posted by Adam Levitin
The media has been abuzz recently with articles (here and here and here and also here) about rate-jacking--the often arbitrary increases in cardholders' interest rates. At first glance rate jacking makes little sense. Why raise rates on a good, paying customer? The cardholder might decide to close the account. Or the customer might not be able to service the higher rate debt and default? Why mess with a paying customer these days?
To understand rate-jacking, you have to understand two factors about credit cards: lock-in and the incentive to increase account volatility created by card securitization.
(1) Lock-in. Most consumers do not close accounts and switch cards because of higher rates. They do not perceive the switching costs as worthwhile--the transactional hassle, loss of accrued rewards points, and a ding to their credit score. There may also be a hyperbolic discounting/optimism bias at work--why bother to switch cards because of a higher rate if you don't think you're likely to pay interest? Whatever the reasons, it appears that consumers don't switch cards for less than $150 of increased costs (at least in some models). So there's a strong lock-in effect.
(2) Securitization Encourages Rate-Jacking. Credit card securitization is key to rate-jacking. Credit card securitization is different from mortgage securitization in three important ways.
First, unlike most mortgage securitizations, card issuers have some skin in the game. A typical credit card securitization requires the issuer to maintain a 7% untranched interest in the securitization trust; most issuers keep a larger interest, say 10%-20%. This, say 10%, of the trust's securities is paid pari passu with the say 90% that are represented by outside investor certificates. (The investor certificates are tranched amongst themselves.) The skin in the game is a good thing, but it might lull the market into missing the net impact of securitization.
Second, card issuers retain the "excess spread"--the revenue in surplus of what is needed to pay the asset-backed securities. Sometimes this is the case with mortgages, but credit card excess spread is likely to be larger as a percentage of revenue; the revenue from credit card receivables is harder to predict than from mortgages because it is much more dependent upon contingent fees. Therefore, the asset-backed securities are likely a smaller percentage of total revenue lest there be a shortfall.
Third, card issuers only securitize the receivables; they do not securitize the accounts. A mortgage securitization involves the sale of the entire loan. For cards, however, the issuer retains the account, and hence controls the terms of the account.
These three factors explain why rate-jacking makes a lot of sense. The card issuer holds only a limited piece of the downside: say there are $100 is asset backed securities and the issuer has a 10% pari passu stake. If there is a revenue shortfall of 10 for the securitization trust, the issuer loses only 1. But if there is a revenue surplus of 10, the issuer receives 20. Credit card securitizations give issuers imbalanced upside benefits and downside risks.
This situation is as if the issuer has both a call and a put option on the receivables: the issuer can buy the receivables at 100, even if they're worth 110, but can sell them at 99, even if they're worth 90. In this situation, it's basic option pricing theory that increased volatility increases the value of the options. And because card issuers control the terms of accounts, they can exploit this upside/downside imbalance. If issuers jack up rates, it will mean greater volatility for the receivables. There will be more cardholders who default, but the ones who don't will pay more. With the 10% hypothetical pari passu stake, the issuer could incur a 10% increase in defaults for every 1% increase in revenue and still be revenue neutral. And don't forget that issuers have tremendous volumes of consumer data on which to base volatility predictions.
With an imbalanced deal like that, who wouldn't jack up rates? Indeed, the question is why this isn't done more often. Part of the answer is that not all card issuers engage in securitization; generally it is only the largest ones. Part of the answer is that not all card receivables are securitized even for the issuers that do engage in securitization. Part of the answer is that card securitization isn't as simple as I just described. Sometimes there's a first loss position for the issuer as well, which reduces the rate-jacking incentive. And part of the answer is that too much rate jacking will also result in account closings as well as defaults. But I think this starts to explain the rate-jacking phenomenon.
Finally, it might be that the market understands this and discounts the purchase price of the ABS, so that rate-jacking is value neutral to the issuer (or even value negative, if it doesn't rate-jack enough). But that doesn't mean that it is value-neutral to the consumer, of course.
Notice that none of the proposed legislation gets at this problem. It isn't just a matter of retroactive rate increases. It is a matter of lock-in and rate increases in general and also of the application of back-end fees. I'm excited to see regulatory initiatives pick up steam, but I worry that they might miss the forest for the trees. The card industry's problems aren't a few ugly billing practices so much as a much more deeply seeded price structure problem that discourages transparent pricing.
http://www.creditslips.org/creditslips/2009/04/why-card-issuers-engage-in-ratejacking.html#more