The Profitability of Credit Card
Operations of Depository Institutions

Section 8 of the Fair Credit and Charge Card Disclosure Act of 1988 directs the Federal Reserve to transmit annually to the Congress a report about the profitability of credit card operations of depository institutions.¹ This is the eighth annual report. The analysis here is based to a great extent on information from two sources: the Consolidated Reports of Condition and Income (Call Report) and the Functional Cost Analysis (FCA).

Call Report Data
Every insured commercial bank files a Call Report each quarter with its federal supervisory agency. The Call Report provides a comprehensive balance sheet and income statement for each bank; however, it does not allocate all expenses or attribute all revenues to specific product lines, such as credit cards. Nevertheless, the data may be used to assess the profitability of those banks established primarily to issue and service credit card accounts. These specialized banks are referred to here as "credit card banks."

For purposes of this report, credit card banks are defined by two criteria: (1) the bulk of their assets are loans to individuals (consumer lending), and (2) 90 percent or more of their consumer lending involves credit cards or related plans. Given this definition, it can reasonably be assumed that the profitability of these banks primarily reflects returns from their credit card operations.

The first credit card banks were chartered in the early 1980s, and the vast majority have been in operation only since the mid-1980s. For this reason, the Call Report offers a relatively limited time series on credit card profitability. To provide a more reliable picture of the year-to-year changes in the profitability of the credit card operations of card issuers, this report limits its focus to credit card banks having at least $200 million in assets. Most of these institutions have been in continuous operation for several years, particularly those with assets exceeding $1 billion, and are well beyond the initial phase of their operations.

As of December 31, 1996, forty-two banks with assets exceeding $200 million met the definition of a credit card bank. At that time, these banks accounted for roughly 77 percent of outstanding credit card balances on the books of commercial banks and in pools underlying securities backed by credit cards. Both the dollar amount and the share of all credit card debt held by large credit card banks increased substantially from 1995. Outstanding credit card balances at large credit card banks increased about 25 percent from 1995 to 1996 and the share of all credit card debt held by these banks increased about 6 percentage points from 71 percent.

In 1996, credit card banks with assets in excess of $200 million reported net earnings before taxes of 2.14 percent of outstanding balances adjusted for credit card-backed securitization.² As table 1 shows, returns on credit card operations for these large credit card banks were lower in 1996 than in 1995.³ Returns on credit card operations in 1996 were also well below their peak level of 4.06 percent attained in 1993. Although credit card profitability has fallen substantially since 1993 for the large credit card banks, credit card earnings still compare favorably to returns on all commercial bank activities. The average return on all assets, before taxes and extraordinary items, for commercial banks in 1996 was 1.86 percent.⁴

Lower returns to credit card operations reflect factors affecting both the expense and revenue sides of the income statement.⁵ First, as in 1995, earnings at large credit card banks suffered as credit losses and provisions for future losses increased. Net charge-offs were up 20 percent as a percentage of assets from 1995 levels and provisions for loan losses were up 7 percent. Both increases reflect sharply higher delinquency rates among users of credit cards. For example, 30-day to 89-day delinquency rates for credit card plans at large credit card banks increased from 2.80 percent in 1995 to 3.27 percent in 1996. Both figures were up from the 2.48 percent delinquency rate in 1994.

Second, non-interest expenses such as labor costs and costs to process accounts increased during the year, reflecting a substantial increase in both the number of accounts and the number of credit card transactions.⁶ On the revenue side of the income statement, interest income as a percent of assets declined some from the 1995 level, as did non-interest income, perhaps reflecting the widespread use of low introductory interest rates for new accounts or transferred balances and the waiver of annual fees.⁷

1.  Net before-tax earnings as percentage of outstanding
     balances for large credit card banks
      Adjusted for credit card securitizations, 1986-1996¹

Year     Earnings2 1986     3.45 1987     3.33 1988     2.78 1989     2.83 1990     3.10 1991     2.57

Year     Earnings2 1992     3.13 1993     4.06 1994     3.98 1995     2.71 1996     2.14

    1.  Credit card banks are defined as commercial banks that
have assets greater than $200 million, have the bulk of their
assets in loans to individuals (consumer lending), and have
90 percent of their consumer lending in credit cards and
related plans. For credit card banks, outstanding balances are
adjusted to include balances underlying credit card securities.
Outstanding balances reflect an average of the four quarters
for each year.
    2.  Figures may differ from those presented in prior year
reports as the result of revisions to the Reports of Condition and
Income.
    Source.  Reports of Condition and Income, 1986-1996,
and data on securitizations.

