate whether it is probable that the debtor would be in payment default on any of its debt in the foreseeable future without the modification. That is, a creditor may conclude that a debtor is experiencing financial difficulties, even though the debtor is not currently in payment default.
The debtor has declared or is in the process of declaring bankruptcy.
There is substantial doubt as to whether the debtor will continue to be a going concern.
The debtor has securities that have been delisted, are in the process of being delisted, or are under threat of being delisted from an exchange.
On the basis of estimates and projections that only encompass the debtor’s current capabilities, the creditor forecasts that the debtor’s entity-specific cash flows will be insufficient to service any of its debt (both interest and principal) in accordance with the contractual terms of the existing agreement for the foreseeable future.
Without the current modification, the debtor cannot obtain funds from sources other than the existing creditors at an effective interest rate equal to the current market interest rate for similar debt for a nontroubled debtor.

In determining when the practical expedient for collateral-dependent financial assets can be applied, an entity should also consider the FASB’s rationale for changing previous U.S. GAAP to limit the application of this expedient. Since the CECL model is built on the foundation that expected credit losses are estimated over the contractual life of a financial asset, the FASB was concerned that allowing an unrestricted practical expedient for collateral-dependent financial assets could enable entities to defer the recognition of credit losses that are expected to occur over the asset’s contractual life as a result of declines in the collateral’s fair value. Consequently, under the guidance, the practical expedient can only be elected if the borrower is experiencing financial difficulties, since only at this point does the collateral’s fair value signify a reasonable expectation of the recoverability of the financial asset. This guidance is consistent with paragraph BC64 of ASU 2016-13, which refers to the acceptability of measuring expected credit losses on the basis of the fair value of the collateral “because fair value reflects the amount expected to be collected.” Before the point at which a debtor is experiencing financial difficulties, the collateral’s fair value would not necessarily reflect the net amount of the financial asset expected to be collected because of an entity’s prolonged exposure to declines in that fair value.

Entities may wish to implement accounting practices under which they perform an objective determination of when a borrower is experiencing financial difficulties (e.g., when a loan is a certain number of days past due or becomes subject to the entity’s nonaccrual policy). The reasonableness of such practices would depend on the specific facts and circumstances, including the financial asset type, and any such objective evidence would need to be supplemented by qualitative considerations since an entity must use judgment and consider its particular facts and circumstances in determining when a borrower is experiencing financial difficulties.

Since ASC 326 does not require entities to use the practical expedient for collateral-dependent financial assets before foreclosure is probable, the principal risk in the application of objective evidence lies in prematurely measuring expected credit losses on the basis of the collateral’s fair value. Rather, to apply this practical expedient when foreclosure is not probable, an entity must appropriately assess all relevant facts and circumstances to determine whether the borrower is experiencing financial difficulties. The entity may consider both quantitative and qualitative indicators in performing this assessment. While the mere fact that a borrower is past due on its payments would generally indicate that it is experiencing financial difficulties, this may not always be the case. For example, a borrower may be past due on a loan because of a payment delay was temporarily caused by a natural disaster or other similar event that did not have a significant bearing on the borrower’s wherewithal to make the contractually required payments on the loan receivable.

Keep in mind that the entity is only allowed to use the practical expedient (when foreclosure is not probable) if the borrower is experiencing financial difficulty. As a result, the entity would be permitted to continue to use the practical expedient only if it concludes that the borrower continues to experience financial difficulty. In other words, the conditions related to using the practical expedient when foreclosure is not probable must be met in every reporting period. To the extent that the entity determines that the borrower is no longer experiencing financial difficulty (or that repayment will no longer be substantially provided through the collateral’s sale or operation), the entity would need to measure expected credit losses by using another appropriate measurement method discussed in ASC 326-20-30-3.

