In addition to providing its own analyses, the Payment Cards Center also wants to assist other researchers with an interest in payments and consumer credit. For example, the Center has developed reference tools for researchers (see box) to provide information on various data sources. Industry Specialist Mark Furletti and Research Assistant Christopher Ody have also produced a series of three papers that examine publicly available data on consumer credit. One paper analyzes the different sources of data on credit card chargeoffs, reviews the methodologies of each, and compares the resulting data series. Another analyzes the sources and methodologies used to compile the Federal Reserve Board’s G.19 statistic on revolving consumer credit. Finally, they analyze the Federal Reserve Board’s credit card comparison guide, documenting many of the changes in credit card pricing that have had an impact on the guide’s usefulness to consumers and researchers.
Credit card chargeoffs are loans that are written off by card issuers as no longer collectible because they are in default. The percentage of credit card loans charged off by card issuers during a particular month or quarter is an important metric because it provides insights into the financial health of the credit card industry and the U.S. consumer. While this measure is commonly cited, few understand the nuances of how chargeoffs are calculated. In this paper, Mark Furletti examines the sampling techniques, frequency, availability, and calculation methods of five different publicly available measures of credit card chargeoffs.
While chargeoff statistics should be a measure of charged-off loans divided by all loans, a variety of nuances in the five measures can complicate analysis. Numerators differ as to whether they include or exclude recoveries on loans the bank has already charged off. Neither method is unequivocally superior. Including recoveries gives a better idea of how much banks are losing. However, losses and recoveries on one loan occur at different times, complicating the interpretation of net numbers. Denominators vary as to whether they use average daily balance, the balance on one day of the period, or an average of the bank’s balance for multiple periods. Most measures weigh chargeoffs by the size of the balance charged off, but one takes a simple average for each bank in the sample. Other methodological issues include dealing with foreign versus domestic loans and adjusting chargeoffs for mergers between issuers.
While there are differences in the sampling techniques, frequency, availability, and calculation methods of the five chargeoff measures, Furletti determines that until 2000, they all tend to move together. However, between 2000 and the paper’s publication date, the on-balance-sheet and off-balance-sheet chargeoff measures began to diverge. When these five measures moved together, understanding the detailed differences between them was less important than it is now that their movements have become less correlated.
In this paper, Mark Furletti and Christopher Ody describe the source data, sampling methods, and calculations used to compile the Federal Reserve System’s monthly estimate of revolving consumer credit as published in the G.19 statistical release. The G.19 is the most widely used and cited measure of nonmortgage consumer credit outstandings and the prevailing interest rates on certain loan types. Furletti and Ody’s paper focuses on the G.19’s measure of revolving consumer credit outstandings because of its importance to the credit card industry and credit card consumers.
One challenge the Federal Reserve System faces in compiling the G.19 is the wide variety of issuers of revolving credit. Banks, finance companies, thrifts, credit unions, and nonfinancial businesses (such as department stores) all issue credit cards. The paper catalogues the various regulatory reports that these different lenders file and how each is used for the G.19. Further complicating matters, many credit card loans are not held on an institution’s balance sheet but are instead sold in the securitization market. Some off-balance-sheet loans are not captured on regulatory forms at all. Therefore, the Federal Reserve System uses a variety of alternative sources, catalogued in the paper, to collect supplemental information on securitized lending.
While the authors conclude that the current methodology nicely balances providing timely, accurate information with avoiding unnecessary burden on lenders, they also offer several suggestions for modifications to the report that they believe will be especially useful to research analysts.
Every six months, about 150 U.S credit card issuers provide the Federal Reserve System with interest rate and fee data for their most popular credit card plan open to new customers. Pursuant to a 1988 amendment to the Truth in Lending Act (TILA), the Federal Reserve System uses these data, commonly known as the Terms of Credit Card Plan (TCCP) data, to compile a survey of credit card plans that is published on the Board’s website. The underlying TCCP data are also made publicly available to researchers, and they have been used in a variety of papers. Furletti and Ody analyze how recent changes in industry pricing practices have affected the nature of the reported data.
The authors identify three industry developments that have made it more difficult to conclude that the most common plan will represent the actual product offer most applicants will receive. This, they find, is especially true for large issuers with a vast array of pricing and other product features. First, at the account level, a wide variety of new fees and differing APRs have become common since the TCCP was created. Second, large issuers charge their customers different prices, depending on their assessment of the relative riskiness of any borrowers. Third, credit cards, as products, have become highly differentiated customized by color, reward type, co-brand partner, and affinity group with differing implications for pricing.
These relatively recent industry developments make it harder to compile and simply report information about pricing practices. First, there are significantly more pricing terms and combinations of possible pricing elements that must be taken into account. Second, as a result of risk-based pricing and an increase in the number of card products, an issuer’s most common plan becomes less and less common.
The authors conclude that these industry changes create new difficulties in designing an effective tool for consumers and good data for researchers. They argue that, in many ways, the data have remained more useful to researchers than to consumers. For consumers the authors note that the qualitative information included along with the statistical data can be a helpful guide to understanding many of the nuances inherent in contemporary credit card pricing practices.
In addition to the data paper series, the Center has built other tools to aid researchers interested in consumer payments and credit.
The Payment Cards Center maintains an online searchable bibliography of academic articles, books, reports, and publications related to consumer credit and payments.
Gathering data on the payment card industry can be difficult for researchers because there are often multiple sources for seemingly similar statistics. To address this problem, PCC staff developed a dictionary of industry statistics and, where necessary, worked with analysts at source institutions to confirm definitions and calculations.