The harsh economic downturn that has chastened credit-happy consumers, along with increased scrutiny by regulators, will force card issuers to rethink their business models as the economy begins to recover, according to Wharton faculty and credit industry analysts.
Since Bank of America first introduced credit cards in California in 1958, the industry has grown to nearly $1 trillion in revolving credit outstanding, with more than 700 million U.S. credit card accounts. During the economic boom that took place between 2005 and 2008, a credit-fueled consumer-spending binge brought the U.S. savings rate, effectively, to zero. In May, however, that savings rate hit a 15-year high of 6.9%, boosted in part by one-time federal stimulus payments made to taxpayers last year. In 2009, for the first time in the 40 years that the Federal Reserve has tracked the industry, revolving credit is expected to decline.
“Even without changes in legislation, credit card companies would be changing lending patterns. They are already tightening credit and raising standards because the risk has gone up,” says Wharton finance professor Nicholas Souleles.
Souleles predicts the future of the credit card industry will be shaped in large part by broad macroeconomic patterns. The current drop off in the use of credit by consumers is due in part to a tightening of credit standards by issuers stung by a default rate of up to 8% — double the figure for 2006. At the same time, demand has been cut short as consumers are more reluctant to take on debt when they are uncertain about the value of their home and other assets, as well as the security of their jobs.
During the recent housing boom, Souleles points out, it made sense for homeowners with access to home equity loans to borrow and save less. “Going forward … access to credit will improve, but consumer desire to borrow might very well have changed for a long time, even as we come out of the recession,” he says.
In the short run, individuals’ new-found willingness to save may stall recovery, he adds, since as much as 70% of the U.S. economy hinges on consumer spending. However, over time, increased savings could provide the capital to invest in new innovation that will ultimately create new jobs and higher economic growth, he predicts.
Credit card issuers are now in search of new models to account for changing economic realities, according to Wharton statistics professor Robert Stine, who hosted a recent Wharton Financial Institutions Center conference titled, “Modeling Retail Credit Risk after the Sub-Prime Crisis.” Stine says industry players had developed elaborate credit risk models that were relatively successful until they were hit by the economic crisis and resulting changes in consumer behavior. For example, mortgage debt was traditionally considered sacred by consumers, who made their payments even under dire financial pressure rather than risk losing their home. But once consumers began to think of their home more as an investment, rather than a place to live and establish roots in a community, they were more willing to walk away from mortgage debt if the home’s value fell near to or less than the amount of their outstanding mortgage.
“The credit industry has come to view its retail business much more as a portfolio of assets,” says Stine. “Each customer represents a collection of debts, and customers vary in how they treat those obligations.”
Getting to Know Borrowers
Meanwhile, credit card companies had developed ways to evaluate individual applicants, but they had little ability to cross-check how many other credit card issuers or lenders had also extended credit to the same applicant, according to Stine. “We’ve learned you can’t just look at giving one credit card. If the borrower had many credit lines, that would add up to much greater risk than any one credit card company would be able to track.” Credit bureaus attempt to provide this type of information, but in many cases they do not have enough details about individual arrangements that lenders might have with a customer who has fallen behind on their payments, or enough behavioral data to predict how an individual would balance a credit portfolio. In addition, certain laws limit the amount of information credit bureaus can collect, including information about income.
In the future, he says, large financial institutions that have multiple relationships with individual customers may be able to better track income and behavioral patterns in order to improve their ability to gauge risk. For example, even if a company could not get complete income information, it might be able to look at checking account records for regular deposits indicating income.
Stine notes that the old ways of determining risk, primarily the FICO score developed by the Fair Isaac Corporation to gauge creditworthiness, were successful in determining whether one individual might be a better risk than another. However, he says, the FICO score is not in synch with changing economic conditions. For example, a FICO score of 600 might indicate a 10% risk of default in a strong credit cycle, but 40% in a slump. “It’s still a FICO score, but we don’t know how to use it,” says Stine. “The score itself is interesting, but the only way to make a credit decision is to connect it to how much money I risk by having you as a customer.”
Credit risk analysts are now in the early stages of trying to find new ways to build models that link information about a credit card user to make a better assessment of default risk, according to Stine. “I think the models will be stronger as a result of the crisis.” Companies will continue to use models because they are efficient and can accommodate large-scale predictions “better than going to talk to Joe at the corner bank. But the [old] models had lost degrees of calibration with the real world around them. I think the industry is willing to grow and improve and will find new sources of data to allow that.”
The economic collapse and change of administration in Washington have already led to new curbs on the credit card industry. In May, President Obama signed legislation which prohibits issuers from imposing fees on customers who inadvertently exceed their credit allowance, and limits the fees companies can charge for late payments. In addition, credit card companies will be able to raise rates on existing debt only if consumers have paid their bill more than 60 days late. The law also limits marketing aimed at those under age 21.