An entity must use judgment in determining whether repayment of a loan is expected to be provided solely by the underlying collateral. On October 24, 2013, the federal financial institution regulatory agencies jointly issued “Interagency Supervisory Guidance Addressing Certain Issues Related to Troubled Debt Restructurings” (the “Interagency Guidance”10), which provides the following additional interpretation of the definition of a collateral-dependent loan:

An impaired loan is collateral dependent if “repayment is expected to be provided solely by the underlying collateral,” which includes repayment from the proceeds from the sale of the collateral, cash flow from the continued operation of the collateral, or both. Whether the underlying collateral is expected to be the sole source of repayment for an impaired loan is a matter requiring judgment as to the availability, reliability, and capacity of sources other than the collateral to repay the debt. Generally, repayment of an impaired loan would be expected to be provided solely by the sale or continued operation of the underlying collateral if cash flows to repay the loan from all other available sources (including guarantors) are expected to be no more than nominal. For example, the existence of a guarantor is one factor to consider when determining whether an impaired loan, including a TDR loan, is collateral dependent. To assess the extent to which a guarantor provides repayment support, the ability and willingness of the guarantor to make more-than-nominal payments on the loan should be evaluated.

The repayment of some impaired loans collateralized by real estate may depend on cash flow generated by the operation of a business or from sources outside the scope of the lender’s security interest in the collateral, such as cash flows from borrower resources other than the collateral. These loans are generally not considered collateral dependent due to the more-than-nominal payments expected to come from these other repayment sources. For such loans, even if a portion of the cash flow for repayment is expected to come from the sale or operation of the collateral (but not solely from the sale or operation of the collateral), the loan would not be considered collateral dependent.

For example, an impaired loan collateralized by an apartment building, shopping mall, or other income-producing property where the anticipated cash flows for loan repayment are expected to be derived solely from the property’s rental income, and there are no other available and reliable repayment sources, would be considered collateral dependent because repayment is expected to be provided only from the continued operation of the collateral. However, an impaired loan secured by the owner-occupied real estate of a business (such as a manufacturer or retail store) where the anticipated cash flows to repay the loan are expected to be derived from the borrower’s ongoing business operations and activities would not be considered collateral dependent because the loan is not expected to be repaid solely from cash flows from the sale or operation of the collateral. Nevertheless, if the borrower’s condition worsens so that any payments from the operation of the business are expected to be nominal and repayment instead is expected to depend solely on the sale or operation of the underlying collateral, the loan would then be considered collateral dependent. [Footnotes omitted]

Under this guidance, a loan would be collateral-dependent if it is a nonrecourse loan and the creditor expects to foreclose on it, because in such circumstances, the creditor may only look to the collateral for satisfaction of the loan. Conversely, if the creditor is unsure whether it will foreclose on the collateral, the loan would not be considered collateral-dependent unless it is expected to be repaid solely through the continued operation of the collateral.11

However, if the loan is collateralized by real estate and the creditor has recourse to the general credit of the borrower or to a guarantor, the assessment is less clear. The creditor would need to (1) consider whether the borrower’s financial ability to satisfy the obligation would include repayment sources other than the collateral and (2) formulate an expectation regarding the borrower’s future actions. If a debtor has no means of repaying the loan other than through operation of the collateral, the creditor should conclude that the loan is collateral-dependent. However, if more-than-nominal payments are expected to come from other repayment sources, the loan would generally not be considered collateral-dependent. A similar assessment should be performed when a source of repayment may come from a guarantor. The ability and willingness of a guarantor to make more-than-nominal payments on a loan would result in a conclusion that the loan is not collateral-dependent.

4.4.9.1.1 Considering Collateral Values in the Measurement of Expected Credit Losses

For collateral-dependent financial assets, ASC 326-20-35-4 requires an entity to measure expected credit losses on the basis of the collateral’s fair value if the entity determines that foreclosure is probable. Further, even if a financial asset does not qualify for the practical expedient for collateral-dependent financial assets, the collateral value is still relevant to the estimate of expected credit losses on the financial asset. ASC 326-20-30-10 states, in part:

Except for the circumstances described in paragraphs 326-20-35-4 through 35-6, an entity shall not expect nonpayment of the amortized cost basis to be zero solely on the basis of the current value of collateral securing the financial asset(s) but, instead, also shall consider the nature of the collateral, potential future changes in collateral values, and historical loss information for financial assets secured with similar collateral.