Functional Cost Analysis Data
Initiated in the late 1950s, the FCA is a nationwide cost-accounting program conducted by the Federal Reserve Banks to provide cost, revenue, profitability and productivity data along product and functional lines for participating depository institutions. Credit card operations are one of the product lines included in the FCA program. Consequently, these data can be used to assess both the profitability of credit card operations of participating institutions over time and the profitability of credit card activities relative to the returns on other types of loans.

The FCA data have some limitations. First, the program currently includes only a relatively small number of savings and loan associations and credit unions that offer credit card programs. Consequently, the FCA data do not permit a reliable assessment of the profitability of the credit card operations of these types of depository institutions. Second, participants in the FCA program change over time as some institutions join the program and others drop out. Thus, caution is warranted in drawing conclusions from intertemporal changes in earnings. Finally, FCA participating banks are overwhelmingly small- to medium-size institutions that offer a range of banking services to their customers. Historically, few large commercial banks (those with assets in excess of $1 billion) have participated in the FCA program. Currently, none of the banks participating are among the largest credit card issuers. Thus, FCA data depict the credit card programs of banks with assets of less than $1 billion.

Earnings performance could differ between the credit card activities of FCA banks and banks with very large credit card operations if, for example, significant economies of scale are present or if larger issuers generally employ different credit standards and product pricing strategies than smaller issuers.⁸ Together, the FCA and Call Report data present a more complete picture of the profitability of credit card operations of commercial banks of all sizes.

For 1996, 63 commercial banks that participated in the FCA program submitted detailed information on their credit card activities. Among the 63 credit card issuing banks, 39 participated in the FCA program in both 1995 and 1996.

Taken together, the 63 participating banks had net losses before taxes on their credit card plans equal to 3.75 percent of credit card balances outstanding in 1996 (table 2). In 1995, the FCA participating banks had net losses of 2.93 percent. The decline in reported earnings between the two years stems primarily from an increase in expenses, as total income from credit card activities remained about the same. The increase in expenses can be traced to increased credit losses and higher data servicing costs.

To further evaluate changes in credit card profitability, an analysis was conducted of the earnings of the banks participating in both the 1995 and 1996 reporting years. Such an analysis is useful because it helps identify the extent to which changes in credit card profitability may be due to changes in the specific banks participating in the FCA program from year to year. That analysis also found that credit card earnings declined from 1995 to 1996, but to a lesser extent. Net earnings before taxes for this constant sample declined from -4.15 percent to -4.72 percent. The decline in earnings for the constant sample banks can be traced to several factors, including a fall in interchange fee and other fee income, an increase in credit losses, and higher costs for data services.

For FCA participating banks, returns on credit card plans were lower than returns on other major types of lending in 1996. Net earnings on credit card plans (-3.75 percent) were below those on real estate mortgage lending (3.06 percent), commercial lending (1.99 percent), and installment loans (0.99 percent). Historically, credit card operations have not been exceptionally profitable for FCA participating banks. From 1974 through 1996, the annual net earnings of bank credit card plans averaged 1.77 percent of balances outstanding (table 2). Over the same period, average net returns on other major types of bank lending have been somewhat higher: 2.48 percent on real estate mortgages, 2.25 percent on commercial and other loans, and 2.15 percent on installment loans.

As table 2 illustrates, FCA participants have experienced substantial year-to-year variation in credit card profitability. The standard deviation of credit card earnings (a measure of their variability) over the past 23 years has equaled 2.05 percentage points. The variability of credit card earnings over this period has exceeded that on the other types of bank lending.

