No one ever said it was easy to make money in the stock market. You have to earn your keep--especially at times like last week, when the Dow Jones industrial average fell more than 100 points on each of three days.

The hard work is to hold fast, to resist selling your stocks when all your nerves are screaming, "Get out!"

Certainly, there are reasons to sell. You should sell if the underlying business has deteriorated. Sell if a management change isn't working out, or if a product fails or if tough new competition enters the market.

Sell if a wonderful opportunity opens up and you want to exchange a good company for a much better one. Sell if your asset allocation has gotten out of whack and you own, for example, too many growth stocks and not enough bonds.

But don't sell because the price of a stock is going down. To sell then is to believe you can tell what the stock will do next. You can't. In the short term, the price of a stock--in the famous phrase of Princeton economist Burton Malkiel--is a random walk, which means that its next step is utterly unpredictable.

J.P. Morgan & Co. (symbol: JPM), the money-center bank, fell $3.75 on Thursday to $135.12 1/2. At the end of the day, the reason for the decline (presumably, worries about interest rates) was reflected in the stock's price.

Think of that price as the considered judgment of the thousands of players in the market who own or are thinking about owning Morgan stock. By selling at $135.12 1/2 when the market opened on Friday, you would implicitly be saying that you knew that Morgan would keep falling Friday. But how could you possibly have knowledge that the rest of the market lacked? (In fact, Morgan rallied on Friday to $139.31 1/4, thus more than making up for Thursday's decline.)

It is panic selling that causes investors to perform poorly in the stock market. In an important study, Dalbar Inc., a Boston research firm, found that, while the Standard & Poor's 500-stock index returned 820 percent from 1984 to 1997, the average investor scored returns of just 148 percent.

"The gap," the study concluded, "is explained by the behavior of equity fund investors. In their attempt to cash in on the impressive stock market gains, investors jump on the bandwagon too late, and switch in and out of funds trying to time the market. By not remaining fully invested for the entire period, they do not benefit from the majority of the equity market appreciation."

The urge to "time" the market--to act in the belief that you know where the next random step is headed--is especially strong today, thanks to a combustible mixture of high volatility and unrealized capital gains.

In other words, stocks seem to be jumping up and down wildly at the same time many investors have substantial profits in their portfolios and are sorely tempted to lock them in. So why not sell and take your profits?

For one thing, you'll have to pay a federal capital gains tax of 20 percent, plus state levies. Second, what are you going to do with the proceeds? Stocks are the best place to put your money in the long term. Third, you can't possibly tell when a stock has hit a top or bottom, but a casual survey of your friends will probably show you that their greatest regrets are selling companies that later soared.

One little example: The Gap Inc. (GPS). This wonderful company has taken several dives in its history. One occurred in 1994, after the stock had tripled in the previous three years. In 1994, Gap fell from $11 to $6.50. Time to sell?

By 1996, the stock was back up to $16, but again it dipped sharply, this time to $9.50. Time to sell?

In 1998, shares rose to $45 but then plunged to $30. Time to sell?

By last month, Gap had soared to $77.37 1/2, but by Thursday it had fallen to $60.75. Time to sell?

Or is there a lesson here?

The truth about stocks is that they are indeed volatile from day to day, month to month. Last week was a good example. On four of the five trading days, the Dow rose or fell at least 124 points. Friday broke the string with an increase of 92.81 points. The stretch over Labor Day 1998 (Sept. 4-14) was even more volatile: five 100-plus days in a row, three up and two down.

"The investor," wrote Benjamin Graham, the financial genius who was mentor to billionaire Warren Buffett, among others, "may as well resign himself in advance to the probability, rather than the mere possibility, that most of his holdings will advance, say, 50 percent or more from their low point and decline the equivalent of one-third or more from their high point at various periods in the next five years."

Actually, for many stocks, it's worse than that. High volatility can be compressed into short periods. Take General Electric Co. (GE), the quintessential steady earner. Year after year, GE's profits rise between 12 percent and 15 percent. Yet, over the past 52 weeks alone, GE's price has bounced between a low of $69 and a high of $117.43 3/4. GE closed Friday at $101.68 3/4.

There is no cure for volatility. It comes with the territory. But there is a way to deal with it: Ignore it. Or, better yet, as Buffett himself suggests, let it work to your advantage by purchasing stocks at good prices after they have been hammered.

