During 1995, the credit card charge-off ratio at banks rose sharply, almost a full percentage point. This surprised some observers, since 1995 was the fourth year of economic expansion in the U.S.
- What’s usual about the current situation?
- What’s unusual about the current situation?
During 1995, the credit card charge-off ratio at banks rose sharply, almost a full percentage point. This surprised some observers, since 1995 was the fourth year of economic expansion in the U.S. A comparable year-to-year increase in the charge-off ratio had not been seen since the 1990-1991 recession, and though the level in 1995 was below the peak reached in the early 1990s, it still was near the highs of the previous two decades. In this Economic Letter, I point out what was usual and what was unusual about the situation by discussing some of the recurring and special factors that likely are influencing the current credit card charge-off ratio.
Some observers have suggested that the recent deterioration in credit card loan quality is unusual and due mainly to special circumstances. For example, commentators have suggested that it is due to banks taking on too much risk in the face of increasing competition from nonbanks and the need to maintain growth. They point to the aggressive marketing of credit cards as a manifestation of this trend and reason that it has led inevitably to less creditworthy customers and higher charge-off rates.
While such explanations may play a role, they overlook the importance of cyclical influences tied to the economy at large and, in so doing, miss the fact that we have seen increases in the credit card charge-off ratio during economic upturns before.
This is evident from Figure 1. The charge-off ratio is the percentage of credit card loans outstanding that banks have, in an accounting, as well as a colloquial, sense, “written off.” In the figure, two series of credit card charge-off ratios are graphed together to cover data from early 1971 through the end of 1995 (the shaded bars mark recessions). The series that begins in 1971 is from a paper by economist Lawrence Ausubel (1995), while the series that begins in 1984 is based on data from the bank regulatory Report of Condition and Income. Both series are annualized charge-off ratios for the previous quarter. Although the data sources are different, the two series are quite close.
Figure 1 shows that the credit card charge-off ratio goes through ups and downs, much like the business cycle. However, the timing of upturns in the charge-off ratio, including the most recent one, is not closely tied to recessions. Rather, the 1995 deterioration in credit card loan quality appears to fit a general pattern wherein, soon after recessions, the charge-off ratio declines for a time before embarking on a sustained upward trend that starts before the next recession and is maintained throughout the recessionary period. Note also that the time lag between troughs in the charge-off ratio and the arrival of recessions is quite variable. The trough in the charge-off ratio at the beginning of 1973 came only three quarters before the 1974-1975 recession arrived, whereas the trough at the end of 1983 came 26 quarters before the onset of the 1990-1991 recession. In addition, the charge-off ratio did not increase steadily between 1983 and 1990, but dipped between 1987 and 1989. So, the relationship between the pattern of credit card charge-offs and the pattern of recessions is somewhat loose.
As Figure 2 depicts, the pattern of credit card charge-offs shows a much tighter relationship to the pattern of employment growth. Figure 2 shows the credit card charge-off ratio along with the year-over-year payroll employment growth rate. In general, when the credit card charge-off ratio is decreasing, the year-over-year payroll employment growth is increasing, and when the credit card charge-off ratio is increasing, employment growth is decreasing. In keeping with the pattern, the nearly 1 percentage point increase in the charge-off rate in 1995 coincided with a decrease in employment growth of more than 1.5 percentage points.
Moreover, the timing of the turning points in the two series is quite close, that is, peaks in one series are close to troughs in the other. In fact, during the period shown, the turning point in one series is always within three quarters of the turning point in the other series. Most recently, employment growth peaked in the third quarter of 1994, and the credit card charge-off ratio bottomed only one quarter later.
The link between employment growth and the credit card charge-off ratio makes sense: When economic conditions deteriorate, employment growth should slow while charge-offs should increase, as more and more people fall behind on their credit card payments. On the other hand, when conditions improve, employment growth should increase, while credit card charge-offs should decrease, as more and more people regain financial stability. One specific link may be through small businesses. For example, it has been reported that many small businesses use credit cards instead of regular business loans to obtain start-up capital and operating funds, and small businesses are an important source of employment growth. So, for example, those periods when employment growth is increasing and credit card charge-offs are falling also may correspond to periods when small business conditions are improving.
