[Home] [Papers] [Courses] [Contact]

Jamal Munshi, Sonoma State Univesity
All rights reserved

The nature of market forces and competiton

The heartless 19th century Darwinian view of capitalism described by Herbert Spencer, that undesirable social effects constituted the price of progress, has given way to regulatory mechanisms designed to mitigate them and to build a kinder gentler capitalism. The Sherman Anti-Trust Act of 1890, the formation of the SEC, the FDIC, and the Federal Reserve and enactment of the slew of laws following the 1930s depression to regulate the financial industry were intended to moderate volatility and attrition problems.

Regulation Q is a good example. Banks operate by holding high interest assets and low interest liabilities. They profit from the spread. The spread also absorbs interest rate volatility. Regulators believe that a contributing factor to the financial meltdown of 1929 was the unrealistically thin spread brought about by intense and uncontrolled competition among banks. Regulation Q effectively sets a minimum to the spread. It is, as are all such regulations, "anti competition". We adopt such regulations because we believe that the net long term benefit to the society and to the economy is positive.

This sort of relationship between competition and volatility may be viewed as a risk-return metric. In perfect and efficient markets expected returns from risky assets are related to the risk of holding those assets. Risk is defined as the uncertainty in the returns projection. The CAPM (capital asset pricing model) (Sharpe 1964) describes such a relationship. The CAPM, though controversial (Fama French 1992) is nevertheless accepted as mainstream financial theory today (Munshi 1994). It states in an efficient market higher expected returns can only be achieved at a higher risk and that the economy at any time establishes a consensus price of risk thru its market mechanism. When competition is fierce, it forces some firms into dangerous waters to "bottom fish". In this state, the economy (or the industry) appears to be healthy but is metastable in the sense that a small external shock can cause severe attrition because of the amount of risk that has been taken. In this paper I argue, using various measures of risk, that the credit card industry today (1998) is in such a state.

The architecture of the credit card industry

The credit card industry was born in the 1960s. Credit card banks (i.e. banks that exist primarily to issue credit cards) and non-bank issuers were an innovation of the '80s. The current architecture of the credit card industry evolved when the Bank of America renamed its "Bank Americard" as "Visa" and invented the payment network.

A simplified model of this architecture is shown in Figure 1. The payment network is the glue that binds the system together. The major firms in this arena are Visa, Mastercard, American Express, and Dean Witter. (American Express and Dean Witter are also issuers). The payment network maintains a merchant base which consists of merchants who accept the card and to whom the network forwards prompt payment when a cardholder makes a purchase. The network charges the merchant a fee of approximately 2% of the sale for this service. This fee is the primary source of revenue for the network and the "merchant acquirers" who manage the merchant base.

The network also maintains "member financial institutions" who are banks that issue the network's credit card. The banks in turn "solicit" and build up a customer base of cardholders who will use their credit cards to make purchases from the merchant base and to obtain cash advances. The bank borrows funds from the money market and other sources and pays for the purchase at time t and then collects from the cardholder at time t+n days where n is typically between 30 and 120. The assets of the credit card issuer consist of the accumulatd debt of the cardholders to the bank. The liability side consists of loans it took from the money market to pay the network for the purchase. It profits from a spread in the effective interest rate between the asset and liability sides and from seasoning its assets, that is, maintaining a high value of n. It may also generate revenue by charging various types of fees.

A more complete picture of the bank's cash flows is depicted in Figure 2. The spread that the bank actually earns is significantly reduced by charge-offs and back office expenses. The management of these expenses, particularly charge-offs, is the crucial factor in credit card bank management in a competitive environment. A spread of 7% can rapidly diminsh to less than 2% with charge-offs in the 4% range. In a highly competitive arena, back office costs arise principally from origination and collection.

Spread management

The gross spread of the banker is the difference between the interest charged on loans and the cost of debt that supports these loans. Figure 18 shows the trend in these rates. In 1996 the average industry rates were: credit card rate = 15.5%, cost of funds = 15.2%, and gross margin = 10.3%. Yet the net margin earned by commercial banks that participated in a Federal Reserve Survey showed net margins of 2.14% in 1996. Rising bank earnings set a record of $14.8 billion in 3Q 1997 (Figure 12) but this come from revenue increases in spite of shrinking net spreads (Figure 12) and falling "efficiency ratios" (Figure 12). The net spread therefore represents a difference between two large and variable numbers.

