A PDF version of this publication is attached here: Not-for-profits and the Current Expected Credit Loss (CECL) model (PDF 393.15kb)

ASC 326-20-15-2 and ASC 326-20-15-3 provide detailed information on the scope of the CECL guidance, which is summarized in the table below.

Figure 1-1
Scope of CECL

Contributions (pledges) receivable are specifically excluded from the CECL standard. This would include any research receivables that are accounted for under the ASC 958-605 contribution model, which will likely include most federal research awards.

Although available-for-sale (AFS) debt securities are also outside the scope of the CECL guidance, the impairment model for these financial assets has been modified in connection with the issuance of the CECL guidance and codified in ASC 326-30. The new model requires an entity to use a discounted cash flow approach when estimating expected credit losses. See Section 5 for further details on the new impairment model.

Reporting entities should record the expected lifetime credit losses for financial instruments within the scope of CECL through the allowance for credit losses account. As a result, the financial statements will generally reflect the net amount expected to be collected on recognized assets. The allowance is measured and recorded upon the initial recognition of the in-scope financial instrument, regardless of whether it is originated, purchased, or acquired in a business combination.s

Prior to the adoption of CECL, credit loss expense (or what was commonly referred to as “bad debt expense”) was recognized only when it was probable that a credit loss had been incurred. Under the CECL model, an allowance is recorded at the point of initial recognition of the asset for all expected losses over the life of the asset.

The CECL model requires that an entity's estimate of expected credit losses includes a measure of the expected risk of credit loss even if that risk is remote. Figure 2-1 provides a comparison of the current expected credit loss model under ASC 326 to the previous incurred loss model.

Figure 2-1
Comparison of CECL to previous incurred loss model

ASC 326-20-30-5 requires a reporting entity to determine the allowance for credit losses based on the amortized cost of the financial asset. ASC 326-20-20 defines the amortized cost basis.

ASC 326-20-30-2 requires a reporting entity to measure expected credit losses on a collective (pool) basis when similar risk characteristics exist across multiple financial instruments. These pools are also sometimes referred to as portfolio segments. If a financial instrument does not share similar risk characteristics with other assets subject to CECL, expected credit losses may be measured at the individual asset level. ASC 326-20-55-5 provides a list of risk characteristics that may be used to pool assets. The pools established are not static and should be reassessed each reporting period.

CECL does not prescribe a specific method to calculate the allowance for credit losses. The selection of a method will depend on the reporting entity’s facts and circumstances, including the complexity and significance of the financial instruments being evaluated, as well as other relevant considerations.

All entities, including NFPs, may continue to determine the allowance using their existing methodologies, such as aging schedules (also sometimes referred to as provision matrices) or a loss-rate method. However, due to CECL, the inputs into that methodology, the factors considered, and additional qualitative adjustments will likely change as the principles of what constitutes a credit loss have changed significantly.

A reporting entity should develop an estimate of credit losses based on historical information, current conditions, and reasonable and supportable forecasts. It may begin the process of measuring expected credit losses by analyzing its historical loss experience for financial assets with risk characteristics similar to the assets being measured. However, as discussed in ASC 326-20-30-8, this information must be adjusted, as necessary, to reflect the extent to which management expects current conditions and reasonable and supportable forecasts to differ from the conditions that existed for the period over which historical information was evaluated and due to differences in the composition of the current portfolio. In evaluating conditions that may merit an adjustment to the historical data used to measure expected credit losses, a reporting entity should consider the risk factors relevant to the assets being measured. These may include data that are specific to the borrower, specific to a group of pooled assets, relevant to all assets at a macro-economic level, or some combination.

If an entity’s quantitative models and historical data do not reflect current conditions or an entity’s reasonable and supportable forecasts, these factors should be included through qualitative adjustments such that the estimate in total is reasonable. CECL requires an entity to estimate and recognize an allowance for credit losses for a financial instrument, even when the expected risk of credit loss is remote.

ASC 326-20 requires an entity to estimate lifetime expected credit losses. In doing so, a reporting entity is not required to develop economic forecasts over the asset’s life if such estimates are not reasonable and supportable, but instead, the entity may use a combination of economic forecasts and reversion to historical loss experience in arriving at its estimate. ASC 326-20-30-7 requires a reporting entity to evaluate both internally-generated data and reasonably-accessible external data to estimate credit losses. However, ASC 326-20-30-7 also states that a reporting entity may determine that using its internally generated data is sufficient. The estimate should consider all relevant data that is reasonably available to an entity at the balance sheet date without undue cost and effort.

