On Wednesday, the Federal Reserve announced plans to lend $40 billion to banks. By my count, it's the fourth high-profile attempt to rescue the financial system since things started falling apart about five months ago. Maybe this one will do the trick, but I wouldn't count on it.
In past financial crises - the stock market crash of 1987, the aftermath of Russia's default in 1998 - the Fed has been able to wave its magic wand and make market turmoil disappear. But this time the magic isn't working.
Why not? Because the problem with the markets isn't just a lack of liquidity - there's also a fundamental problem of solvency.
Let me explain the difference with a hypothetical example.
Suppose that there's a nasty rumor about the First Bank of Pottersville: people say that the bank made a huge loan to the president's brother-in-law, who squandered the money on a failed business venture.
Even if the rumor is false, it can break the bank. If everyone, believing that the bank is about to go bust, demands their money out at the same time, the bank would have to raise cash by selling off assets at fire-sale prices - and it may indeed go bust even though it didn't really make that bum loan.
And because loss of confidence can be a self-fulfilling prophecy, even depositors who don't believe the rumor would join in the bank run, trying to get their money out while they can.
But the Fed can come to the rescue. If the rumor is false, the bank has enough assets to cover its debts; all it lacks is liquidity - the ability to raise cash on short notice. And the Fed can solve that problem by giving the bank a temporary loan, tiding it over until things calm down.
Matters are very different, however, if the rumor is true: the bank really did make a big bad loan. Then the problem isn't how to restore confidence; it's how to deal with the fact that the bank is really, truly insolvent, that is, busted.
My story about a basically sound bank beset by a crisis of confidence, which can be rescued with a temporary loan from the Fed, is more or less what happened to the financial system as a whole in 1998. Russia's default led to the collapse of the giant hedge fund Long Term Capital Management, and for a few weeks there was panic in the markets.
But when all was said and done, not that much money had been lost; a temporary expansion of credit by the Fed gave everyone time to regain their nerve, and the crisis soon passed.
In August, the Fed tried again to do what it did in 1998, and at first it seemed to work. But then the crisis of confidence came back, worse than ever. And the reason is that this time the financial system - both banks and, probably even more important, nonbank financial institutions - made a lot of loans that are likely to go very, very bad.
It's easy to get lost in the details of subprime mortgages, resets, collateralized debt obligations, and so on. But there are two important facts that may give you a sense of just how big the problem is.
First, we had an enormous housing bubble in the middle of this decade. To restore a historically normal ratio of housing prices to rents or incomes, average home prices would have to fall about 30 percent from their current levels.
Second, there was a tremendous amount of borrowing into the bubble, as new home buyers purchased houses with little or no money down, and as people who already owned houses refinanced their mortgages as a way of converting rising home prices into cash.
As home prices come back down to earth, many of these borrowers will find themselves with negative equity - owing more than their houses are worth. Negative equity, in turn, often leads to foreclosures and big losses for lenders.
And the numbers are huge. The financial blog Calculated Risk, using data from First American CoreLogic, estimates that if home prices fall 20 percent there will be 13.7 million homeowners with negative equity. If prices fall 30 percent, that number would rise to more than 20 million.
That translates into a lot of losses, and explains why liquidity has dried up. What's going on in the markets isn't an irrational panic. It's a wholly rational panic, because there's a lot of bad debt out there, and you don't know how much of that bad debt is held by the guy who wants to borrow your money.
How will it all end? Markets won't start functioning normally until investors are reasonably sure that they know where the bodies - I mean, the bad debts - are buried. And that probably won't happen until house prices have finished falling and financial institutions have come clean about all their losses. All of this will probably take years.
Meanwhile, anyone who expects the Fed or anyone else to come up with a plan that makes this financial crisis just go away will be sorely disappointed.
Paul Krugman is Professor of Economics at Princeton University and a regular New York Times columnist. His most recent book is The Conscience of a Liberal.
Copyright 2007 The New York Times Company