If you pay too much attention to all the talk about the Dow Jones industrial average, you may be more frightened of volatility than you should be. With the Dow above 10,000, a drop of 100 points is a drop of less than 1 percent--the equivalent of a $50 stock falling by 50 cents.

The 235.23-point decline on Thursday was the 13th decline of more than 200 points since Jan. 9, 1998, and it was the smallest percentage decline in the group--just 2.2 percent, which means that a $100 stock fell to about $97.75.

Perhaps it would help assuage your fears if you understood why stocks are so much more volatile than such assets as residential real estate. The reason is that the value of a stock is the value today of all the cash that the stock will throw off during its lifetime, reduced by a discount rate.

Sorry for the jargon. What this means is that the market is constantly assessing the profits that a company will make in the future. And tiny differences in the rate of growth, thanks to compounding, mean a lot. Also, there's the discount rate, which is an interest rate somewhat higher than what Treasury bonds pay. That rate, too, is constantly changing. Lately, investors fear that it's going up.

Assume you win the lottery. Your payoff is $1 million, but the catch is that you'll receive checks of $50,000 annually over 20 years. What is that flow of cash worth today? In other words, what would an investor pay you in a lump sum right now in order to get $50,000 yearly for 20 years? The answer is about $500,000, but the figure can vary with your assumptions about interest rates. If rates are higher, then the present value is lower.

Now assume that you win the lottery, but your payoff isn't fixed at $50,000 annually; instead, it varies according to a formula that's linked to the profits over the next 20 years of Gap or GE. Clearly, it's much tougher to come up with a present value for such a variable flow of cash. This example comes close to describing the situation for the average stock, and it shows why prices bounce around by 1 percent or 2 percent a day. It's a wonder prices aren't more volatile.

This little model also explains why Internet stocks are particularly volatile--since many of them don't have any earnings at all, which makes predicting future profits extremely difficult. Look at Amazon.com Inc. (AMZN). Over the past 52 weeks, it has traded between a low of $13.75 and a high of $221.25. It closed Friday at $118.75, up about 4 percent--a fairly calm day for the online retailer.

The performance of a stock that produces a return (price increase plus dividends) of 10 percent every year for 10 years is exactly the same as that of a stock that rises 51 percent in half the years and falls 20 percent in the other half (it's true; try it on your calculator). But the latter stock is far more volatile.

Which would you rather own? In theory, if you have a long time horizon (that is, if you are saving for a retirement that's a decade or more away), it shouldn't matter to you. But in real life, it does. Stocks that are volatile encourage investors to misbehave--to sell on the dips rather than buying (or holding) on them. For that reason, it's smart to try to find stocks with a history of low volatility.

The past can't predict the future, but it's the best tool we have. A good measure of a stock's volatility is its "beta," which compares the extremes of its ups and downs with those of the market as a whole. You can find betas in the Value Line Investment Survey, available at libraries, or on Internet investment World Wide Web sites such as www.quote.com.

A beta of 1.0, as in the case of American Eagle Outfitters Inc. (AEOS), indicates that a stock is just as volatile as the market, while 1.70, as in the case of CompUSA Inc. (CPU), indicates that it is 70 percent more volatile. When the market drops 10 percent, CompUSA tends to drop 17 percent, and when the market rises 10 percent, CompUSA tends to rise 17 percent. Gap's beta, by the way, is 1.35; GE's is 1.1.

In addition to American Eagle, a specialty retailer of casual apparel whose earnings are growing at a rate of more than 30 percent a year, other low-beta stocks with high marks from Value Line include WestPoint Stevens Inc. (WPSN), maker of Martex and other brand-name towels and sheets, with a beta of 0.75, making it one-fourth lower in volatility than the market; the Pep Boys--Manny, Moe & Jack (PBY), with a beta of 1.0; The Limited Inc. (LTD), at 1.1 with a timeliness rating from Value Line of "1" (tops); and Montana Power Co. (MTP), which, like most electric and gas utilities, has a very low beta (in this case, 0.5).

But the most distinctive feature of volatility is inevitably. Live with it.

Glassman's e-mail address is jkglassman@aol.com; he welcomes comments but cannot answer all queries.