No. CECL requires an entity to use historical data adjusted for current conditions and reasonable and supportable forecasts to estimate expected credit losses over the life of an instrument. Only for the period beyond which an entity is able to develop a reasonable and supportable forecast can an entity revert to unadjusted historical loss information. While some entities may be able to develop reasonable and supportable forecasts for longer periods than other entities, it is not acceptable for an entity to assert it cannot develop a forecast and use only historical loss information.

At each reporting period, a reporting entity should update its estimate of lifetime expected credit losses and adjust the allowance accordingly. Increases and decreases are both classified as credit loss expense.

ASC 326-20-35-8 requires reporting entities to write off individual financial assets (or a portion thereof) in the period in which a determination is made that the financial asset (or portion) is uncollectible. This generally occurs when all commercially reasonable means of recovering the balance have been exhausted. Factors a reporting entity may consider include (1) significant changes in the borrower’s financial position such that they can no longer pay the obligation or (2) an assessment that the proceeds from collateral will not be sufficient to repay the loan. However, the term “uncollectible” is not defined and continues to require the application of judgment.

ASC 326-20-30-10 does not require a reporting entity to measure expected credit losses on an instrument, or a pool of instruments, if historical information adjusted for current conditions and reasonable and supportable forecasts result in zero expected credit losses in all scenarios. In these situations, the risk of default may be greater than zero, but the amount of the expected loss is zero.

ASC 326-20-55-49 and ASC 326-20-55-50 use a US Treasury security as an example of an asset that may result in an expectation that the risk of non-payment of the amortized cost basis (i.e., the loss given default) is zero because they are explicitly guaranteed by a high-quality sovereign entity and have a long history of no credit losses. However, the FASB did not provide a list of other assets that may qualify for zero expected credit losses.

Disclosures related to credit losses are intended to enable users of financial statements to understand the credit risk inherent in a reporting entity’s portfolio, how management monitors this risk, management’s estimate of expected credit losses, and the changes in the estimate during the period.

Required disclosures include a description of management’s policy and methodology for developing the allowance for credit losses, a rollforward of the balance, an aging analysis of past due loans, and credit quality indicators. The past due loans and credit quality indicator disclosures are not required for trade receivables due in less than one year. Example 3-1 provides sample policy and rollforward disclosures for trade receivables.

As discussed in ASC 326-20-50-3, each reporting entity needs to determine how much detail to provide based on its specific facts and circumstances. The disclosure should strike a balance between (1) too much aggregation of non-homogenous asset classes and (2) excessive disaggregation and details that could obscure material information.

CECL is effective for all not-for-profit entities for fiscal years beginning after December 15, 2022. For entities that have not yet adopted CECL, the effective dates for standards that modify or clarify CECL (e.g., ASU 2019-04, ASU 2019-05, and ASU 2022-02) are the same as the effective date for CECL. For entities that have early adopted CECL, ASU 2022-02 is effective for fiscal years beginning after December 15, 2022.

The CECL guidance in ASC 326-20 should be applied to financial instruments that are in scope (other than certain debt securities) using a modified retrospective approach. A not-for-profit entity will be required to recognize the cumulative effect of initially applying the impairment standard as an adjustment to opening net assets in the period of initial application. The cumulative effect adjustment will represent the difference between today’s model and the CECL impairment model.

The transition guidance for debt securities can be complex, and depends on whether a security reported an other-than-temporary impairment prior to the adoption of CECL. Section 13.3.1 of the Loans and Investments guide details other considerations on the transition for debt securities.

A trade receivable arising from revenue recognized under ASC 606 is likely one of the more common asset classes to which the CECL model will apply for entities outside of the financial services sector. The allowance for credit losses should be applied to the trade receivables balance, including contract assets, determined after application of the guidance in ASC 606 for estimating variable consideration, including both explicit and implicit price concessions. The initial amortized cost basis for a receivable or contract asset is the amount recognized under ASC 606. This concept is especially relevant for patient accounts receivable held by not-for-profit healthcare providers, as discussed later in this section.

For the purposes of pooling trade receivables with similar risk characteristics, factors to consider may include customer credit rating, receivable aging (e.g., 30, 60, 90 days past due), customer industry, geographical location of the customer, product line, and other factors that may influence the likelihood of the customer not being able to pay for the goods or services.

While a provision matrix (e.g., a loss rate applied to each receivable aging bucket, as illustrated in Example 3-1) may still be used to develop a credit loss allowance under the CECL guidance, the historical loss data from such models will need to be combined with an evaluation of current conditions and reasonable and supportable forecasts of future losses to determine estimated credit losses. Because of this, even receivables that are not yet past due will have an allowance for expected credit losses.

