Financial reformers are pointing to the collapse of the $41 billion MF Global brokerage house as evidence of why we need Dodd-Frank’s “Volcker Rule” to prohibit FDIC-insured banks and their affiliates from making proprietary bets on the markets. Fortunately, MF Global was not a bank or bank affiliate, and its failure has not cost taxpayers a dime. And I, for one, am very happy to see a major, well-connected market player eat its losses (while being equally dismayed by the apparent regulatory lapses that have let hundreds of millions in customer money go missing).

But what if MF Global had been an FDIC bank? Would the Volcker Rule have protected the government purse? Well, it’s not clear.

MF Global took proprietary positions in European sovereign debt through what Wall Street calls “repo to maturity” transactions. It technically sold the European bonds to other firms, agreeing to repurchase them at a premium when they matured in 2012. MF hoped to make money by pocketing the difference in the rate it paid its trading partners and the higher rate paid on the bonds themselves. As market prices on the bonds fell, MF Global’s trading partners demanded more collateral. Given MF’s extreme leverage — about 40 to 1 — the collateral calls quickly brought it down.

Note that MF Global apparently used repo financing to move those high-risk bonds off its balance sheet, making its capital position look better than it was. Hmm. Didn’t Lehman also use the artifice of repo financing to dress up its balance sheet? Why don’t accountants fix this stuff?

But I digress.

Under the 300-page Rube Goldberg contraption of a regulation recently proposed by federal agencies to implement the Volcker Rule, “repo” transactions like MF Global’s are not generally treated as verboten proprietary trades. Thus, even if MF Global had been a bank, it arguably could have used this exception to gamble away, putting the FDIC at risk. Indeed, the proposed rule has so many loopholes that seem to permit proprietary trading with government-insured deposits that former MF Global CEO Jon Corzine might consider commercial banking for his next career move.

In fairness, regulators have a very difficult task. The basic intent of the Volcker Rule is to prohibit insured banks and their affiliates from making directional bets on the markets. But in 1999, Congress repealed Glass-Steagall, allowing common ownership of FDIC-insured banks and securities firms. And making (or losing) money on market movements is exactly the kind of thing securities firms do. To try to reconcile those conflicting policies, Congress granted exceptions from the Volcker Rule for traditional securities functions such as investment banking and market making. But where the exceptions end and banned proprietary trading begins is far from clear.

So here is an idea. Regulators should scrap the mind-boggling complexity in the proposed rule and focus instead on the underlying economics of a transaction. If the transaction makes money the old-fashioned way — the customer paying the institution for a service through interest, fees, and commissions — then it passes the test. If profitability (or loss) is driven by the direction of markets, then it fails. Inevitable gray areas, such as marketmaking, need to be done outside of the insured bank and be supported by a truckload of capital. Securities firms should be allowed to maintain adequate inventory to make liquid markets.

Most important, regulators should tell executives and boards that they will be held personally accountable for monitoring and compliance. Bank leadership must make clear to employees that they are supposed to make money by offering good customer service, not by speculating with the firm’s funds.

Complex rules are easy to game and hard to enforce. If regulators can’t make this work, then maybe we should return to Glass-Steagall in all of its 32-page simplicity.

—Fortune contributor Sheila Bair is a former chair of the FDIC.

This article is from the December 26, 2011 issue of Fortune.