Accordingly, if the practical expedient for collateral-dependent financial assets in ASC 326-20-35-5 is not applied, an entity should consider, among other factors that may affect expected credit losses on the financial asset (or group of similar financial assets), how the collateral’s fair value may affect the estimation of such losses over the contractual life of the financial asset (or group of financial assets).12 If the practical expedient is not applied, an entity cannot measure the allowance for credit losses solely on the basis of the reporting-date fair value of the collateral or avoid recognizing an allowance for credit losses solely because the reporting-date fair value of the collateral equals or exceeds the amortized cost basis of the financial asset (or group of financial assets).

Nevertheless, the collateral security on a financial asset would be expected to affect an entity’s estimation of expected credit losses. In this regard, as discussed in ASC 326-20-30-10, an entity should consider the nature of the collateral (including the nature of the security and the seniority thereon), potential future changes in the collateral’s fair value, and historical losses (adjusted for current conditions and reasonable and supportable forecasts) for financial assets secured with similar collateral. ASC 326-20-55-15 further indicates that debt-to-value ratios and collateral affect the credit quality indicators that are pertinent to financial assets. Accordingly, the collateral securing a financial asset is likely to have an impact on both the probability of default and the loss-given default on the financial asset.

If the practical expedient for collateral-dependent financial assets is not applied, there could potentially be no allowance for credit losses on financial assets, although such situations are not typical. Paragraph BC63 of ASU 2016-13 states:

The Board decided that an entity should consider the expected risk of loss, even if that risk is remote, and that an entity need not measure an expected credit loss when historical information adjusted for current conditions and reasonable and supportable forecasts results in an expectation that the risk of nonpayment of the amortized cost basis is zero. The Board decided not to explicitly state which financial assets are appropriate to have a zero allowance for expected credit losses. The Board understands that an expectation of zero loss is entirely based on the nature and characteristics of a financial asset, which may change over time. As a result, the Board concluded that a “bright-line” approach would be inappropriate for all facts and circumstances and decided not to provide explicit guidance on what specific assets are appropriate for zero expected credit losses. The Board decided that an entity should determine at the reporting date an estimate of credit loss that best reflects its expectations (or its best estimate of expected credit loss).

For example, it may be appropriate not to recognize an allowance for credit losses on certain secured financial assets in which the loan-to-value ratio is so low that it reduces the expected credit losses to zero. To make such a determination, an entity would need to appropriately consider the relevant facts and circumstances, including, but not limited to, the following:

The remaining contractual term of the financial asset (and, if relevant, the entity’s ability to call the financial asset upon a decline in the collateral’s fair value).
The nature of the collateral.
The nature and terms of the security (including any subordination provided by other interests in the collateral).
Past experience with similar collateralized financial assets.
Current conditions and reasonable and supportable forecasts that may affect the prior historical loss information (e.g., potential reasonable and supportable declines in the fair value of the collateral that have not occurred in the past). An entity is not required to consider extremely remote scenarios but should consider those that are reasonably possible.

However, if an entity is aware of a historical default on a particular asset or an asset with similar risk characteristics, regardless of whether the asset was held by the entity or another entity, it would be difficult for the entity to conclude that it is not required to recognize an allowance for credit losses for that asset or asset class. In addition, an entity should consider that paragraph BC71 of ASU 2016-13 states, in part, that “it would be inappropriate to measure credit losses for financial assets on an individual basis to arrive at a zero expected credit loss when a pool of financial assets with similar risk characteristics exists that would indicate otherwise.”

4.4.9.2 Consideration of Costs to Sell When Foreclosure Is Probable

In estimating expected credit losses, an entity that believes foreclosure is probable should consider the costs to sell the collateral securing the financial asset. ASU 2019-04 amended ASC 326-20-35-4 to clarify that an entity is required to adjust the fair value of the collateral by the estimated costs to sell if it intends to sell rather than operate the collateral when it determines that foreclosure on a financial asset is probable.

