One of the first rules of investing is that if something seems too good to be true, then it probably isn't true.

Which brings us to the idea floating around Washington that we can fix the Federal Deposit Insurance Corp.'s busted deposit insurance fund -- at no cost to anyone -- by having the nation's banks, many of which are themselves busted, buy $25 billion of FDIC preferred stock.

It sounds great, like printing money in the basement and being able to get someone to accept it. You recapitalize the deposit insurance fund by putting $25 billion into it, but you don't charge the banks anything because they can count the preferred stock as an asset instead of a deposit insurance premium. Premiums, of course, have to be counted as an expense when banks compute their profits.

Sounds too good to be true, doesn't it? Because it's not true. It's not a solution. It's like having two drunks -- the insolvent insurance fund and the banks that are insolvent -- trying to prop each other up.

The problem is simple: The $25 billion that the plan would raise isn't capital -- it's just borrowed money. The preferred stock would be a liability on the FDIC's balance sheet, offsetting the $25 billion of cash. So at the end of the day, the FDIC wouldn't be any less insolvent than it is -- it would just have bigger numbers on both the asset and liability sides of its balance sheet.

Even though accountants count some kinds of preferred stock as capital, that wouldn't be the case here if everyone is reasonably honest. Here's why. If the FDIC has to pay a reasonable dividend on the preferred stock or has to redeem it some day, then the $25 billion can't possibly be capital -- it's a loan.

If the preferred stock really is capital because the FDIC doesn't have to pay dividends on it or repay it, then the stock isn't worth face value. And banks should have to charge off some or all of it against their earnings and capital, like any loan to an insolvent borrower.

So rather than being a solution, the preferred stock plan is part of the problem: the idea that borrowing will fix the deposit insurance fund.

Think of it as writing a $10,000 check on your Visa account to repay the $10,000 that you owe MasterCard. You're not any less under water than you were. You're just in hock to a different lender.

If you talk to the intellectual godfather of the preferred stock plan, Lowell Bryan of the McKinsey & Co. consulting firm, you discover that Bryan has never presented this plan as a cure. Bryan, the head of McKinsey's bank practice, says the preferred stock plan is a temporary expedient to hold the fort until the banking system and deposit insurance are changed to limit the deposits that would be insured, the interest rates banks could pay on such deposits and the loans they could make with the money. Don't hold your breath.

The point of the preferred stock, he says, is to "recapitalize the FDIC so it can invest in solvent but under-capitalized institutions in partnership with outside investors." In other words, none of the $25 billion could be used to bail out depositors of failed banks, which is what the FDIC needs money for.

Notice that Bryan is talking about the real, long-term solution to the problem: Bringing in new capital -- real capital, billions of dollars worth -- to recapitalize the banking system.

Where will this come from? The solution may well involve what FDIC Chairman L. William Seidman has been pushing for months: allowing corporations to buy banks.

As things stand now, any corporation that controls a bank is subject to Federal Reserve Board regulation under the Bank Holding Company Act, and can't engage in any business that's not closely related to banking. This means, for instance, that General Electric Co. or Exxon Corp. can own a savings and loan association but can't own a bank.

"The Bank Holding Company Act is archaic, and ought to be done away with," Seidman said in an interview. "There are only two constituencies for keeping things the way they are -- the bankers who go to sleep at night and know that Mr. {Carl} Icahn won't be on their doorstep in the morning, and the Federal Reserve Board, which will lose its franchise" if the laws are changed.

The Fed, obviously, has a different view, which we won't go into now. And there are problems with letting corporations into banking, such as the risk of putting too much economic power in the hands of a GE or an Exxon, or having sleazy corporations misapply federally insured bank deposits the way sleazy S&L operators misapplied federally insured S&L deposits. But these are problems that can be dealt with.

Regulators are trying to jump-start the system to get banks to start making loans again. Regulators have reduced the amount of money that banks have to keep interest-free at the Fed, lowered short-term interest rates and liberalized accounting for bad loans. Even if these moves work -- and they may not -- they are temporary fixes for the banks and the deposit insurance system. A long-term solution involves pain: stopping insolvent banks from paying dividends, merging and closing banks, bringing in new investors with deep-pockets who will hack and slash.

Pretending to solve the problem by having the FDIC sell preferred stock to banks is worse than nothing. Not only won't it work, but it will allow politicians and regulators to declare the problem solved and to sweep it under a rug. Which means that the problem, untreated, will metastasize and one day emerge as the next S&L disaster.

Remember what happened in 1988, when the Federal Savings and Loan Insurance Corp., of blessed memory, tried to borrow its way out of insolvency by raising $10 billion in the capital markets? It didn't work for the FSLIC -- and it won't work for the FDIC, either.

Allan Sloan is a columnist for Newsday.