Several variables ar responsible for the erosion of the gross margin. The most important of these is the charge-off rate. It represents the percentage of the outstanding that will not be collected due to bankruptcy (about 50%), fraud, or other reasons. Historically the charge-off has been around 3% for credit card users. Since 1992 the charge-off rate and the rate of personal bankruptcies have been rising during a time of sustanied economic growth, in what appears to be a paradox. At the end of 1997 credit card outstandings stood at $450 billion and the chargeoff rate at around the 5% mark. By Feb 1998, the chargeoff rate had risen to 6% with some banks reporting 8%.

Another source of margin erosion is the convenience user. Most credit card terms and conditions allow for a "grace period" before interest is charged on purchases. There is normally no grace period for cash advances. The grace period requires the issuer, in essence, to make an interest free float of up to 45 days on average to the cardholder. Some cardholders take advantage of this float and pay off the entire bill within the due date and do not carry a balance. Issuers whose revenues are generated mostly by interest are vulnerable to this behavior because they rely on seasoned balances for their income.

The "convenience user" rate has risen from 10% in 1990 to over 31% in 1997. Figure 11 shows the trend in convenience use reported by CardTracks at ramresearch.com. Figure 17 shows how a 31% convenience user rate can reduce a 13% gross margin to 1% under certain conditions. Securitization raises the effective rate of convenience users since the securitized portion is free of grace period float. The managed portion must absorb all the grace period costs.

Back office costs include solicitation costs, credit check and scoring costs, and collection costs. Competition has forced all of these costs to go up as bankers send out more solicitation letters (2.8 billion in 1997) and reach into riskier demographics for new accounts. These accounts necesarily require better and costlier screening and also increase collection costs.

On the portion of the portfolio that is securitized, the bank incurs origination and "convenience user" float costs but saves on collection costs and may be able to earn a 1% spread. On its own managed portfolio, it may struggle to break even with chareoffs and back office costs.

In such cases the issuer must rely on fee income for revenue. Annual fees are contentious and a hard sell with consumers who hold multiple cards. Currently there are over 120 million cardholders in America with an average of 4 cards per cardholder. Event fees have become popular to circumvent the annual fee problem. Events that generate fees may include delinquency, cash advance, and going over the limit.

Generating interest income from a credit card portfolio requires the bank manager to play a dangerous game of "chicken" with the consumers and their credit scores. The bank wants the consumer to carry a balance for as long as possible but not so long that it increases the probabilty of default beyond a given threshold. The banker's seasoning dilemma is shown graphically in Figure 3. The optimal seasoning lies at the intersection of the two curves but since the impact of seasoning on default is not known very well conservative bankers will stay well behind it. It may be that competition is leaving some banks with few options but to extend their seasoning policy to higher risk levels.

Credit risk management

Low profile but watershed innovations are changing the consumer credit industry in fundamental ways (Derivatives Strategy 1998a). These innovations include scoring, convergence, and credit derivatives.

Pressed by competition to hold near-default assets in their portfolio, bankers can no longer rely on conventional credit ratings and have resorted to costly models that use a technique called "credit scoring". Scoring models use classification and regression techniques to identify variables that differ significantly between defaulters and non-defaulters and to identify the predictors of default. Scoring costs significanly reduce the already thin margin of issuers even though their classification accuracy is uncertain especially when projected to economic conditions that might destabilize the industry.

The term "convergence" refers to the growth of securitization of bank loans. Securitized credit card debt has grown from zero to over $ 220 billion in 1997 (Figure 15). By then about half of the credit card debt was securitized (Figure 9). Some banks have securitized more than two-thirds of their credit card portfolio (Capital One 1998). In a typical deal an investment banker such as CS First Boston will underwrite $100 million in credit card receivables at a discount and create asset backed securities (ABS) against the projected cash flows. These securities are then floated and traded in fixed income securities markets and rated by Fitch Investment Services. Securitization allows bank managers to remove selected segments of their credit card portfolio from the books and to generate cash flow when it is needed; and it narrows the distinction between the bond market and the bank loan market.

