Government is saving the credit card industry from itself.

You'd think that this kind of customer-gouging would at least be profitable. It is not. On the contrary, it has been an unmitigated disaster. Last week, Advanta announced that it would shut down all of its customers' credit lines in June and close down its credit card business. Advanta's losses have already hit 20 percent—the worst-case scenario envisioned in the government's bank stress tests—and are certain to rise. Advanta, which two years ago actually issued a press release bragging about its "stellar" results, is now nearly worthless, with a share price at about $1.

What makes Advanta a telling and important example for the banking industry is that what happens when credit card issuers raise their rates to sky-high levels is largely uncharted territory. In general, the experience of credit card issuers that focused on marginal customers (such as Household Bank, bought by HSBC (HBC) and dragging down HSBC's results) is not promising. The difference with Advanta is that its customers were not "subprime": A look at a typical pool of Advanta cards shows that 80 percent of its customers had prime credit, and another 11 percent are what the industry classifies as "near prime." Half its customers have credit scores of 720 and above, the upper end of the prime range. What Advanta shows is that the strategy of starting off by offering attractive rates—Advanta focused on small-business customers—and then gouging your way to profitability may be even more ill-advised.

Bad as this is, this isn't the whole story. In addition to the reliable customers that banks rip off and alienate with arbitrary rate increases, there is another set of customers facing higher rates: the ones whom the banks want to alienate. These are people with overextended credit and declining credit scores. For these customers, the banks' fondest hope is that they will be ticked off enough by the new rates to pay off their balances.

For a single bank to raise rates to a sky-high level is a dangerous strategy. When many banks adopt it, however, it gets much worse. Customers who see the rates on all of their credit cards rise become much less able to pay any of them. If banks try to one-up one another's rate increases to get paid the fastest, the likelihood is that in the end fewer will get paid.

Do banks see this? I would guess they do. One reason I think this is the evidence of my own credit cards. Several of them have a "universal default" clause in their contracts, which lets them raise my rates if I'm late on any loan or credit card. None of them, however, have exercised that option: I suspect their data crunching tells them (correctly) that if all my rates were to suddenly rise, the likelihood is not that I would magically repay my debts but that I would no longer be able to afford to pay even the minimum.

For every bank to raise rates at once hoping to be first in line to be repaid—essentially what many customers are facing now—amounts to a bankers' suicide pact. Yet, pressed to deliver short-term results to prove that they are back on their feet and desperate to push risky borrowers into repaying credit cards while they still have money available, banks persist in playing the rate-increase game.