2.  Net before-tax earnings as percentage of outstanding balances
     for selected types of bank credit, 1974-96¹
      Diversified banks in the Functional Cost Analysis

Year Credit card lending Real estate mortgage lending Commercial and other lending Installment lending 1974    .77 2.21 3.49 1.56 1975 1.58 2.74 2.60 2.34 1976 2.73 2.85 1.84 2.45 1977 3.09 3.18 1.86 2.75 1978 2.55 2.70 2.86 2.82 1979 1.62 2.06 4.02 2.32 1980 -1.61   1.65 4.58 1.57 1981 1.00   .73 5.38 1.69 1982 2.32   .91 3.26 2.81 1983 2.36 2.16 1.49 3.17 1984 3.42 2.10 1.95 2.81 1985 3.97 2.86 1.40 2.70 1986 3.28 2.37    .97 2.57 1987 3.38 3.05 1.34 2.31 1988 2.53 2.70 1.96 2.23 1989 1.20 2.67 2.43 2.21 1990 1.51 1.66    .79 1.92 1991 3.12 2.72 1.12 1.72 1992 2.92 3.16    .77 2.02 1993 4.22 3.20 1.35 2.14 1994 1.44 3.18 1.90 1.45 1995 -2.93   3.10 2.29 1.01 1996 -3.75   3.06 1.99   .99 Memoranda: 1974-96 average 1.77 2.48 2.25 2.15 Standard deviation 2.05   .71 1.21   .59

    1.  Net earnings rates are weighted averages for three size categories of banks reported in the National Average Commercial Bank Report, Functional Cost Analysis, annual issues, 1974-96.
    Source.  National Average Commercial Bank Report, Functional Cost Analysis, annual issues, 1974-96.

General Discussion
Generally speaking, competition in the credit card market remains intense, with thousands of firms offering bank cards to consumers.⁹ Prior to the early 1990s, card issuers competed primarily by waiving annual fees and providing credit card program enhancements. Since then, however, interest-rate competition has played a much more prominent role. Many credit card issuers, including nearly all of the largest issuers, have lowered interest rates on many of their accounts below the 18 to 19 percent levels commonly maintained through most of the 1980s and early 1990s. Credit card interest rates in general have become more responsive to issuers' costs of funds (currently, about 70 percent of the card issuers tie their interest rates directly to an index that moves with market rates). Some issuers have segmented their cardholder bases according to risk characteristics, offering reduced rates to existing customers who have good payment records while maintaining relatively high rates for higher-risk, late-paying cardholders. Moreover, many issuers have attempted to gain or maintain market share by offering very low, temporary rates on balances rolled over from competing firms. Trends in credit card pricing are discussed in more detail below.

Over the past several years, competition has led to substantial shifts in market shares among the industry's largest firms. Several of the more rapidly growing firms in recent years appear to have attracted market share by offering comparatively low-rate cards. Others have gained market share through co-branding and associated rebate strategies, typically combined with waivers of annual fees.¹⁰ Finally, some of the larger issuers have grown by acquiring credit card portfolios from smaller issuers or by merging with other firms.

Aggressive competition for new customers during 1996 was at least partly the cause of a 6 percent increase from 1995 in the number of VISA and MasterCards in circulation, to a total of 383.4 million. The number of credit cards per cardholder also increased slightly, rising to 3.9 credit cards per person, up from 3.8 in 1995.¹¹ The large number of direct mail solicitations, although off some from 1995 levels, demonstrates the continuing desire of card issuers to expand their base of cardholders.¹² This aggressive marketing of new customers, as well as the segmentation of card holders into different risk classifications, has been facilitated by the growing use of credit scoring techniques. While credit card holding continued to grow from 1995 to 1996, the rate of growth has moderated, consistent with the view that the market is becoming relatively saturated.

Recent Trends in Credit Card Pricing
Aside from questions about the profitability of credit card operations, considerable attention has been focused on credit card pricing and how it has changed in recent years. Analysis of the trends in credit card pricing in this report focuses on credit card interest rates because they are the most important component of the pricing of credit card services. Credit card pricing, however, involves other elements, including annual fees, fees for cash advances, rebates, minimum finance charges, over-the-limit fees, and late payment charges.¹³ In addition, the length of the "interest-free" grace period, if any, can have an important influence on the amount of interest consumers pay when they borrow on their credit cards.

