Thanks to a diligent reader, a crucial adjustment to the stay-in-stocks strategy

By Peter Lynch

Last September's cover story, "Fear of Crashing," drew more response than any other article I've written. The subject was how to weather a stock-market correction, which many pundits said was imminent as the Dow hit 4700. By mid-February of this year, the Dow had risen to 5600, so a 16 percent correction would have taken it back to the 4700 that people were worrying about when the piece was written.

This supports the main point I was trying to make: Corrections are unpredictable. By selling stocks to avoid pain, you can miss the next gain.

The most thoughtful and challenging response to "Fear of Crashing" came from Scott Burns, a columnist for The Dallas Morning News. Burns found a flaw in my argument that an investor who needs income can use a stock portfolio as a bond substitute, boosting the yield when necessary by selling shares out of the portfolio. His column ran under the headline "Peter Lynch's Stock Theory May Be a Bust."

Here's the plan as I presented it. Let's say you have $100,000 to invest and you want to produce a $7,000 income stream. That's 7 percent a year. The normal course of action would be to buy a 7 percent bond that matures at a much later date: 20 years or longer. But suppose you reject that option and put the $100,000 into a portfolio of dividend-paying stocks or into a growth-and-income mutual fund that owns dividend-paying stocks.

At the end of the first year, you get $3,000 worth ofdividends from your portfolio, leaving you $4,000 short onthe income stream. You make up the difference by dipping into capital and selling $4,000 worth of shares inthe stocks or the mutual fund you bought 12 months earlier.

In financial circles, dipping into capital is a no-no, but in this case, if you accept certain assumptions, it makes sense. You have to believe that stock prices will continue to rise over time and that companies will continue to raise dividends at the same rate. You have to believe that stocks will produce a total return of 10 to 11 percent a year, on average, as they have in the past.

You might be forced to dip into capital for several years to engineer the 7 percent return, but as the companies in your portfolio raise their dividends, eventually you'll get $7,000 without having to sell any shares. Twenty years later, if stocks behave as advertised, the portfolio will be worth $349,140, so you will have more than tripled your money on top of the annual $7,000 you extracted. If you bought a $100,000 bond, the most you could hope for is your to get $100,000 back.

In a bull market, this plan is a cinch to succeed--last year, for instance, when the total return from stocks was 37 percent, you could have withdrawn $7,000 and still come out $30,000 richer. My chief concern was how it would work in a bear market. I asked a number cruncher to crank up the computer, and we tested two nightmare scenarios. In the first, stock prices drop 25 percent the day after you invest the $100,000. In the second, stocks and dividends rise at half the normal rate for the next 20 years. As I reported in "Fear of Crashing," the computer's verdict was thumbs up. In the first hypothetical case, you could take home your annual $7,000, and two decades later you'd be looking at a portfolio worth $185,350. In the second, you'd end up with $100,000, so you'd get back the same amount as you would have from a bond. This is where Scott Burns had his doubts.

With the help of a couple of brokerage houses, Burns tested the plan on the Standard & Poor's 500 and other stock indexes going back to the 1960s. Using data from real life, Burns found a worse worst-case scenario than my two imaginary ones: the Papa Bear market of the early 1970s. In that disaster, stock prices dropped a quick 40 percent and didn't regain their lost ground until the early 1980s. Any hapless investor who had bought at the top and followed the plan, withdrawing $7,000 a year, would have gone broke. By the time the next bull market rolled around, there would have been no money left in the account.

On the off chance that Burns's computers had caught a virus, I sought a second opinion on his second opinion of my first opinion. John McAllister, at the Boston-based Keystone group of mutual funds, agreed to run a test on Keystone's Growth and Income Fund, also known as S-1, which goes back to 1935 and has had a habit of paying a regular dividend. The results were disappointing. If you put $100,000 into S-1 at the peak of the market in January 1973, and extracted $7,000 every year thereafter, you were penniless by 1991.

Burns deserves kudos for bothering to figure this out. Clearly, it's not safe to withdraw $7,000 from a stock portfolio or a stock mutual fund if you had the bad luck to buy on the eve of a 40 percent correction leading to a 10-year bear market.

Now, I chose this 7 percent figure arbitrarily. If I'd given it more thought last fall, I would have remembered that hospitals, museums, universities, etc., customarily take a 5 percent annual draw, as they call it, from their endowments. The fiduciaries who manage those endowments are a cautious bunch. They must have chosen 5 percent for a reason.

