The seminal question about any stock is when to sell it.

It's also the hardest question to get answered.

I've read hundreds of articles, whole books, even, on the subject, and it's a study in frustration.

So imagine my pleasure when, in reading a new book, "Secrets of the Investment All-Stars" (Kenneth A. Stern, Amacom Books, 1999), I encountered the following quote from Foster Friess, manager-owner of the Brandywine Fund:

"Our sell discipline is based on the 'pigs-in-the-trough' theory. Remember, I grew up on a farm in Wisconsin. If you watch the pigs, you will notice that when a pig approaches a group at the feeding trough, it has to shove aside a weaker or less hungry one to get in. The same survival of the fittest philosophy holds true for the stronger stocks that replace the weaker ones. In that regard, usually we will sell a stock when a better one comes along that we want to buy."

Oink. My sentiments exactly. And it has nothing to do with the fact that I, too, come from Wisconsin.

Many writers on investing treat the notion of selling stock as if it were a war crime. They've got a single answer to the question of when to sell: "Never," or its second cousin, "Hardly ever," or finally, "Only when you need to reallocate assets or seek some tax advantage."

They drum you out of the Investor Corps for even thinking about selling and condemn you disdainfully as a "trader," a lower breed. Horrors. Trading soon leads to day trading, which then leads to--I don't even want to think about it.

Some other experts you can't understand because they speak another language. When to sell? When the 200-day moving average equals the standard deviation in the Black-Scholes formula and the support level meets resistance from the long bond and Libor. Thanks.

Then there are the articles typical of journals like that of the American Association of Individual Investors, which I respect. They hem, haw, clear their throats, apologize and give you all the reasons not to sell. And then, finally, they say, when the stock no longer meets the strict criteria for which you chose it, then you may consider selling.

Let's face it, friends. We would all like to brag that we applied strict criteria to all our purchases. Congratulations to you if you have a system. But all of us may not have been such good little girls and boys. We've bought stocks on loose criteria instead of strict criteria or we've bought them on hardly any criteria at all.

If you deploy your crummy methodology for buying, it doesn't make sense to deploy it again for selling.

So I am very grateful for Foster Friess's pigs. I'm sure I'm not doing them, or him, justice. You should read the book to get his version. Here's mine--with the caveat that I am not urging anyone to sell anything, and certainly not without full consideration of the tax implications, which can be painful.

It's just that, in the face of the incantations of the investing priesthood, it takes courage to sell.

Most people I know buy stocks for one reason only: to make money. They seek no meaningful relationships with the companies they own. When the company's making money, the stock price tends to appreciate or pay ever-increasing dividends and it's worth holding on to.

If it's a large-cap company, well known and well followed, and if the market's trend is upward, I personally would not tolerate deterioration or stagnation or underperformance compared with other available stocks for more than a year, two years at most, unless there's a good reason.

There are good reasons. One, for example, is that the company is spending money to make more money in the future in a way that may turn off fund managers or analysts but which, in the long run, may make the company more competitive. I would not be selling Amazon.com, for example, the way many have, merely because it's losing ever-larger amounts of money because of its spending on advertising, among other things. I certainly would not have sold IBM because its mainframe sales were delayed by Y2K fears. If people want the mainframes, they'll buy them next year.

But if I can't find a good reason, I would sell and plow the proceeds into a company that is doing better and has prospects of continuing to do better.

Why should I hold on to a stock that's not growing, or growing modestly when others are growing by 10, 15 or 20 or more percent a year? Yes, I do believe in investing for the long term. But I also believe in making the best use of the finite sums I have for stock buying.

For example, I bought Sara Lee in November 1994. In the following two years, the price appreciated by about 45 percent. That's not bad. But by 1996 lots of investors were noticing that technology stocks such as Cisco Systems were doing much better. I sold Sara Lee and bought Cisco. Since then, Sara Lee has appreciated roughly 40 percent; Cisco's price has gone up by 390 percent. It would have been nice to hold on to both of them, but I couldn't.

What would you do? What's the best use of a sum of money? Needless to say, that was not a hard call. Needless to say, I haven't always been so wise. I drank Starbucks before it was fashionable, but I neglected to buy the stock.

Smaller, less well-known companies raise different issues and are often worth holding considerably longer even when they appear to be "rusting," in the apt words of Marcus W. Robins, president and editor in chief of the Red Chip Review (www.redchip.com), a small-cap-oriented newsletter.

A small stock that's down or inactive may simply mean that the marketplace has yet to understand it. If it's increasing earnings and management appears on course, it's worth a greater degree of patience, which may also bring a greater reward.

At the same time, Robins says, it's simpler to make the sell decision on a small-cap than on a large, well-known company. If a top executive, leaves, say, American Express, the company may not change much.

With a small, unknown company, he says, "if there's a departure or a personal crisis of the chairman, the president, or the CEO, that can stall a small-cap company for a long time," like possibly forever, or worse.

Another sell sign with a small unknown may be when it becomes known: "Sell the company when it's truly discovered. If it's a small company that didn't get followed, and today they're getting written up by Merrill Lynch, you have to believe that all the good news that's to be had is in the price of the stock."

People who specialize in small companies, mind you, often have very different investment objectives. I may be hunting optimistically for stocks that look as if they will appreciate 10 percent a year for a long time.

The hunters of small-cap growth stocks are looking for doubling or tripling or quadrupling.

Your objective is important in the sell decision. Some people buy stocks to make money to buy better stocks. They actively speculate (trade), or they see such an opportunity to make a decent return relatively quickly, for example, a great company beaten down--such as Motorola just a year ago. It's gone from about $40 a share then to $114 now.

When I was trading stocks 18 months ago (my job as business editor precludes trading now), I found irresistible 15-point jumps in Internet stocks over the course of a week or two.

But the main goal is rationality and common sense. The single best book I've read in pursuit of those elusive qualities with regard to all things financial is "Why Smart People Make Big Money Mistakes" by Gary Belsky and Thomas Gilovich (Simon & Schuster, 1999).

The book is a review of the field of behavioral economics. It was like a trip to a good therapist--rather than preaching at me, it merely suggested that whenever I'm confronted with a decision, I question my reasoning and understand that I'm subject to natural human biases that don't always make sense.

It reminded me that inaction--that is, not selling--is action, that perfectly smart people have a problem letting go of an investment just because they have it.

Breaking up is hard to do.

The fact is that some stocks just do better than others. If you're convinced you've got one of the losers, concede your mistake, dump it and find something better. Don't be afraid to take a loss. Don't be afraid to take a profit either. And don't forget to subtract the capital gains tax you will pay when you figure out whether it's a good move.

It's true, you might sell the next Dell Computer.

On the other hand, you might sell the next Iridium (the satellite phone people, now in Chapter 11 protection).

This is the quintessential problem with the "when to sell" question. Our experience teaches us nothing. I know that doesn't sound right. We're supposed to learn from our mistakes.

But for every share you shouldn't have sold, there's probably another you should have. As soon as you "learn from your mistakes" and do the opposite thing, the opposite thing becomes the mistake.

If you're too afraid to risk a mistake, you'll always be paralyzed. Don't look back.