Over the past several years, pricing practices in the credit card market have changed significantly. Many card issuers that in the past offered programs with a single interest rate now offer a broad range of card plans with differing rates depending on credit risk and consumer usage patterns. Moreover, as noted, many issuers have also moved to variable-rate pricing that ties movements in their interest rates to a specified index such as the prime rate.

At present, the Federal Reserve collects information on credit card pricing through two surveys of credit card issuers. Because of the significant changes in the pricing of credit card services, the Federal Reserve initiated the Quarterly Report of Credit Card Interest Rates (FR 2835a) at the end of 1994. This new survey collects from a sample of credit card issuers information on (1) the average nominal interest rate and (2) the average computed interest rate. The former is the simple average interest rate across all accounts; the latter is the average interest rate paid by those card users that incur finance charges. These two measures can differ because some cardholders are convenience users who pay off their balances during the interest-free grace period and therefore do not typically incur finance charges. Together, these two new interest rate series provide improved measures of credit card pricing. The Federal Reserve also collects detailed information on the pricing features of the largest credit card plan of a sample of credit card issuers through the Survey of Terms of Credit Card Plans (FR 2572).¹⁴

Because information from the FR 2835a survey does not have an extended historical interest rate series for comparison purposes, this report on credit card profitability also presents data from the survey that preceded and was replaced by the FR 2835a, the Federal Reserve's Quarterly Report of Interest Rates on Selected Direct Consumer Installment Loans (FR 2835). Data from the FR 2835 indicate that credit card interest rates fell sharply from mid-1991 through early 1994 after being relatively stable for most of the previous twenty years (table 3).¹⁵ Since early 1994, credit card interest rates have fluctuated in a narrow range between 15.25 and 16.25 percent. It should be emphasized that the interest rates reported after August 1994 are based on the new survey and are not directly comparable to the interest rates reported on the older survey.

The general decline in credit card interest rates from mid-1991 is the result of many factors, including much more pronounced competition based on this aspect of credit card pricing. The decline in rates also reflects, in large measure, the sharp drop in credit card issuers' costs of funds in the early part of this period.

Additional evidence on changes in credit card interest rates comes from the
FR 2572. Although not precisely comparable from period to period because of some changes in the sample of reporters, this statistical series reveals a general decline in credit card interest rates in recent years. For example, only 11 percent of the respondents reported interest rates below 16 percent on their largest credit card plan as of September 1991, but 50 percent did so as of January 1997. In addition, the proportion of card issuers reporting that they utilize variable-rate pricing has also increased substantially since September 1991. As of September 1991, 23 percent of issuers used variable-rate pricing; by January 1997 the proportion had increased to 70 percent. The increased use of variable-rate pricing suggests credit card rates are likely to be more responsive to changes in market interest rates in the future than they have been in the past.

3.  Average Most Common Interest Rate on Credit Card Plans,
     1972-August 1994, and the Interest Rate Assessed on Accounts
     Incurring Interest Charges, November 1994-1996¹
      Percent

Year Interest rate 1972 17.21 1973 17.21 1974 17.20 1975 17.16 1976 17.05 1977 16.88 1978 17.03 1979 17.03 1980 17.31 1981 17.78 1982 18.51 1983 18.78 1984 18.77 1985 18.69 1986 18.26 1987 17.92 1988 17.78 1989 18.02 1990 18.17 1991 18.23 1992 17.78 1993 16.83 1994 15.77 1995 15.79 1996 15.50

Month Interest rate 1991 February 18.28 May 18.22 August 18.24 November 18.19     1992 February 18.09 May 17.97 August 17.66 November 17.38     1993 February 17.26 May 17.15 August 16.59 November 16.30     1994 February 16.06 May 16.15 August 16.25 November 15.77     1995 February 15.29 May 16.23 August 15.94 November 15.71     1996 February 15.41 May 15.41 August 15.64 November 15.52

    1.  Prior to November 1994, interest rates were those reported in the Quarterly
Report of Interest Rates on Selected Direct Consumer Installment Loans
(FR 2835). Beginning in November 1994 interest rates are those reported on the
Quarterly Report of Credit Card Interest Rates (FR 2835a) for those credit
card holders incurring interest charges.
    Source.   Board of Governors of the Federal Reserve System

NOTES

     1   P.L. 100-583, 102 Stat. 2960 (1988).

     2   Calculations are adjusted for securitizations because earnings as reported on the Call Report reflect revenues and expenses from outstandings both on the books of the institutions and in off-balance-sheet pools backing securities.