Banks also use swaps and other derivatives to manage the credit risk of their portfolios. The size of this market exceeded $50 billion by the end of 1997 (Derivatives Strategy 1998b). In addition to vanilla fixed for variable asset swaps and total return swaps, banks also use innovations called "credit default swap" and "credit spread options" which are used to carve out the credit risk portion of an ABS and sell it separately to a contraparty who wishes to take the risk.

Revolving credit data from the Federal Reserve's G19 report (FRB 1998) shows the innovation and rapid growth of securitized revolving credit. The rate of transfer of the outstandings from bank balance sheets to securitized debt slowed between 1992 and 1994, the profitable years, but has accelerated since then. By 1997, had broken even with commercial banks with each holding approximately 40% of outstanding revolvers. The data are depicted graphically in Figure 9.

Competition in the credit card industry

Competition between issuers of credit cards reduces their margins by increasing solicitation and scoring costs on the one hand and simultaneuosly increasing the chargeoff rate and decreasing revenues on the other side of the spread. The squeeze on the spread is exacerbated by competition motivated priorities such as low introductory rates, balance transfer, and low or absent annual fees.

It is difficult to maintain both market share and spread; and in their attempt to do so the banks have incrementally increased the complexity of the fee structure. In addition to annual fees fee structure may contain different kinds of fees such as over-the-limit fee, late payment fee, cash advance fees, and fees that may be imposed under certain conditions. The interest portion may include low introductory fees, punitive rates under certain conditions, varying grace periods, and variable rates with floors but no caps. The complexity and variety of credit card terms and conditions confuse consumers and they complicate cash flow and spread management by bank managers.

The rapid growth in credit card debt since the end of the 1990 contraction (see Figure 16) actually understates the impact on household debt because of the "home equity loan" (HEL) innovation and its derivative called the "125% LTV". Credit card debt that has been "consolidated" with HEL no longer show up as consumer loans but they have not gone into increasing the value of houses. The market for HEL is nearing saturation because possibly of this reason and also because it has been very actively marketed by banks as a tax shield and debt consolidater.

The 125LTV innovation essentially extends the HEL market to 125% of the value of the home. Some firms such as Fidelity First Financial Corporation (CardTracks 1997) have targeted the 125LTV market as a credit card consolidation program and others such as ContiAsset (American Banker 1998) have started a program of issuing credit cards against a pre-approved HEL or 125LTV. ContiAsset was formed by parent HEL specialist ContiFinancial. The 125LTV portfolio grew from $3 billion in 1996 to over $10 billion in 1997. A successful LTV125 marketing program will increase household debt levels which in turn will increase the number of defaults for a given rate of economic slowdown should it occur.

The use of low interest introductory rates and balance transfer has significantly reduced the industry's interest earnings. The introductory rates are typically 5% or 6% while the normal rates are in the 16 to 20% range. In a balance transfer, one bank cannibalizes a performing seasoned account of another. Each such transfer reduces the net profitability of the industry as a whole.

Chargeoff, bankruptcy, and "competition risk"

In his study of consumer debt Garner (1998) found that the effect of rising consumer debt on the economy is uncertain because the available data do not have a clear interpretation and because recent changes in the financial industry makes it difficult to determine the relationship between economic growth and consumer debt. He concludes that the rise in consumer debt is not sufficient predictor of a recession especially in light of healthy growth in employment, disposable income, and savings.

Although the correlation between personal bankruptcies and consumer loan defaults is intuitive and empricially validated (Bishop 1998, FRB 1997), the direction of the causality is not clear and is possibly complex. In most analyses (Bishop 1998) bankruptcy is treated as a purely exogenous process that generates chargeoffs. However, the data (see Figure 5) are also consistent with a competition theory of bankruptcy.

Competition can force issuers who have access to scoring and other techniques, to seek out marginal consumers who will carry a balance. This process itself can generate the observed coincident rise of chargeoffs and bankruptcies. The rise in bankruptcies is not a simple exogenous process but one that is created, at least partially, by the competition strategies of credit card issuers.


Consumer debt both as credit card debt and as home equity loans used to consolidate credit card bills, has doubled since the 1990 contraction during a period of sustained growth in GDP and employment. Simultaneously, personal bankruptcies and credit card charge-off rates have also risen to record levels amid intense competition among hundreds of credit card issuers. The thin spreads of issuers and high leverage of consumers creates a metastable system which will be unable to sustain even a small economic downturn without significant volatility, attrition, and social costs.