NFP University (the University) will adopt the CECL guidance in its financial statements for the fiscal year ending June 30, 20X4. The University bills tuition at the beginning of each term, with balances typically due in full within 30 days. For purposes of this example, installment plans and other tuition payment arrangements that the University may offer will not be considered.

In accordance with ASC 326-20-30-2, the University determined that similar risk characteristics existed for receivables with similar days outstanding, so that the pools for which the allowance for credit losses was calculated were based on aging buckets. This is consistent with their methodology under the previous incurred loss model.

The University compiles historical loss information for its student receivables annually and updates its credit loss percentages based on the previous 5-year average writeoff history to calculate loss rates using a “roll-rate” methodology. Under a roll-rate methodology, the historical loss rates for each aging bucket take into account the migration of each balance through the various aging buckets to determine the appropriate credit loss rate for each aging bucket. The University considered other available external data and management’s reasonable and supportable forecasts of current and future economic conditions over the life of the receivables and adjusted the historical loss information.

Management determined that historic loss rates will increase by 10% over the next fiscal year due to rising inflation and indicators of a potential broader economic recession. Management’s five-year historical loss rate information does not include periods of rising inflation or recession. Management also considered the need for any additional qualitative adjustments and concluded that their adjustment to the inputs to the quantitative model was sufficient.

Management applies these forecasts to create the following credit loss matrix to calculate the allowance for credit losses at June 30, 20X4:

To determine the impact to opening net assets as of July 1, 20X3, management performed a similar analysis based on the receivables balance as of June 30, 20X3. Under the previously-applied incurred loss model, the University did not record an allowance against receivables that were current (i.e., not past due). Management determined that the cumulative effect of adopting CECL at July 1, 20X3 was $35,843 (the calculation of which is not shown). This transition impact would be reported as an adjustment to opening net assets as of July 1, 20X3 in the financial statements for the year ended June 30, 20X4.

ASC 326-20-50-10 and ASC 326-20-50-11 require disclosure of management’s policy and methodology for developing the estimate of the allowance for credit losses. In this example, a sample narrative disclosure could be:

ASC 326-20-50-13 requires a rollforward of the allowance for credit losses. In the year of adoption, the rollforward is required to include the impact of adoption recognized as a cumulative effect adjustment as of the beginning of the period. The rollforward is not required for the prior comparative period in the year of adoption, but is required for each period for which a statement of financial position is presented in subsequent periods. The rollforward disclosure in this example would be:

There may be complexity in applying the CECL model to patient accounts receivable. In determining the transaction price under ASC 606, health care entities often estimate explicit and implicit price concessions. Implicit price concessions differ from explicit price concessions in that they (1) typically are not specific to individual patients and (2) are not contractually specified or determinable at the inception of the contract. The determination of the transaction price should take into account payments the provider expects to receive from third-party payers (e.g., private health insurance plans, Medicare, Medicaid). There are a number of factors that can affect the amount for which those parties are responsible. Healthcare providers will estimate those amounts at the time services are delivered and may include explicit, or contractual, price concessions. The portion to which the healthcare provider is entitled from the patient (either through deductibles and coinsurance, or because patients are uninsured) can vary significantly and, even once determinable – after consideration of the patient’s access to insurance – often carries a high risk of uncollectibility. In addition, given the nature of healthcare and certain regulatory requirements that mandate the provision of care in some cases, providers often perform services without knowing whether patients will be able to pay. Thus, to determine the transaction price pursuant to the principle in ASC 606 of “the amount of consideration to which [the provider] expects to be entitled in exchange for transferring promised goods or services,” amounts that are not expected to be collected from patients are considered implicit price concessions and will reduce the transaction price “up front.” While, in theory, the nonpayment of some of these amounts may be due to deteriorations in patients’ ability to pay after services are delivered, in many cases, the noncollection is simply a function of the manner in which healthcare services are delivered and the complex web of payer responsibilities.

In developing the estimate of implicit price concessions, the provider should consider its expectations of eventual cash collections from all sources based on all information (historical, current, and forecast) that is reasonably available to the provider. Once the entity has determined its best estimate of the transaction price under ASC 606, if there are any incremental amounts that would be recognized as credit losses under the CECL model, an additional allowance for credit losses may be required.