     3   One problem that arises in assessing changes in profitability over time is that the sample of credit card banks may change somewhat from one year to the next. Thus, overall changes in profit rates from year to year reflect both real changes in activity and changes in the sample. To evaluate the effects of sample changes, the profitability of the specific banks included in the 1996 sample were examined over the period from 1986 to 1996 as well. Although the level of reported profitability is somewhat different from that shown in table 1, for such a consistent panel of banks, the intertemporal pattern of profitability remains essentially as shown in the table.

     4   See William R. Nelson and Ann L. Owen, "Profits and Balance Sheet Developments at U.S. Commercial Banks in 1996," Federal Reserve Bulletin, vol. 83, no. 6 (June 1997), pp. 465-489.

     5   See, James J. Daly, "Down for the Count," Credit Card Management,
May 1997, pp. 46-47.

     6   It is estimated that the number of credit card accounts at banking institutions increased 7 percent and the number of transactions increased 12 percent from 1995 levels. Source. The Nilson Report, March 1997.

     7   For example, the proportion of credit card issuers reporting they do not assess an annual fee on their largest credit card plan increased from 47 percent on January 31, 1996 to 54 percent as of January 31, 1997. Source. "Terms of Credit Card Plans," Board of Governors of the Federal Reserve System.

     8   For a more complete discussion of the differences in cost structures and pricing behavior of large and small credit card issuers, see Glenn B. Canner and Charles A. Luckett, "Developments in the Pricing of Credit Card Services," Federal Reserve Bulletin, vol. 78, no. 9 (September 1992), pp. 652-666.

     9   Currently, roughly 6,800 depository institutions issue VISA and MasterCard credit cards and independently set the terms and conditions on their plans. Close to 10,000 other institutions act as agents for card-issuing institutions. In addition to the firms issuing cards through the VISA and MasterCard networks, two large nonbank firms, American Express and Dean Witter, issue independent general purpose credit cards to the public.

     10   Under co-branding programs, the credit card bears the name of and is marketed to consumers of the co-branded product(s). Through use of the card, consumers typically accumulate "points" good for rebates on purchases of the co-branded product(s). One popular type of co-branding is with airline companies; in this case, "frequent-flier miles" are earned through credit card purchases.

     11   Figures exclude debit cards. Source. The Nilson Report, March 1997.

     12   An estimated 2.5 billion direct mail solicitations were sent by issuers during 1996, off from the historical high level of 2.8 billion reached in 1995. The response rate on credit card solicitations in 1996 was estimated to be 1.8 percent, up from 1 percent in 1995 but down from an average of 2.5 percent over the period 1990 to 1993. Source. James J. Daly, "Saving on Postage," Credit Card Management, May 1997, pp. 68-71.

     13   In June 1996, the Supreme Court ruled that states may not regulate the fees charged by out-of-state credit card issuers. States have not been permitted to regulate the interest rates out-of-state banks charge. In making its decision, the Court supported the position previously adopted by the Comptroller of the Currency that a wide variety of bank charges, such as late fees, membership fees, and over-the-limit fees, are to be considered interest payments. This ruling will likely ensure that banks will continue to price credit cards in multidimensional ways rather than pricing exclusively through interest rates. Source. Valerie Block, "Supreme Court Upholds Nationwide Card Charges," American Banker, June 4, 1996.

     14   The information in the FR 2572 survey is published twice a year by the Federal Reserve. Historically, the data were made available in a statistical release, the E.5 "Report of the Terms of Credit Card Plans." Beginning in 1995, the E.5 statistical release was discontinued and data are now included in a consumer brochure, entitled "Shop: The Card You Pick Can Save You Money." This brochure is available on the Internet at http://www.bog.frb.fed.us/pubs/shop.

     15   For a comprehensive discussion of the factors that account for the levels and changes in credit card interest rates see, Glenn B. Canner and Charles A. Luckett, "Developments in the Pricing of Credit Card Services"; also U.S. General Accounting Office, U.S. Credit Card Industry (GAO/GGD-94-23, 1994).