The good, the bad and the doubtful in credit card reform

You can’t get a good understanding of the true impact of the cards legislation without a broader look at the historical context. What we have seen in the industry for the last two decades is a steadily increasing ability of card issuers to design products that are profitable on a stand-alone basis. Generally, this involves segmenting the population into ever smaller groups at which more specific products are directed. Thus, instead of “revolvers” paying interest that subsidizes the cards issued to “transactors” who pay their bills each month, we now have distinct types of cards with different fee structures. 

The cards held by people who tend not to pay interest often have annual fees or affinity programs that generate revenues that are a substitute for annual fees. And the issuers that are best at marketing those products are very good at ensuring that most of their cards get used frequently, so that most of the cards generate enough revenues from interchange to offset the fixed costs of having an account. 

Similarly, card issuers that specialize in products that result in high levels of borrowing are skilled at offering products that use features like teaser rates and low balance-transfer fees to attract customers for whom borrowing is highly likely. It surely is the case that there are more and less profitable customers within each of those product classes, but the groups within which the cross-subsidization occurs are much smaller than they used to be. 

So, in the light of the new Credit Card Accountability, Responsibility and Disclosure (CARD) Act of 2009 signed into law by President Obama last month, the most important question is how the bill will affect the fees and products that issuers are going to offer. 

The most obvious outcome – one that is happening right now -- is that card issuers are responding to the likelihood that the options they currently use to design specialized products will be truncated in important ways. Here, I would distinguish two groups of changes in the bill: changes that enhance the transparency of pricing and changes that limit the ability to price risk.

Change One: Shrouded Product Design 

The first set of changes in the bill outlawed a set of product attributes that (at least to the academic outsider) seem to draw their attractiveness largely from their complexity. These are the “shrouded” attributes like double-cycle billing and minimum finance charges. Over the life of an account those charges can produce a substantial amount of income for the issuer, but because they are not apparent at the front end the customer does not account for them in estimating the price of the card when it is acquired and used. I would also put in that category “sharp” practices like procedures for receiving payments that enhance the likelihood that payments will be treated as late even when customers in fact place them in the mail before the date on which they are due. 

On the surface, there is a lot to be said for banning those practices. If they make the product more profitable to the issuer only because customers systematically misunderstand the cost, the market produces a race to the bottom, in which those issuers seeking profit most zealously will have a strong incentive to maximize their use of those attributes. To be sure, the market does produce countervailing incentives. Cardholders may learn over time to avoid aggressive issuers, and the reputation of issuers whose products have too great a balance of back-end costs may suffer. We have seen in advertising for years how important it is for issuers to develop a reputation for having benign products. But how sure can we be that the reputation for benign products reflects (or depends on) anything more than careful advertising and marketing? 

In any event, the forces of publicity and reputation already had persuaded most of the largest issuers to abjure many if not most of the practices banned by the legislation. And given the increasing concentration of credit card lending (the top six lenders now have about 90% of the credit card debt), practices abandoned by the largest issuers have relatively little market significance. 

To that extent, the major impact of the legislation will be to alter the practices of the smaller issuers who are least likely to be affected by publicity. To the extent the smaller issuers are issuing cards to people who have severely constrained credit options, the bans can substantially improve the transparency of the most expensive products. BREAK

The Problem with Banning “Harvester” Cards

The provisions banning “harvester” cards are illustrative. Those cards often trumpet amazingly low interest rates in the range of 15-20% for people with very bad credit; they recover the costs by charging to the card up-front fees that limit the available credit to a small share of the amount allegedly available on the card. Thus, in fact, the effective interest rates on those cards are more commensurate with those of payday lenders than those of conventional lending products. Banning those cards will have no effect at all on the practices of the JPMorgan Chases and Capitol Ones of the world, but it will change the marketing and pricing of the cards on the cutting edge of risk. 

The most troubling thing about the strategy of banning these kinds of practices is the likely futility of the statute. As noted above, to the extent the bill identifies practices that ever have been in the portfolio of mainstream lenders, it bans things that for the most part those lenders already have abandoned. But nothing in the legislation alters the motivations of issuers to rely on the kinds of “shrouded” terms that legislators found objectionable in the statute. And so the legislation makes sense only if the problem of shrouded terms is a static problem of a few specific bad practices into which some issuers have drifted. 

But that ignores the behavioral aspects of the problem. The most important difficulty that issuers face in garnering revenues based on shrouded terms is that consumers that use the product have a financial incentive to learn about the terms over time – they are, after all, costly to the consumer that does not understand them. What this means is that over time the revenues from any particular shrouded term will decline as consumers learning of the term alter their behavior to avoid the conduct that results in higher charges. 

The history of late and over limit fees provides a good example of this phenomenon. Issuers in the early 1990’s identified these fees as a good source for enhanced revenues. The fees are particularly attractive as part of the revenue structure for relatively risky low-balance cardholders for whom even a $25-$30 fee is a major return on the outstanding balance. Thus, the total revenues from those fees doubled during the 1990’s (from about 70 basis points to 140 basis points, as a share of outstanding receivables). But as consumers learned to adapt their behavior to those charges, issuers were no longer able to increase revenues from those fees. Thus, despite frequent press reports about continuing increases in the fees charged for late payments or over limit transactions, the revenues from those charges have been stagnant for this whole decade. In 2008 -- not the best year for consumers -- issuers received only 120 basis points in revenues from those fees, according to stimates calculated from the annual Cards Profitability Survey of Cards and Payments. 