No matter what lies behind the close relationship between credit card loan quality and employment growth, it is apparent that credit card loan quality is linked to cyclical influences that also underlie trends in the economy at large. Banks know this, of course, and their willingness to make consumer installment loans, which include credit card loans as well as automobile and other personal loans, is very cyclical. Regular surveys of the senior loan officers of some of the nation’s largest banks show that, on the whole, banks are least willing to make consumer loans during recessions and most willing immediately following recessions. Since about mid-1994, soon before the credit card charge-off ratio turned up, banks have become somewhat less willing to make consumer loans.
In one respect, recent charge-off rates are unusual. In no other expansionary period during the past 25 years have credit card charge-off ratios been so high while in an increasing portion of their cycle. This is not due to a particularly fast rate of increase recently, nor to an increase over a longer period of time. Rather, it is a manifestation of the long-run secular (as opposed to cyclical) increase in the charge-off ratio that began around the mid-1980s. This secular increase meant that the recent cyclical upturn came on top of an already high level of charge-offs.
An aggressive marketing push for credit cards since the mid-1980s may help to explain the long-run secular increase in credit card charge-offs over roughly the same period. Figure 3 shows the total number of the major credit and debit cards between 1979 and 1994. (When a debit card is used for payment, funds are deducted from the payer’s bank account. There are many fewer debit cards than credit cards, and, although debit cards have grown, they have not had a significant influence on the pattern of total credit and debit card growth.) Growth accelerated sharply around 1984 and again in 1993 and 1994. It is not unreasonable to suppose that each increase in growth put credit cards in the hands of an increasingly high proportion of customers who were more likely to default, thereby contributing to the long-run secular increase in the charge-off ratio. In addition to aggressive marketing, a change in the bankruptcy law in 1984 also may have contributed to a secular increase in the credit card charge-off ratio. The revision stipulated that private employers could not discriminate against individuals who were or had been bankrupt.
With banks less willing to make consumer loans since about mid-1994, the growth of credit cards declined slightly in the first half of 1995 (not shown). (Data for the second half of 1995 are not yet available.) However, banks still are maintaining a relatively aggressive credit card marketing stance and may continue to do so. Consequently, we may continue to see relatively high credit card charge-off ratios compared to past cycles. This would be consistent with research by Furlong and Levonian (1995), which indicates an upward shift in the riskiness of banks’ portfolio of assets in the 1990s. Furlong and Levonian suggest that this shift may be relatively long-lasting. It is important to note, though, as Furlong and Levonian point out, that banks’ increases in portfolio risk have been matched by significant increases in capitalization, which serves as a buffer against loan losses.
Recently, the credit card loan charge-off ratio has been high and rising. In this Economic Letter, I point out that an increase in the credit card charge-off ratio is not unusual at this stage of the business cycle. The charge-off ratio appears to be closely linked to the economy, and, in particular, to employment growth. Employment growth has been slowing recently, and, accordingly, the charge-off ratio has been rising. On the other hand, the recent increase in the credit card charge-off ratio came on top of an unusually high level of charge-offs that was the result of a long-run secular increase. It is possible that aggressive credit card marketing since about the mid-1980s may have contributed to this secular increase in the credit card charge-off ratio. Thus, both recurring, cyclical influences and special factors seem to be playing a role in the current credit card loan quality situation.
Ausubel, L. 1995. “The Credit Card Market Revisited.” University of Maryland. Unpublished manuscript.
Furlong, F., and M. Levonian. 1995. “Reduced Deposit Insurance Risk.” Federal Reserve Bank of San Francisco Weekly Letter95-08.
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Anita Todd and Karen Barnes. Permission to reprint portions of articles or whole articles must be obtained in writing. Please send editorial comments and requests for reprint permission to firstname.lastname@example.org