The social cost of competition extends beyond attrition. In the competitive arena, banks are targeting low income families, Hispanics, and the elderly with home equity. The increasing number of convenience users who tend to be on the upper layers of the income distribution remove wealth from the system by benefitting from the float. A segment of low income households may "benefit" from the chargeoffs. These "benefits" do not come at the cost of the banks who are already operating on thin margins but at the cost of the hard working, perhaps mostly low income families, unable to pay off their credit card debt and carrying several thousand dollars per family of high interest credit card debt.

The confusing pricing structure of credit cards is the result of competition. It is confusing to consumers and perhaps even to the bankers themselves who are uncertain of the revenue and spread dynamics of portfolios consisting of myriads of pricing structures. Perhaps consolidation and structural changes in the credit card industry will help to alleviate the social cost of competition and nurture a kinder gentler capitalism.


Advanta Holdings, 10-Q filing, third quarter 1997, SEC
American Banking Association, ABA, 1998
American Bankruptcy Institute, ABI, 1998
Paul Bishop, "A time series model of the U.S. personal bankruptcy rate", in Federal Deposit Insurance Corporation "Bank Trends: Analysis of emerging risks in banking", FDIC, Feb 1998
Valerie Block, "Supreme court upholds nationwide card charges", American Banker, June 4, 1996
James Bothwell, "U.S. Credit Card Industry: Competitive developments need to be closely monitored", United States General Accounting Office, GAO/GGD-94-23, 1994
CardTrack, "The road ahead", Ram Research 1998
CardRrack, "Card debt drives high LTV mortgage", Ram Research ,1997
Derivatives Strategy, "Modeling credit risk", v3 n1, derivatives.com, 1998a
Derivatives Strategy, "A credit derivatives primer", v3 n1, derivatives.com, 1998b
Glenn Canner and Charles Luckett, "Developments in the pricing of credit card services", US General Accounting Office report number GAO/GGD-94-23, 1994
Capital One, 10-Q filing, third quarter 1997, SEC
James Daly, "Savings on postage", Credit Card Management, May 1997
Eugene Fama and Kenneth French, The cross section of expected stock returns, Journal of Finance, v47:2, 1992, p427
Federal Deposit Insurance Corporation, FDIC 1998
FDIC, "Commercial Banking Performance: third quarter 1997", Feb 1998, FDIC
Fitch Investor Services, Fitch, 1998
FRB, Surveys of consumer finance, 1992
FRB, Report FR2835, Credit card interest rates, 1997
FRB, Report H15, Historical interest rates, 1998
FRB, Report G19, Consumer credit, 1998
Federal Reserve Board of Governors, "The profitability of credit card operations of depository institutions", August 1997, FRB
C. Alan Garner, "Can Measures of the Consumer Debt Burden Reliably Predict an Economic Slowdown?", in "Economic Review", Federal Reserve Bank of Kansas City, 1998
John Maynard Keynes, "The general theory of employment, interest, and money", Harcourt Brace, NY, 1936
Jamal Munshi, "Fama French 92 and the beta puzzle", http://munshi.sonoma.edu/working/ff92.html, Sonoma State University, 1994 William Sharpe Capital asset prices: a theory of market equilibrium under conditions of risk, Journal of Finance, v19, p425, 1964
A.P. Murray, "Debt and the consumer", Business Economics, April, 1997
Spencer Nilson, "The Nilson Report", March, 1997
Spencer Nilson, "The Nilson Report", January, 1998
Spencer Nilson, "The Nilson Report", February, 1998
RAM Research Group, ramresearch.com
Robert Solow, "Economic history and economics", American Economic Review, v75 n2, 1985 Adam Smith "The wealth of nations", Modern Library, NY, 1937
David Wheelock and Paul Wilson, "New evidence on returns to scale among U.S. commercial banks", Federal Reserve Bank of St. Louis, 1997
Peter Yoo, "Charging up a mountain of debt: Households and their credit cards", Federal Reserve Bank of St. Louis, 1996

Figures and Tables
Please send comments about this paper to jamal@munshi.sonoma.edu

[Home] [Working papers]