4.4.9.2.1 Collateral Maintenance Provisions

ASC 326-20

35-6 For certain financial assets, the borrower may be contractually required to continually adjust the amount of the collateral securing the financial asset(s) as a result of fair value changes in the collateral. In those situations, if an entity reasonably expects the borrower to continue to replenish the collateral to meet the requirements of the contract, an entity may use, as a practical expedient, a method that compares the amortized cost basis with the fair value of collateral at the reporting date to measure the estimate of expected credit losses. An entity may determine that the expectation of nonpayment of the amortized cost basis is zero if the fair value of the collateral is equal to or exceeds the amortized cost basis of the financial asset and the entity reasonably expects the borrower to continue to replenish the collateral as necessary to meet the requirements of the contract. If the fair value of the collateral at the reporting date is less than the amortized cost basis of the financial asset and the entity reasonably expects the borrower to continue to replenish the collateral as necessary to meet the requirements of the contract, the entity shall estimate expected credit losses for the unsecured amount of the amortized cost basis. The allowance for credit losses on the financial asset is limited to the difference between the fair value of the collateral at the reporting date and the amortized cost basis of the financial asset.

In arrangements in which the borrower must continually adjust the collateral securing the asset to reflect changes in the collateral’s fair value (e.g., reverse repurchase arrangements), the entity is permitted to measure its estimate of expected credit losses on these financial assets only on the basis of the unsecured portion of the amortized cost as of the balance sheet date (i.e., the difference between the amortized cost basis of the asset and the collateral’s fair value). In such arrangements, the entity must reasonably expect that the borrower will continue to replenish the collateral as necessary. Under the practical expedient for the collateral maintenance provision, the entity may assume that there will be zero losses on the portion of the asset’s amortized basis that is equal to the fair value of the collateral as of the balance sheet date. If the fair value of the collateral is equal to or greater than the amortized cost of the asset, the expected losses would be zero. If the fair value of the collateral is less than the amortized cost of the asset, the expected losses are limited to the difference between the fair value of the collateral and the amortized cost basis of the asset. Example 7 in ASC 326-20 illustrates application of this practical expedient.

ASC 326-20

Example 7: Estimating Expected Credit Losses — Practical Expedient for Financial Assets With Collateral Maintenance Provisions

55-45 This Example illustrates one way an entity may implement the guidance in paragraph 326-20-35-6 for estimating expected credit losses on financial assets with collateral maintenance provisions.

55-46 Bank H enters into a reverse repurchase agreement with Entity I that is in need of short-term financing. Under the terms of the agreement, Entity I sells securities to Bank H with the expectation that it will repurchase those securities for a certain price on an agreed-upon date. In addition, the agreement contains a provision that requires Entity I to provide security collateral that is valued daily, and the amount of the collateral is adjusted up or down to reflect changes in the fair value of the underlying securities transferred. This collateral maintenance provision is designed to ensure that at any point during the arrangement, the fair value of the collateral continually equals or is greater than the amortized cost basis of the reverse repurchase agreement.

55-47 At the end of the first reporting period after entering into the agreement with Entity I, Bank H evaluates the reverse repurchase agreement’s collateral maintenance provision to determine whether it can use the practical expedient in accordance with paragraph 326-20-35-6 for estimating expected credit losses. Bank H determines that although there is a risk that Entity I may default, Bank H’s expectation of nonpayment of the amortized cost basis on the reverse repurchase agreement is zero because Entity I continually adjusts the amount of collateral such that the fair value of the collateral is always equal to or greater than the amortized cost basis of the reverse repurchase agreement. In addition, Bank H continually monitors that Entity I adheres to the collateral maintenance provision. As a result, Bank H uses the practical expedient in paragraph 326-20-35-6 and does not record expected credit losses at the end of the first reporting period because the fair value of the security collateral is greater than the amortized cost basis of the reverse repurchase agreement. Bank H performs a reassessment of the fair value of collateral in relation to the amortized cost basis each reporting period.

4.4.9.2.2 Frequency of Collateral Replenishment

ASC 326-20-35-6 states, in part, that “[a]n entity may determine that the expectation of nonpayment of the amortized cost basis is zero if the fair value of the collateral is equal to or exceeds the amortized cost basis of the financial asset and the entity reasonably expects the borrower to continue to replenish the collateral.”

