But banks’ ordinary businesses may be just as dangerous.

Citigroup (C) last week got rid of its $100 million man, Andrew Hall, selling his Phibro commodities trading group to Occidental Petroleum (OXY) and so sidestepping the headache of explaining how a taxpayer-supported company can pay one trader well over 2,000 times an average family's income. Yesterday the New York Times threw in the parting gift of an editorial that fellow TBM contributor Heidi N. Moore nicely summarized on Twitter as “good riddance.” The Times believes, as do others such as former Federal Reserve chairman Paul Volcker, that banks (which we now see the government has little choice but to bail out) don't have any business being in high-stakes speculative trading businesses in the first place.

There are some genuine arguments for this. For more than 60 years, commercial banks such as Citi and Bank of America (BAC) couldn't engage in the speculative trading that investment banks do because of the Glass-Steagall. This may have been a good thing, because businesses like those of investment banks and hedge funds have an unfortunate tendency to blow up. We've seen that happen many times even before the financial crisis (I wrote about how the seemingly venerable Lehman Bros. that blew up in this crisis was really a new company pulled out of the ashes of the earlier Lehman Bros. Kuhn Loeb that imploded in the 1980s).

The problem with this theory, however, is that it's not totally clear that the supposedly high-risk, speculative, overpaid cowboy businesses of the investment banks' trading desks are really that much more prone to disaster than, well, the plain old lending of the commercial banks. Lehman Bros. and Bear Stearns failed, Merrill came close, and the biggest losses at Citi did, in fact, come from the megabanks' bond underwriting and trading side. But, then again, the collapse of Wachovia and Washington Mutual—not exactly bit players—as well as huge losses at Bank of America and Wells Fargo (WFC) came out of plain old commercial bank lending.

Even if you look only at the trading houses, it's not at all clear that the casino caricature—traders happily playing a game in which they would reap huge profits if they got lucky and lose other people's money if they didn't—holds up. It is true that Lehman and Bear were brought down by holdings of complex and risky derivatives; long before this crisis, one major investor described Bear to me as “basically a giant hedge fund.” Yet at the same time, as Marc Hodak of the Hodak Value blog points out, few chief executives were as closely identified with their companies or stood to lose (and, in fact, did lose) as much as Lehman's Richard Fuld and Bear Stearns' James Cayne. The traders at those houses had every incentive to be at least as conscious of risk as anyone at a traditional bank: They knew they were gambling with their bosses' money.

So how did they not see them? A widely popular account of the financial crisis tends to paint what happened as a study in systemic risks, with interlinked deals that no one understood causing companies like Lehman and AIG to collapse like dominoes (Calvin Trillin spins out a brilliant comic version of this). So, the thinking goes, if we can just untangle that mass of impossible-to-gauge risks—or at least separate them from the trillions of dollars in government-insured deposits—we can prevent this from happening again.

Don't be so sure. This is a hopeful and comforting myth, but the problem with bubbles like the lending and housing bubble we just saw is that insanity appears normal, and boring commercial banks get pulled into them just as eagerly as the ostensibly high-flying risk kamikazes of the trading desks. We're still seeing this play out: It's not just the banks that the government is supporting but the vast guarantees and continuing bottomless losses at Fannie Mae (FNM).