The distinction between an implicit price concession and a credit loss is important because it affects the timing of recognition and classification in the income statement. Under ASC 606, subsequent changes to estimates of implicit price concessions are accounted for as increases or decreases in the transaction price for the revenue transaction and are recorded as an adjustment to revenue. However, a credit loss recognized pursuant to ASC 326 on a receivable is an expense, and a decrease to expected credit losses is recorded as a reduction of credit loss expense. The key consideration in making this distinction is whether the adjustment is due to a change in the amount the provider was willing to accept in exchange for services provided (implicit price concession) or if the adjustment is an amount to which the provider believed they were entitled, but ultimately were unable to collect due to a credit loss. Providers will need to apply judgment to determine what is a credit loss and what is an additional implicit price concession. This judgment is no different than that required prior to the adoption of the CECL model for distinguishing between implicit price concessions under ASC 606 and the incurred loss model for bad debts. Based on our experience, the vast majority of amounts ultimately not collected for patient services have been classified as implicit price concessions.

Yes. While the credit loss considerations in ASC 326 apply to all receivables, providers should apply judgment in determining if subsequent adjustments to a patient account are due to changes in implicit price concession estimates under ASC 606 or credit losses under ASC 326.

Some NFPs provide loans as part of their operations. For example, some colleges and universities provide uncollateralized loans to students as part of an institutional student loan program or loans to faculty and staff for the purchase of a primary residence. Loans generally differ from tuition receivables, in that they accrue interest and are for a longer duration. In addition to the basic measurement principles outlined in Section 2, there are additional considerations for loans receivable. These include:

ASC 326-20-30-5 allows a reporting entity to measure the estimate of expected credit losses by measuring components of the amortized cost basis on a combined basis or by separately measuring the credit loss for the following components of the amortized cost basis:

When a reporting entity assesses a financial asset for expected credit losses, it should consider how any accrued interest could be affected by its expectation of future defaults. Some approaches (such as a discounted cash flow approach) might consider accrued interest as part of the calculation. An entity can either assess accrued interest separately from the other components of a loan’s amortized cost basis, and, if assessed separately, can elect to not measure an allowance for credit losses if uncollectible accrued interest is written off in a timely manner. ASC 326-20-30-5A provides guidance on when it would be appropriate to exclude accrued interest from the calculation of the allowance for credit losses. The writeoff of accrued interest is subject to a further accounting policy election in ASC 326-20-35-8A, which allows entities to write off accrued interest receivables by reversing interest income or by recognizing a credit loss expense. For either election, the timing of when to write off accrued interest will be a matter of judgment.

For example, consider the example of a university that does not consider accrued interest in its allowance for credit losses. It stops accruing interest on loans to faculty and staff for the purchase of primary residences that are over 90 days past due, and at that time, writes off all previously accrued interest. The writeoff of accrued interest is recorded as a credit loss expense.

There are five distinct decision points (some of which are accounting policy elections) embedded within this scenario:

Conclusions on these five decision points could lead to a different policy, and therefore, a different accounting presentation that is also acceptable under the CECL model if it fits the facts and circumstances of the loan portfolio.

The length of the forecast period of a loan is informed by its life (e.g., 5, 10 or 30 years), adjusted for expectations of prepayments. When forecasting credit losses for student loans that have payments due over a number of years, management should evaluate how long it expects current and future economic conditions to impact data relating to historical repayment rates, by risk pool. CECL does not include prescriptive guidance on the length of the reasonable and supportable forecast period or how this should be developed, however, a reporting entity is required to support its selection of the forecast period (as well as its expected credit losses estimate in its entirety). The length of the period is judgmental and should be based in part on the availability of data on which to base a forecast of economic conditions and credit losses. The process should be applied consistently and in a systematic manner. Changes in factors such as macroeconomic conditions could cause the time horizon for reasonable and supportable forecasts to change.

Management’s forecast of estimated credit losses may cover the entire life of the loan and may include some element of reversion to historic means as part of a forecast. However, if management’s reasonable and supportable forecast does not cover the entire life of the loan (after considering the impact of prepayments), then management should apply the guidance in ASC 326 for reverting to historical loss levels. ASC 326-20-30-9 provides entities with flexibility in selecting a reversion methodology for periods beyond the reasonable and supportable forecast. Reversion methods include: immediately reverting to unadjusted historical loss experience, the use of a straight-line basis for the remaining period beyond the reasonable and supportable forecast period, or another rational and systematic basis. Management should evaluate whether the historical data is relevant to the current characteristics of the risk pool and does not require adjustment based on other factors, such as changes in the contractual terms of the loan product. Entities should evaluate whether they can support that the reversion method selected is systematic and rational. The reversion technique should be evaluated in conjunction with all other judgments made in the entity’s estimate and in the context of the estimate as a whole.

