CECL 2023-1: What is Current and Expected Credit Loss and What Has Changed?

CECL Background

CECL refers to the credit impairment model provided in Accounting Standards Update (“ASU”) 2016-13, Financial Instruments — Credit Losses (Topic 326): Measurement of Credit Losses on Financial Instruments, as subsequently amended.

The ASU requires credit losses on most financial assets carried at amortized cost and certain other instruments to be measured using an expected credit loss model (referred to as the CECL model). Under this model, entities will estimate credit losses over the entire “contractual term” of the instrument (i.e., considering estimated prepayments) from the date of initial recognition of that instrument. The FASB clarified that any extension or renewal options (except those recognized as derivatives) that are not unconditionally cancellable by the entity should be considered in the contractual term.[1] The initial measurement of expected credit losses, as well as any subsequent change in the estimate of expected credit losses, is recorded as a credit loss expense (or reversal) in the current period income statement. The objective of CECL is to provide financial statement users with an estimate of the net amount the entity expects to collect on those assets.

When measuring credit losses under CECL, financial assets that share similar risk characteristics (e.g., risk rating, effective interest rate, type, size, term, geographical location, vintage, etc.) should be evaluated on a collective (pool) basis, while financial assets that do not have similar risk characteristics must be evaluated individually.[2] The ASU provides an indicative list of risk characteristics, which includes both credit and non-credit-related characteristics.[3] The ASU indicates that financial assets can be aggregated into pools based on any one or a combination of risk characteristics. However, in practice, it is expected that some credit-related characteristic would be considered. Further, the ASU does not prescribe a specific methodology for measuring the allowance for expected credit losses. 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.

However, the ASU does require that an entity base its estimate on:

  • Available and relevant internal and/or external information about past events, e.g., historical loss experience with similar assets,

  • Current conditions, and

  • Reasonable and supportable forecasts that affect the expected collectability of the reported amount of financial assets.

For periods beyond which the entity is able to make or obtain reasonable and supportable forecasts of expected credit losses, an entity should revert to historical loss information that is reflective of the contractual term of the financial asset. An entity may revert to historical loss information immediately, on a straight-line basis or using another rational and systematic basis, depending on its facts and circumstances. The reversion method is not a policy election; an entity should support the reversion methodology and period it uses to develop its estimates of expected credit losses. The expected credit loss is recorded as an allowance for credit losses, adjusted for management’s current estimate as updated at each reporting date.

In contrast, current US GAAP is based on an incurred loss model that delays recognition of credit losses until it is probable the loss has been incurred. CECL removes the threshold of “probable” and requires recognition of credit losses when such losses are “expected.” That is, even though a credit loss event may not have occurred yet, lifetime losses would still be recorded on day one (i.e., origination or purchase of the asset) under CECL based on expected future losses. A reserve is generally required even if the risk of loss is remote. Further, current practice for estimating credit losses is generally focused on the past i.e., historical loss experience and current conditions, whereas CECL also requires consideration of reasonable and supportable forecasts and, if necessary, reversion to historical loss information as mentioned earlier. In other words, CECL is based on the entire expected life of the asset. Accordingly, it is anticipated that credit losses will be both recognized earlier and for a different amount under the new CECL model than under the prior incurred loss model.

Key changes from existing guidance to the new CECL model are summarized below:

When to recognize credit losses When probable that loss has been incurred, generally subsequent to initial recognition of the asset. When losses are expected, in almost all cases upon initial recognition of the asset.
Period to consider Not an explicit input to incurred loss model Contractual term
Information to consider Historical loss and current economic conditions Historical loss, current economic conditions, reasonable and supportable forecasts about future conditions (with reversion ot historical loss information for future periods beyond those that can be reasonably forecast)
Unit of Account Pooling generally not required, but permitted Pooling required when assets share similar risk characteristics

If an asset’s risk characteristics changes such that it no longer shares similar risk characteristics with other assets in the existing pool and there is no other pool having similar characteristics, that asset should be evaluated individually. For instance, if a customer files for bankruptcy, that asset is unlikely to share similar risk characteristics as collection would now be based on that customer’s facts and circumstances.


Historical Loss Information
Segments or pools are created based on common loan characteristics. A combination of both internal and external information, including macroeconomic variables, are used to establish a relationship between historical losses and other variables.
Current Conditions
To reflect current asset-specific risk characteristics, adjustments to the historical data will need to be considered. These adjustments are usually done through a combination of both qualitative and quantitative factors.
Reasonable & Supportable Forecasts
The forecast period to project expected credit losses should be reasonable and supportable. Document the rationale and provide evidence supporting the reliability and accuracy of economic scenarios and forecasts.
Revision to History
Entities should revert to historical loss information when unable to make reasonable and supportable forecasts. The reversion method applied must be well documented and is not a policy election.
Expected Credit Loss
The results should represent the current expected credit loss over the remaining contractual term of the financial asset or group of financial assets.



[1] See ASU 2019-04, Codification Improvements to Topic 326, Financial Instruments—Credit Losses, Topic 815, Derivatives and Hedging, and Topic 825, Financial Instruments.
[2] See ASC 326-20-30-2
[3] See ASC 326-20-55-5


About the Authors

Eric D. German
Eric D. German
Partner, Assurance and Advisory


Stay up-to-date with the latest news and information delivered to your inbox.

Subscribe Now