The guidance is not clear on how frequently an entity must post additional collateral to meet the condition that it “continually” replenish the collateral. While “continually” may mean that the replenishment must occur on a daily basis, certain arrangements require that collateral be replenished only if a specific minimum threshold is reached (i.e., if the change in the fair value of the collateral is greater than a specified amount). In such arrangements, the requirements to use the practical expedient would never be met, which we believe was not the FASB’s intent. Therefore, we think that an entity should use judgment when evaluating the terms of the arrangement with respect to posting additional collateral. In performing this evaluation, the entity should consider whether there are certain situations in which the posting of additional collateral is either required or prohibited when a specified threshold is reached.

In addition, we believe that the entity should also consider the collateral’s nature as well as how frequently it evaluates the collateral’s adequacy. The adequacy of the collateral’s value should be evaluated continually, even daily. If the evaluation is performed less frequently than daily, we believe that it would be more difficult for the entity to support its use of the practical expedient. Furthermore, the entity should consider the nature of the collateral because the more liquid and observable the collateral’s value is, the easier it is to support the collateral’s adequacy in the arrangement. The less liquid and observable the value of the collateral is (e.g., Level 3 for fair value disclosure purposes), the harder it will be for the entity to demonstrate that it has evaluated the adequacy of the collateral in attempting to qualify to use the practical expedient.

4.4.10 U.S. Treasury Securities and Other Highly Rated Debt Instruments

ASC 326-20

Example 8: Estimating Expected Credit Losses When Potential Default Is Greater Than Zero, but Expected Nonpayment Is Zero

55-48 This Example illustrates one way, but not the only way, an entity may estimate expected credit losses when the expectation of nonpayment is zero. This example is not intended to be only applicable to U.S. Treasury securities.

55-49 Entity J invests in U.S. Treasury securities with the intent to hold them to collect contractual cash flows to maturity. As a result, Entity J classifies its U.S. Treasury securities as held to maturity and measures the securities on an amortized cost basis.

55-50 Although U.S. Treasury securities often receive the highest credit rating by rating agencies at the end of the reporting period, Entity J’s management still believes that there is a possibility of default, even if that risk is remote. However, Entity J considers the guidance in paragraph 326-20-30-10 and concludes that the long history with no credit losses for U.S. Treasury securities (adjusted for current conditions and reasonable and supportable forecasts) indicates an expectation that nonpayment of the amortized cost basis is zero, even if the U.S. government were to technically default. Judgment is required to determine the nature, depth, and extent of the analysis required to evaluate the effect of current conditions and reasonable and supportable forecasts on the historical credit loss information, including qualitative factors. In this circumstance, Entity J notes that U.S. Treasury securities are explicitly fully guaranteed by a sovereign entity that can print its own currency and that the sovereign entity’s currency is routinely held by central banks and other major financial institutions, is used in international commerce, and commonly is viewed as a reserve currency, all of which qualitatively indicate that historical credit loss information should be minimally affected by current conditions and reasonable and supportable forecasts. Therefore, Entity J does not record expected credit losses for its U.S. Treasury securities at the end of the reporting period. The qualitative factors considered by Entity J in this Example are not an all-inclusive list of conditions that must be met in order to apply the guidance in paragraph 326-20-30-10.

ASC 326-20-30-10 states, in part, that “an entity is not required to measure expected credit losses on a financial asset . . . in which historical credit loss information adjusted for current conditions and reasonable and supportable forecasts results in an expectation that nonpayment of the [financial asset’s] amortized cost basis is zero.” On the basis of Example 8 in ASC 326-20-55-48 through 55-50, we believe that the FASB may have contemplated U.S. Treasury securities and other similar financial assets when it decided to allow an entity to recognize zero credit losses on an asset.

Keep in mind, however, that the FASB did not specifically exclude certain financial assets from the scope of the CECL model on the basis of the level of an asset’s credit risk. Rather, Example 8 illustrates that although it may be easy to conclude that the risk of nonpayment is zero on a U.S. Treasury security, the entity should still apply the CECL model to financial assets, even if the risk of loss associated with those assets is low. Consequently, the entity should apply the CECL model consistently to all of its financial assets regardless of the credit risk associated with each asset. That is, the entity should measure expected credit losses by considering all available relevant information, including details about past events, current conditions, and reasonable and supportable forecasts and their implications related to such losses.