New Regulatory Environment
The Obama Administration is working with members of Congress to create a Consumer Financial Protection Agency, which will provide federal enforcement of the credit protections for consumers.
Souleles says it is hard to argue against much of the new consumer protection legislation because it requires greater disclosure and puts a threshold on fees. However, the provisions that limit card companies’ ability to set their own rates is trickier, because — in times of higher risk — companies may need to raise rates to remain in business and continue to lend, he adds.
Wharton finance professor David Musto notes a tension in any policy designed to address problems in consumer credit. Concerns about predatory lending must be balanced against the problem of restricting the flow of credit to low-income people — or people who live in certain neighborhoods — who are able to make their payments and benefit from loans. He points to the debate over the Community Reinvestment Act (CRA), which requires banks to lend in neighborhoods where they take deposits but make relatively few loans. The act has been criticized as contributing to the subprime mortgage mess because it allowed some home buyers to take on more debt than they could handle.
At the same time, others contend that CRA helped deserving families get a toehold on homeownership that will allow them to build assets and contribute, long-term, to economic growth, Musto notes. “Some people might want to rethink how they want banks to help in the communities they take deposits in. Did the Community Reinvestment Act have negative, unintended consequences? That debate is connected to any discussion of consumer credit.”
Musto adds that much of the financial infrastructure that allowed widespread use of credit cards is under stress because the financial collapse has made it impossible to securitize credit card debt. “It is very hard to package [credit card debt] right now. The business of believing you could package subprime loans and isolate the buyer from the risk is going to be rethought.”
Wharton marketing professor Stephen J. Hoch says credit card companies and savvy consumers were both playing games that have come largely to a halt with the freeze on credit expansion. Credit card companies were able to use the fine print in contracts to draw many borrowers into escalating debt, he says. Meanwhile, customers who paid their bills promptly each month reaped the benefit of free, short-term financing, convenient transactions and other perks, such as airline miles, at the expense of the financial institutions issuing their credit cards.
For the majority of consumers, Hoch says, credit cards are valuable because they allow people to borrow against the future. “As long as the future is brighter than the current time period, it’s a rational, normal and appropriate thing to borrow. Unfortunately, people get overly optimistic — or can’t do the compounding math.” Inexperienced or financially unsophisticated credit card holders may have been concerned that they would not be able to carry a certain level of debt, Hoch notes. However, if a large financial institution was offering them credit, they assumed “the credit card company must know more than I do.”
Credit cards are not going away anytime soon, Hoch adds. Plastic makes transaction processing faster and more convenient. “But just as people can’t use their house as a piggy bank anymore, they can’t use credit cards as a piggy bank. The growth period is over, but credit cards are ubiquitous and they are not going away.”
A Mature Industry
Brian Riley, research director for bank cards at TowerGroup, a Massachusetts financial services consulting firm, says that after 40 years of growth, generated largely by changing technology, the credit card industry has now matured.
The industry’s strength is in its payment networks that can also be adapted to debit cards, which draw funds directly from a consumer’s checking account, and newer prepaid credit cards. “The debit card is now overshadowing the credit card,” he says. “The shift to debit is important and we expect the trend will continue. On the credit card side, it’s going to be harder to get a card and harder to maintain it. There is going to be less forgiveness and we expect overall volume to decline.”
Riley’s firm estimates that revolving credit will drop from $960 billion in 2008 to between $777 billion and $885 billion in 2010, depending on the severity of the recession. Going forward, he says, credit card issuers need to sharpen their customer acquisition models to take a more careful approach to credit scores, payment history, credit line management, debt burden and the length of the company’s relationships with the customer.
“Rather than keeping accounts that have high loss potential and limited revenue opportunity, the mission becomes to close out those customers’ active lines and drive them off the books by means of an aggressive account management strategy,” Riley writes in a report titled, “After Boom and Bust: Navigating the Credit Card Industry into the Next Economic Cycle.” Credit card issuers “must walk a fine line when making account adjustments. It is preferable to have a well-planned strategy in place than to devise reactive policies under pressure.”
Stine emphasizes the cyclical nature of the credit card industry, which has enjoyed strong growth and profitability, particularly in the last economic boom. Cards were generating enormous profits when the spread between the cost of money and the rates financial institutions could charge borrowers widened, he notes. The overall rise in economic activity masked the risks. When the cycle turns down, the risk is magnified. “It’s not just one person who defaults,” says Stine. “It’s a lot of people defaulting all together. There is an aspect of contagion.”
The best models that could ever be designed are still vulnerable to human override, Stine notes. The industry has always faced the scenario of a risk officer complaining that potential problems were rising, alongside colleagues who argued: “Hey, it worked last year and we are making a huge profit. Why should we stop?” Stine says. “All of a sudden, I think a lot more people are listening.”