When I look at the stock market these days, I can't help but think of what we used to say about the weather when I worked in Detroit during the 1970s: "Wait five minutes, it will change." You don't have to wait even that long, it seems, for investors' opinions about stocks to change. Last week we had the spectacle of big, widely followed stocks such as Citigroup, IBM and Intel moving more than 10 percent in a single day. And the market as a whole has been amazingly bipolar, lurching from euphoria to depression in the blink of an eye. Swings of 200 Dow points now seem routine. And you never know in which direction.

The market seems to delight in being contrary. It ended September on a horrible losing streak, finishing the worst calendar quarter since 1987. Enter October, renowned as a wretched market month: the Great Crash in 1929, Black Monday in 1987, among other debacles. Oops. Despite all the depressing postmortems on Oct. 1 analyzing the third-quarter slaughter, stocks rose from the rubble to kick off the quarter. But by Oct. 9, the market had hit its lowest point in five years. The Nasdaq composite, at 1950 when the year began, had plummeted past the Magna Carta level (1215, if you've forgotten) and was heading for the Battle of Hastings (1066). With despair in the air, the market turned yet again, producing its best four-day percentage gain since 1933.

Then it dropped sharply, rose sharply, treaded water and rose sharply again.

There is no shortage of explanations for each day's action. IBM and Citigroup reported higher-than-expected profits, so stocks went to the moon. Intel had bad numbers, and stocks cratered. If only markets really were that sensible, and each day's effect really had a rational cause. It's clear that the market is thrashing around, and it's being roiled on a day-to-day basis by momentum players. These people bet that whatever is happening will keep happening. This means that rises and falls tend to go to extremes, because lots of big players pile on, in either direction. Then there are rises and falls that have nothing to do with stocks' inherent value. About a third of the New York Stock Exchange's volume is "program" trades. This means that hundreds of millions of shares change hands daily because of stock prices' relationship to some sort of derivative security that big investors are playing with.

There is a reason I'm taking you through this extended explanation. It's to show you how futile it is to try to time the market, because good days (and bad days) can blow up out of nowhere, like summer storms. The spring of 2000, when stocks were booming and optimism reigned, was a terrible time to buy stocks; the market has since fallen more than 40 percent. But if you bought on Oct. 9, when the market and the news were horrible, you've done spectacularly well. You can trust your gut for lots of things, but with stocks you do so at your peril. When you feel the need to make a drastic change by piling into stocks or bailing out of them, invoke the anti-Nike slogan: Just don't do it.

Another point is that it's silly to spend a whole lot of time figuring out what stocks' daily gyrations presage for the economy. Sure, the market is often a leading indicator, a precursor of what's to come. But "often" isn't "always." That's why stock prices are only one of the 10 factors that go into the Conference Board's widely followed index of leading economic indicators. If stocks were an infallible indicator, we wouldn't need the other nine.

It's not clear whether we can draw any valid conclusions from the market's mid-October turnaround. This may be another "dead-cat bounce," a rally that will go thud like the one in August. On the other hand, when the great bull market started in the summer of 1982, a lot of people didn't believe it. Stocks had been essentially flat since 1964. I used to qualify the term "bull market" with "should one ever appear." In the long run, the market is rational. In the short run, the market is . . . the market. After the fact, the weatherman can explain with perfect certainty why the hurricane veered out to sea, or why it headed inland and took your roof off. But did he know what the storm would do before the fact?

However, I'm supposed to offer predictions. So here goes. Stocks are less risky than a few years ago, because they're so much cheaper. But it doesn't mean they're safe. How long can people keep consumer spending booming by refinancing their houses? Do disasters lurk in the loan portfolios of our giant financial companies?

So I'm going to imitate the weatherman -- the modern-day kind. In the 1970s, they used to go out on a limb and make real predictions. These days, they say "There's a 60 percent chance of rain" or "Snow is possible." So I will boldly say that there's a 51 percent chance the bear market is over. And a good chance that it's not. If it works for the Weather Channel, it works for me.

Sloan is Newsweek's Wall Street editor. His e-mail address is sloan@panix.com.