Credit card companies are getting a lot of grief in the blogosphere (not to mention Congress) lately. Most of these critiques are just a bunch of shaggy dog stories, but the very smart Rortybomb has an extremely numerate post in which he points out that when the interest rate on your plastic goes from 8% to 28% because you’re two days late on a payment, it’s highly unlikely that this is a pure reflection of a change in your probability of default. His analysis indicates that the way this price (i.e., interest rate) change is determined is not by the change in creditworthiness that is indicated by the new piece of information, but instead by the price sensitivity that is indicated by this new piece of information:
One model is that the credit card companies are lying to you – they think of you less as an individual to have a dynamic risk factor dynamically assigned to you, and instead as part of a portfolio to have a specific rate of return extracted from. So they have statisticians and psychologists not to create a credit risk, but instead to figure out who is likely to pay what when, and use that to keep their returns very high. Quants to study how much they can squeeze from someone – not too much, but not too little. So it is less about the awesome part of markets, the price information and the convergence and feedback, and something more feudal.
Update: So after discussing with some people, I think the big problem, as was a problem with subprime mortgages, in invoking the market is that the inital rate is competitive, but the refinancing later is not. Companies bid up and down your initial rate plus rewards package. However once you are locked in, and get a balance going, nobody is bidding against your rate.
His conclusion is that credit card companies are, morally speaking, “scumbags”.
In my experience, and very broadly speaking, he is correct about the logic by which price changes (including interest rates, fees and other contract terms) are determined. The credit card company is making decisions with the intent of maximizing their shareholder value, consistent with the law. (See *1 for more details about method.)
In other words, this is a normal consumer market in which the guy selling you something is not looking out for you, but is trying to make money for himself. This is just like a car company, search engine provider or private university. Why is the guy who sells you a credit contract responsible if you are later unhappy about the decisions that you made?
In the specific example that Rortybomb cites, a reasonably prudent person should be aware that he or she has just signed a contract that gives the counterparty the right to increase the interest rate on a debt contract from 8% to 28%, or to the so-called penalty rate, if you miss a payment. If you have a credit card, go to your cardholder agreement and search for “penalty rate”. In any normal such agreement, you will almost certainly find a specification of the penalty rate, and the conditions under which this rate may be invoked. Expecting that your counterparty will not act to serve their own interests under a contract is the attitude of a child. If you didn’t want this deal, you shouldn’t have signed the contract.
Now, fraud is generally forbidden in these markets, and is for credit cards as well. There can get to be a gray area – what amount and type of disclosure is required and so on. Second, there is normally some kind of (speaking non-technically, and without a specific legal meaning to the term) implied warranty. Even if my purchase agreement with GM doesn’t say “this car will not explode in a ball of flame if you tap the accelerator twice and then hit the brakes”, they are subject to legal action if this occurs.
What we are really debating is where to draw the line on these two questions. That is, to what degree does the issuer have to emphasize risks, what degree of complexity should be allowed in the contract and so forth?
The Center for American Progress is typical of current sophisticated liberal thought in emphasizing this:
Credit cards are convenient, but difficult to use responsibly. Credit cardholder agreements are written in language that is above the level at which about 50 percent of U.S. adults read, and information within them is poorly organized. Moreover, issuers appear to “price” the cost of using credit cards by taking advantage of cardholders’ behavior biases. For example, credit card issuers take advantage of the fact that cardholders underestimate the probability of paying late or going over the credit limit, and punish this behavior with fees and increases in the penalty rate.
The right information at the right moment can help cardholders make better decisions. A text-messaging system by itself would not prevent issuers from continuing to price credit cards however they like, but it would orient cardholders toward the best outcomes, such as paying on time and not going over their credit limit. This approach recognizes that most individuals don’t behave like homo economicus—the “economic man” of economic textbooks who maximizes every financial decision and has perfect information to do so. Most cardholders could benefit from a “nudge” toward a more beneficial choice.
But why is it the credit card company’s job to “nudge” you to “more beneficial choices”?
It is an unfortunate reality that there are many people who are not equipped to get along in a capitalist system. They lack some combination of (rarely) the basic intelligence and (much more frequently) the emotional maturity and self-discipline required to make their way in a world in which others are not looking out for them. Much of the rationale for traditional notions of child-raising, education and social organization is to prepare people to live in such a world. That is, to produce actual adults. To the extent that we can count on people to act responsibly, we can have a less regulated economy that will tend to produce greater freedom and growth. But the problem of how to deal with the semi-incompetent is a real problem that will never go away entirely.
One practical effect of proposed credit card legislation would be to make it illegal for party A to voluntarily engage in a credit contract with party B that has some specific elements that might be abused by an irresponsible person. Why should this freedom of contract be restricted for responsible people? Because the guy who lives down the street might use the same contract to drive himself into personal bankruptcy with Cheetos, beer and big-screen TVs?
Maybe, actually. If (i) the abuse problem were severe enough, (ii) the productive uses of such credit extremely rare and (iii) there were no other practical remedies, this could be a theoretically poor, but practically-workable, compromise. But I don’t think any of these assumptions holds. First, the vast majority of people who use credit cards don’t default, and second, they continue to voluntarily use this source of credit.
Further and most importantly, I think there is a different and better approach. I don’t think our basic strategy should be to forbid contracts that are only suitable for actual grown-ups, but instead to provide safe havens for the less competent. This could, in theory, include things like requirements for a “simple card” alternative and so on. I’ve tried to describe such an overall approach to financial regulation as “walls, not brakes”. It would not eliminate the problem of some sympathetic people getting over their heads in credit card debt, but should reduce it, while not giving up on the dynamism enabled by freedom of contract.
*1 In my experience, the way this is determined is by running structured experiments in which, by illustrative example, 50,000 randomly-chosen late payers are given a large interest rate bump and a matched control group is not. The experiment is run long enough to measure the true life-cycle profitability of each group. The superior rate approach is deemed the “champion”, and future tests pit other potential “challenger” rate, fee or other changes against it. Segmentation of test results is used to fine-grain this to identify sub-groups for whom different approaches work better. While individual scores are developed by individual, it is a lot more complicated than simple data mining to develop propensity scores (for reasons of identification of causality that are complex and beyond the scope of this post).
The “reasons why” such a strategy of raising interest rates dramatically for most people who pay late might work (e.g., it actually indicates a higher default risk; it indicates somebody is more absent-minded, and will be slower to notice and respond to a rate change; it’s a predictor of future low profitability, and the very high interest rate is way to encourage them to go elsewhere, etc.) are complex, overlapping and subject to interpretation. Nobody ever fully understands why any consumer or group of consumers makes a specific decision. (This is a huge practical distinction between the professionalized, wholesale and commoditized bond market that Rortybomb references, and a typical consumer market like this one.) However, the test result – that strategy X makes the issuer more money – is the reliable result.
And while there is often some element of consumer lock-in in various ongoing purchase decisions (this is why razors tend to be surprisingly cheap, and blades surprisingly expensive), it’s not quite true to say that “refinancing later is not [competitive]”. The of the explosion in credit provision to high-balance revolvers was largely predicated on the invention of the balance-transfer with teaser rate business model 15 – 20 years ago; that is, on poaching exactly such (potentially) high-profit customers sitting in the portfolios of incumbent issuers.