The development of estimated credit losses (including reasonable and supportable forecasts) for loan portfolios can be complex. Management must apply significant judgment to determine the appropriate level of disaggregation of its loan portfolio into loan pools with similar risk characteristics.

For example, a starting point might be the various loan products offered (e.g., undergraduate loans, loans to parents, loans to students at various graduate schools such as the law school or the school of medicine, or loans to faculty). Some loan products may require nominal payments while in school, whereas others defer repayment entirely until after graduation. Another relevant starting point is whether the loan is secured by collateral (e.g., a residential mortgage loan to a faculty member) and the remaining term of the loan.

Management must then consider what other factor(s), which typically include borrower-specific factors, have an impact on the estimate of the allowance for credit losses. For student loans, the loan status (e.g., current, past due, greater than 90 days past due) often represents an informative credit quality indicator. Other potentially relevant factors for a student loan portfolio could include employment status of the student, the student’s FICO score, or whether a parent co-signed the loan. Once the model for estimating credit losses has been developed, management then would consider which of the factor(s) in the analysis represent the key credit quality indicators for disclosure in accordance with ASC 326-20-50-4 and ASC 326-20-50-5.

The development of the estimate of credit losses necessarily involves some level of forecasting, in particular for loans with a longer repayment horizon. This is a key difference between the CECL model and the previous incurred loss model under US GAAP. Thus, the factors used by management to estimate expected credit losses under CECL may be different from those used under previous GAAP. As a result, entities may need to track data on loans at a more granular level than in the past. Management should build processes, procedures, and controls to maintain historic data at a detailed enough level to support estimates of credit losses under CECL. Entities may need to obtain relevant third-party data to support their estimates and supplement their internal data.

Although collateral mitigates the economic impact of credit losses, the existence of collateral, in and of itself, does not support an assumption of zero loss of the amortized cost basis, even when the current fair value of the collateral exceeds the amortized cost of the financial asset. This is because the collateral value may decline in the future, exposing the lender to losses in the event of default by the borrower. In addition, the collateral may be illiquid, such as real estate, automobiles, business inventory, equipment, or other assets.

ASC 326-20-35-5 permits a reporting entity to elect a practical expedient for collateral-dependent assets, whereby estimated credit losses are based on the fair value of the collateral (less costs to sell, if applicable). A financial asset is considered collateral-dependent if:

If it is probable that a reporting entity will foreclose on the collateral, use of the fair value of the collateral to calculate the allowance for credit loss is required.

Available-for-sale (AFS) debt securities are not within the scope of the CECL model. ASC 326-30 provides a different impairment model for AFS debt securities, which is a modified version of the model prescribed by prior GAAP.

Under ASC 326-30, an AFS debt security is considered impaired if its fair value is less than its amortized cost basis.

A reporting entity must consider whether it intends to sell, or more likely than not will be required to sell, an impaired AFS debt security before recovery of its amortized cost basis. An entity may be required to sell through legal, regulatory or operational compulsion. If a NFP entity determines that it intends to or more likely than not will be required to sell an impaired AFS debt security, the total impairment loss should be recognized as an adjustment to the amortized cost basis of the investment and reflected in the change in net assets within the performance indicator (if one is presented). That new amortized cost basis should not be adjusted for subsequent recoveries in fair value.

If a reporting entity does not have the intent to sell (and it is not more likely than not that it will be required to sell), the portion of the impairment related to credit losses, if any, must be calculated based on the present value of the expected contractual cash flows to be collected. If the present value of cash flows expected to be collected is less than the security’s amortized cost, a credit-related impairment exists. The difference between the present value of expected cash flows and amortized cost is recorded as an allowance for credit losses. The total amount of the impairment is capped at the excess of the amortized cost over the fair value because an entity could sell the security to recover the fair value. Recoveries of credit losses can be reversed in subsequent periods, either through the reversal of an allowance or application of the recoveries guidance in ASC 326-20-35-8 and ASC 326-20-35-9.

For healthcare entities reporting under ASC 954 and other NFPs voluntarily reporting a performance indicator, non-credit related impairments are reported outside the performance indicator. Credit-related impairments are recorded within the performance indicator.

The most common example of a non-credit related impairment is a change in the market interest rate for a debt security. A decrease in fair value solely related to a change in the market interest rate for a security would be considered an unrealized loss and recorded outside the performance indicator.