Connecting the Dots

Zero Expected Losses

While the FASB did not explicitly exclude specific assets from the CECL model (other than by illustrating that an entity may conclude that the allowance on a U.S. Treasury security could be zero), the AICPA has published guidance indicating that an entity can expect zero losses on U.S. Treasury securities, Government National Mortgage Association (Ginnie Mae) mortgage-backed securities, and agency mortgage-backed securities (Fannie Mae and Freddie Mac). In reaching its conclusion, the AICPA compared indicators that would cause the entity to incur zero losses with indicators that would cause the entity to incur losses greater than zero. While none of the indicators are individually determinative, the comparison illustrated that, in the current economic environment, there was sufficient evidence that the entity could expect zero losses on the financial assets considered.

Footnotes

ASC 250-10-20 defines a change in accounting principle as follows:

A change from one generally accepted accounting principle to another generally accepted accounting principle when there are two or more generally accepted accounting principles that apply or when the accounting principle formerly used is no longer generally accepted. A change in the method of applying an accounting principle also is considered a change in accounting principle.

ASC 250-10-20 defines a change in accounting estimate as follows:

A change that has the effect of adjusting the carrying amount of an existing asset or liability or altering the subsequent accounting for existing or future assets or liabilities. A change in accounting estimate is a necessary consequence of the assessment, in conjunction with the periodic presentation of financial statements, of the present status and expected future benefits and obligations associated with assets and liabilities. Changes in accounting estimates result from new information. Examples of items for which estimates are necessary are uncollectible receivables, inventory obsolescence, service lives and salvage values of depreciable assets, and warranty obligations.

ASC 250-10-20 defines a change in accounting estimate effected by a change in accounting principle as follows:

A change in accounting estimate that is inseparable from the effect of a related change in accounting principle. An example of a change in estimate effected by a change in principle is a change in the method of depreciation, amortization, or depletion for long-lived, nonfinancial assets.

ASC 250-10-45-18 highlights that an entity must often use judgment to differentiate between a change in accounting principle and a change in accounting estimate and discusses certain changes that reflect a change in accounting estimate effected by a change in accounting principle.

The EIR is different for PCD assets. ASC 326 states that when determining the EIR “[f]or purchased financial assets with credit deterioration, however, to decouple interest income from credit loss recognition, the premium or discount at acquisition excludes the discount embedded in the purchase price that is attributable to the acquirer’s assessment of credit losses at the date of acquisition.”

After adoption of ASU 2022-02, this requirement would no longer be relevant because the ASU eliminated the concept of a TDR from a creditor’s accounting.

Discussed at the August 29, 2018, FASB meeting.

Although ASU 2022-02 eliminates the TDR guidance in ASC 310-40, the guidance in ASC 310-40-15-20 has been moved to ASC 310-10-50-45. As a result, a creditor would now refer to ASC 310-10-50-45 when determining whether a debtor is experiencing financial difficulty.

The Interagency Guidance was jointly issued by the Board of Governors of the Federal Reserve System, the FDIC, the National Credit Union Administration, and the OCC.

As indicated in the examples in the Interagency Guidance, an entity must use judgment and consider the specific facts and circumstances, including the nature and functionality of the underlying collateral, in determining whether a loan is expected to be repaid solely through continued operation of the collateral. In OCC Supervisory Memorandum No. 2009-7, the OCC indicated that “[r]esidential real estate loan modifications without evidence of a sustained repayment capacity or cash flows from the borrower rely on the underlying collateral as the sole source of repayment and, as such, would likely be deemed collateral-dependent upon modification.”

Other factors that can affect an entity’s estimation of expected credit losses may include estimated prepayments on the financial asset (or group of financial assets) and expected repayments that may be received from sources other than the collateral (e.g., general recourse to the borrower or an embedded credit enhancement provided by a third-party guarantor).

Confidential and Proprietary — for Use Solely by Authorized Personnel

This publication provides comprehensive guidance; however, it does not address all possible fact patterns, and the guidance is subject to change. Consult a Deloitte & Touche LLP professional regarding your specific issues and questions.