Columbia Business School Research Offers Banking World a Transformative Forward-Looking Model for Estimating Credit Losses
New Metric to Improve Analysis of Bank Loan Portfolios
NEW YORK – As the country’s financial crisis clearly demonstrated, the incurred-loss model in place to assess banks’ loan portfolios is lacking. After the crisis, the financial services industry was forced to grapple with the factors which led to massively over-valued assets and loan defaults that few saw coming. In response, the industry’s top regulator, the Financial Accounting Standards Board (FASB), made one of the biggest accounting changes in banking history and announced the current expected credit loss standard (CECL) to be instituted by 2020.
In a new study proactively confronting the problem – The Expected Rate of Credit Losses on Banks’ Loan Portfolios – Columbia Business School Professors Urooj Khan, The Class of 1967 Associate Professor of Business, Trevor Harris, the Arthur J. Samberg Professor of Professional Practice, and Doron Nissim, the Ernst & Young Professor of Accounting & Finance, introduce ExpectedRCL, a structural model of the expected rate of credit losses which has proven to outperform previous models in predicting credit losses up to one year ahead.
“It’s inherently difficult for outsiders to accurately forecast losses for banks of any size, whether they’re investors, customers, or competitors,” said Columbia Business School Professor Urooj Khan. “Opaque loan portfolios present a significant obstacle to transparency. ExpectedRCL marks a fundamental change to the financial industry, allowing anyone to use publicly disclosed documents to better estimate a bank’s future performance. It’s an important step forward to ensure trust among the industry, its stakeholders, and the public.”
Analyzing 20 years of quarterly FR Y-9C reports for bank holding companies, the researchers explain in detail how they create an estimate of expected credit losses. They develop the metric through a structural model that incorporates different accounting data, including nonperforming loans, net charge-offs, loan portfolio size, and the average balance of loans, among other measures. The researchers use cross-sectional analyses to determine coefficients for estimating each period’s measure of expected credit losses. ExpectedRCL is designed to be able to benchmark and assess the quality of the data that banks are reporting for the allowance for loan and lease losses (ALLL) and the provision for loan and lease losses (PLLL).
The metric presents a number of benefits for the financial industry.
- Outperforming Other Predictors, including Net Charge-Offs: The researchers found that ExpectedRCL is a better predictor of one-year-ahead realized credit losses than any other publicly disclosed credit-risk-related metrics and that it can serve as a benchmark to compare with public disclosures for a better understanding of future portfolio performance.
- Forecasting an Earnings Surprise: Comparing analyst PLLL forecasts with ExpectedRCL, the researchers find that banks on average experience larger earnings surprises when the differences between ExpectedRCL and analyst PLLL forecasts are larger. Similar conclusions can be drawn using a comparison with net charge-offs.
- Predicting Bank Failures: Researchers reviewed the data for 71 banks that fit the study conditions over the time period. Despite a small sample size, researchers found that ExpectedRCL is incrementally useful in predicting bank failure over the next year.
The study, The Expected Rate of Credit Losses on Banks’ Loan Portfolios, published in The Accounting Review is available online here.
To learn more about the cutting-edge research taking place at Columbia Business School, please visit www.gsb.columbia.edu.
About the researchers
Professor Harris' research and practical experience has covered most areas of the use of accounting information for valuation, investment and management decisions, with a...Read more.
Professor Nissim earned his PhD in Accounting at the University of California, Berkeley, and joined Columbia Business School in 1997. He was granted tenure in...Read more.