With that in mind, I asked Keystone to test a 5 percent annual draw from the S-1 fund, assuming one had the misfortune of buying at prePapa Bear market prices in 1973. After five years, a $100,000 investment was reduced to $52,671, but eventually, stocks rallied enough to overcome this double whammy of declining prices and regular withdrawals. After 20 years, the portfolio was worth $107,653, so at least there was a $7,653 profit.

Granted, a 5 percent annual return on a $100,000 investment over 20 years, plus a $7,653 capital gain, is nothing to marvel at. But we're talking about the worst-case scenario since the crash of 1929. Stocks in 1973 had a long way to fall, because they were selling for ridiculous prices--for instance, 90 times earnings for Polaroid, 83 times earnings for McDonald's, and 76 times earnings for Disney.

Today, the Dow is selling at 16 times its 1996 earnings as projected by Wall Street analysts, which puts it in the middle of the range of 10 to 20 times earnings, where it has roamed for 50 years. If the Dow went to 15,000 next week and you put $100,000 in the market, I guarantee you'd be unhappy with the results of a 5 percent withdrawal plan--or a 0 percent withdrawal plan, for that matter. Then the market would be selling for 45 times earnings, and a nasty correction would be inevitable. Otherwise, you'd have stock prices going sideways for 12 to 15 years while the earnings caught up to them. As long as people are willing to pay foolish prices for things, no plan is foolproof.

That said, the 5 percent withdrawal plan seems to work well at least back to 1960; in the worst case you made a little money and in the best case you made a lot of money. If your timing was right and you bought S- 1 at the beginning of the bull market in 1982, then by the end of last year you would have had $360,314.

By the way, Burns wrote a sequel to his second opinion, in which he tested a 6 percent withdrawal rate on the Lipper Growth and Income Fund Index, going back to 1965. In the worst-case scenario involving that index, if you invested $100,000 and withdrew $6,000 a year, you ended up with $133,869, and in the best-case scenario, you amassed $914,682, which is a big improvement over the $100,000 return of principal from a bond.

---------------

THE BASICS HAVEN'T CHANGED

As long as I'm revisiting an old topic, I can't resist a chance to repeat some key points in brief, in case somebody out there missed the sermon. Here are some things to think about.

*If timing the market is such a great strategy, why haven't we seen the names of any market timers at the top of the Forbes list of richest Americans?

*People who exit the stock market to avoid a decline are odds-on favorites to miss the next rally. If you don't believe corporate profits will continue to rise, and you can't stomach a decline in the market, don't buy stocks or equity mutual funds.

*If you were out of stocks in 40 key months over the past 40 years, your annual return on investment dropped from 11.4 percent to 2.7 percent. You underperformed your savings account.

*In this century, we've had 53 corrections of 10 percent or more, roughly one every two years. We've had 15 corrections of 25 percent or more, roughly one every six years. These setbacks are normal and come with the territory.

*A stock certificate is not a lottery ticket. Behind every stock is a company. Stock prices go up 8 percent a year, on average, because corporate profits go up 8 percent a year. Add in the dividend yield of 2.5 percent (today's levels) and stocks give you a total return of 10.5 percent. Dividends are raised, on average, by 8 percent a year, right along with corporate profits.

*Even if we go into a long economic slump during which corporate profits grow at only half the normal rate, or 4 percent a year, stock prices should follow suit, rising an annual 4 percent a year. Assuming the 2.5 percent dividend, you would still get a 6.5 percent return, which is better than a 6 percent bond.

*Stocks outperformed bonds in eight out of the nine previous decades in this century, and they are well ahead halfway through this one.

*Since 1965, if you bought stocks once a year and were unlucky enough to pick the worst day to invest (when stocks were at their highest prices) 30 years in a row, you ended up with an annual return of 10.6 percent. If you were incredibly lucky and invested on the best day of the year 30 years in a row, you ended up with an annual return of 11.7 percent. So the difference between perfect timing and horrendous timing is 1.1 percent. This timing business is much ado about very little.

*In a correction or a bear market, great companies, good companies, mediocre companies, and terrible companies all see the prices of their stocks decline. A correction is a wonderful opportunity to buy your favorite companies at a bargain price.

Peter Lynch writes "Investor's Edge" with John Rothchild, and is vice- chairman of Fidelity Management and Research. Lynch and Rothchild's third investment book, "Learn to Earn," was recently published by Simon & Schuster. The opinions expressed in this column are strictly those of Peter Lynch, and do not reflect the opinions of Fidelity Investments.