The key regulatory problem is that issuers designing their contracts do not face a fixed set of onerous terms from which they make selections based on the predilections of their customers. On the contrary, they are constantly developing new strategies so that they can rely on new terms with which customers have not yet been familiar. This is most obvious from the steady progression through double-cycle billing, the poster-child for aggressive card issuers of five years ago, to minimum finance charges, first introduced a few years ago. 

I see little reason to think that provisions of that sort contributed to the efficiency of card products, and so find nothing wrong in Congress’s decision to ban several of them. But one thing we can be sure of is that the issuers that were motivated to develop those provisions will soon develop newer provisions. Now that the legislative moment has passed, those provisions will be developed in a world in which the prospects for renewed legislation are substantially diminished. 

A more sensible response would look to one of the few things the mortgage industry has done well: standardized the “boilerplate” terms of the transaction and letting issuers compete only on the product attributes that are important to attracting customers: interest rates, annual fees, credit limit, rewards programs, and the like. A standard credit card contract like the Fannie Mae home mortgage would go a long way to facilitating a general understanding by consumers of the basics of how credit card contracts work – because they would all work the same way. This is not to say that I think Congress took a wrong turn in banning many of these terms. It is more akin to regretting a missed opportunity to move the industry forward, an opportunity that is unlikely to repeat itself in an economy that presents Congress with so many burning issues for attention. 

Change Two: Pricing Risk 

Conversely, the most dubious change in the legislation is the limitation on the right of the issuers to alter interest rates from time to time to reflect changes in the risk profile of their customers. The issuers in this instance did a poor job of making a case to Congress to explain the business role of these provisions. The dominant public view, incorporated in most press reports, was that the law bars “arbitrary” interest rate hikes. There is, of course, nothing arbitrary about the interest rate hikes; they rest on determinations by the issuer that the customers facing those hikes are riskier or less desirable in some way than they previously were. The quality of those determinations surely differs from issuer to issuer and from time to time, but I doubt that many of them in recent months have been arbitrary in any meaningful sense. Rather, they reflect the need of issuers to both contract the size of their portfolios and increase the revenues available to them to match the rising losses from delinquencies that they are facing. 

That change will be particularly important to those that issue cards with large credit limits, for which interest revenues are the main way to cover the cost of funds and charge offs. For those issuers, the inability to shift interest rates over time based on the apparent riskiness of the cardholders at any given time necessarily will be replaced by some combination of lower credit limits (something we’ve seen already) and higher up-front fee structure that generate up-front revenues sufficient to cover the expected costs of bad events in the future. 

Those kinds of fees have led to the common charge that issuers are making good customers pay fees to cover the misdeeds of the bad. But it is not as if the issuers are charging the fees only to the “good” customers and not to the “bad”; they are charging them to everybody because they don’t know which people will generate the losses that make the fee prudent. The fee is not really any different from the annual fees charged on many cards already. The problem is that the issuers are being forced into a fee structure where they have to guess farther in advance (when they set the initial interest rate) whether the customer will turn out to be good or bad. An obvious response to that legislation will be to charge a greater number of fixed fees up front, if only because those fees have the virtue of making sure that cardholders that take cards will use them frequently. (Why pay an annual fee for your credit limit if you are not planning to use the credit?) 

Looking Ahead

If anything is safe to predict, it is that we have not yet seen the pricing structures and products that these changes ultimately will motivate. Issuers have relatively little knowledge right now about how their customers will respond to the product changes the new legislation will motivate. I think many issuers were surprised, for example, at the first response of so many of their prime cardholders to economic turmoil: paying down their balances instead of lowering their monthly payments. The products surely will change as issuers develop a more accurate understanding of how cardholders will respond to the rapidly changing product environment. 

More importantly , issuers as yet know very little about the extent to which the crisis will effect a long-term shift in borrowing and spending habits. The rapidly accelerating shift to debit card use in the last 24 months, for example, suggests a shift away from borrowing-based consumption, a shift that would have a much greater effect on the credit card product than anything Congress is contemplating. This is particularly true if that trend is truly age-based – so that it becomes ever harder to identify young and middle-aged households that will use cards as a routine transaction and borrowing vehicle. 

In the end, the most likely outcome is that these changes in consumer behavior will increase even more the advantage of the technologically sophisticated issuers (like Capitol One and JP Morgan Chase) and disadvantage the relatively unsophisticated issuers (like Bank of America and CitiBank). The more sophisticated issuers will do a better job of underwriting products that will be both attractive to the customers and profitable to the issuers, and as they do that their portfolios will steadily grow while those of their less capable competitors will shrink. As the number of issuers steadily shrinks, concerns about the level of competition will become even more serious than they have been in the past. The biggest question, though, is whether the issuers will ever witness the routinized credit card use on which they built their industry for the last decade.

I have my doubts.

About the author: