Current Expected Credit Loss and Consumer Lending 

By | March 3, 2023

In the aftermath of the Global Financial Crisis of 2007-2008, there was a heated public debate about the causes of the crisis and what policies should be adopted to ensure that history would not repeat itself. Many people and policymakers pointed their fingers at the lax accounting standards that seemingly allowed financial institutions to take on large amounts of risk without setting aside sufficient reserves for such risks. In response, accounting standard setters around the world embarked on a years-long journey to overhaul the rules governing how banks reserve for loan losses. 

In the United States, this process culminated with a shift from an incurred-loss (IL) to a lifetime expected-loss model in determining accounting reserves for potential loan losses. Under the IL model, banks could only set aside reserves for potential losses that could be estimated precisely. In practice, that meant that banks could only reserve for short-term losses that were expected to materialize in the next 12 to 18 months. The expected-loss model represented a significant departure from this prior approach; since the adoption of the Current Expected Credit Loss (CECL) approach in the beginning of 2020, many U.S. banks and credit institutions began setting aside loan loss reserves for the full lifetime expected losses of a loan immediately after a loan was originated. This change in accounting standards was expected to have a sizable impact on the size and timing of banks’ loan loss reserves. 

The adoption of CECL was contentious. Accounting standard setters and banking regulators argued that CECL would ensure that banks create robust capital buffers against potential loan losses during good times that would mitigate the adverse macroeconomic impacts of negative shocks to asset values during bad times. But the purported benefits of CECL were met with skepticism from members of the banking industry who expressed deep concerns about its potential impact on banks’ ability to extend credit to the economy. They cautioned that CECL increases banks’ required loan loss reserves and effectively represents an increase in mandatory capital requirements that will raise banks’ cost of capital and force them to increase loan interest rates to consumers. 

In our recent paper, we examine the impact CECL had on loan pricing and access to credit. Did banks change their loan pricing in response to the new rules for computation of credit loss allowances? Did they revise their lending criteria and prices across different types of loans after adopting CECL? And what role did changes in the relative capital charges of different types of loans play in driving changes in banks’ pricing and lending decisions following the transition to CECL? 

We use data from TransUnion, a large U.S. credit bureau covering millions of individual consumer loans, to examine the transition to CECL and its effects on banks’ loan pricing and lending criteria.  These data allow us to examine not only the price and maturity of each individual loan but also which lenders underwrote those loans, the risk characteristics of the respective borrowers, and the performance of each loan over time. Importantly, a large cross-section of banks and other credit institutions report the data to the credit bureau. Thus, we are able to observe the individual characteristics of the borrowers of each loan originated by large banks that had to adopt CECL in 2020 and smaller banks that did not have to adopt the CECL standard until 2023. 

One of the main challenges in evaluating the impact of CECL is that its adoption at the beginning of 2020 coincided with the Covid-19 pandemic, which caused unprecedented uncertainty in credit markets and record levels of job losses. As a result, it is difficult to accurately measure the effects of CECL because the pandemic and subsequent monetary and fiscal policy responses could have different impacts on banks that adopted the CECL approach versus those that did not.  

We exploit an institutional feature of the implementation of CECL in the U.S. to devise a strategy that allays such concerns and allows us to better ascertain if the adoption of the CECL standard affected banks’ loan pricing and underwriting decisions. The adoption of CECL meant that bank lenders would have to set aside relatively more reserves for loans with longer maturities than they did for similar loans with shorter maturities. Figure 1 illustrates why longer-term loans require relatively more loss reserves after CECL. The figure represents the historical loan delinquencies over the lifetime of auto loans with maturities of 36 and 60 months, respectively.  

Figure 1: Aggregate auto loan amounts defaulted after x months 

Note 1: This figure shows the aggregate outstanding balances of auto loans with a default in the nth month after origination. The chart on the left represents historical defaults for auto loans with a maturity of 36 months and the chart on the right represents the historical defaults for auto loans with a maturity of 60 months.

Under the IL model, banks would initially only set aside reserves for the area shaded in blue because those are losses expected to emerge over the next 12 months. Under the CECL model, however, banks have to set aside reserves for the lifetime expected losses, which are represented by the orange and blue areas. While required loss reserves increase for both maturities, the orange area represented a larger fraction of the total lifetime expected losses of longer-term loans. Thus, longer-term loans of CECL banks became relatively more demanding than short-term loans in terms of loss reserve because relatively larger shares of their delinquencies occur after the first twelve months of the loan. 

Our empirical strategy exploits this variation in the intensity of the impact of the CECL standard across loans with different maturities. Specifically, our empirical strategy evaluates the impact of CECL on the prices and quantities of loans by comparing the change in the relative prices and quantities of long- and short-term loans originating from CECL-adopting banks with those from banks that did not adopt the CECL standard. This strategy allows us to rule out the possibility that our results are driven by events that occurred during the pandemic that affected CECL-adopting banks more than other banks. The key assumption of our empirical strategy is that external shocks that affect only specific maturities do not affect CECL and non CECL banks differently. 

Our results do not support concerns that CECL increased the price of credit and reduced credit supply. In particular, we fail to find support for a statistically or economically significant relationship between CECL adoption and loan rates. Our estimates suggest that a one standard deviation increase in the share of long-term delinquencies — which in Figure 1 is represented as the share of the orange area in the total loan defaults — is associated with a post-CECL increase in loan rates of 1.4% with an upper bound for this effect of 4.4% at a 95% confidence level. Thus, a loan with an average interest rate of 6.16% would see an increase in the interest rate of approximately 9 basis points with an upper bound of 28 basis points. Thus, our analyses rule out modest effects of the adoption of the CECL standard on the interest rates of consumer loans. We further developed this analysis by examining whether the effect of CECL might have changed over time. In Figure 2, we plot the estimated coefficients of our analysis on a two-year window before and after the adoption of CECL in January 2020. 

Figure 2: Impact of CECL on loan rates over time 

Note 2: The figure plots the coefficients and respective 95% confidence intervals of a specification examining the impact of CECL over time on the interest rates of consumer loans. 

Figure 2 suggests that around the transition to CECL, there is a slight positive uptick in the interest rates of long-term loans that received a stronger impact from CECL. This effect is, nevertheless, not statistically significant at conventional levels. Moreover, following the adoption of the CECL standard, we do not observe sustained effects of the adoption of CECL on loan interest rates in either direction both in the short and the long run.  

In further empirical analyses, we find similar results when we (i) exclude the months between February 2020 and June 2020, (ii) restrict our attention to banks that are not well capitalized, and (iii) use an alternative data set of interest rates on loans. Finally, we investigated if CECL adopting banks did not change interest rates but instead rationed the loan size that they offered to loan applicants.  

Figure 3 does not indicate that loan amounts of long-term loans relative to short-term loans were rationed when a bank adopted CECL. If anything, the results seem to suggest that the amounts of long-term loans by CECL banks increased, albeit not significantly. 

Figure 3: Impact of CECL on loan size over time 

Note 3: The figure plots the coefficients and respective 95% confidence intervals of a specification examining the impact of CECL over time on the average loan size of consumer loans.

As small community banks are preparing to adopt CECL in the first quarter of 2023, it is more relevant than ever to understand the potential impact that this new standard may have on the supply of credit. Our contribution lies in exploiting a rich and unique data set to devise an empirical strategy that levers specific institutional features to credibly estimate the impact of CECL on the price of credit. Our paper answers calls made by policymakers for in-depth studies of the consequences of CECL. Our finding that CECL did not have a significant impact on the price of credit is novel and not in line with the findings of other papers that have used other empirical strategies and data sets to evaluate how the adoption of expected-credit loss models in other jurisdictions affected the price of credit. In this sense, the paper makes an important contribution to the ongoing policy debate between standard setters and members of the financial industry around the potential effects that CECL might have on access and price of credit. 


Joao Granja is an Associate Professor of Accounting and a Jane and Basil Vasiliou Faculty Scholar at the University of Chicago Booth School of Business.  

Fabian Nagel is an Ph.D. student in Accounting at the University of Chicago Booth School of Business.  


This post was adapted from their paper, “Current Expected Credit Losses and Consumer Loans,” available on SSRN 

Leave a Reply

Your email address will not be published. Required fields are marked *