There'll Always Be A Soros
By Paul Krugman

(FORTUNE Magazine) – Say it ain't so, Josephine! When some lovable grandmothers from the heartland published The Beardstown Ladies' Investment Guide, their book became not only a bestseller but a cultural phenomenon. Now it seems that their claimed 23.4% investment return, unlike the old-fashioned recipes in their book, may have included an unorthodox and questionable ingredient: membership dues. It's like the old joke about the investment guru who advertises, "Send me $10, and I'll tell you how I became rich!" So you send him $10, and he sends back a note saying, "Thanks--I got rich because of people like you."

One lesson from the Beardstown brouhaha is, never trust business bestsellers. You might think people who buy business books would be a hard-headed lot; instead it seems that business--like weight loss--is a subject wherein hope and fear inspire limitless gullibility. True, the public seems finally to have tired of management fads, of being told how to reengineer the search for excellence in the total-quality chaos. But now the lists feature a slew of you-too-can-beat-the-market tracts, with the Beardstown Ladies being only the cutest example.

Which brings me to my second lesson: Don't trust anyone cute, or more generally, anyone who gets press because his life (as opposed to his work) makes a good story. Mr. A grew up middle class in the 'burbs, got good grades in college, and has turned out to be a very good investment analyst. Mr. B grew up in Greenland, spent four years in the French Foreign Legion, climbs the Himalayas, and claims to be able to beat the market. Guess who gets profiled in Greed magazine?

Even if the Beardstown Ladies hadn't exaggerated their record, nobody should have taken them seriously anyway. A history of success is not a reliable, or even useful, indicator that an investor knows what she is doing. After all, stock prices are extremely hard to predict (duh!), so the return earned by any investor, no matter how smart, always contains a large element of pure chance.

Everyone knows that, of course, but the consequences of such randomness can be hard to grasp. Imagine predicting the outcome of ten random events (say, which way the Dow will move over ten years). If you guessed right every time, you'd be beating odds of 1,024 to one. But suppose a thousand people--all clueless--were making predictions: The odds are almost two to one that at least one would have a perfect record.

So it's actually ironic that the Ladies' success may have been due in part to sloppy accounting; sheer luck would have been enough. When their book came out, some 13,000 investment clubs were registered in the U.S. Even if all those clubs had been investing blindly, it was a virtual certainty that at least one would do very well--and that, modern America being what it is, some publisher would step in to spread the good word.

The lesson applies more broadly though: At any given moment, a few money managers are much admired for their success. But isn't it possible that out of the thousands of stock pickers, these folks are simply the day's (or decade's) lottery winners? We've all heard of the brilliant manager who, after years of outperforming the market, unaccountably loses his touch. The question is whether he had any touch to lose, whether his brilliance was ever anything but a statistical illusion.

Does this mean that nobody knows anything? Not necessarily. But how someone makes decisions is often more important than whether those decisions turned out to be right. Suppose Mr. A says, "My research shows that stocks beginning with the letter Z tend to be undervalued because people pick stocks in alphabetical order," while Mr. B can't even begin to articulate his strategy. Well, I would trust A rather than B, even if A has a less impressive track record. A more concrete example: Most people regard George Soros' total inability to explain his investment philosophy as little more than an endearing quirk--after all, look at his track record. Hmmm....

I should end this piece with a disclaimer: As you might have guessed, I grew up in a middle-class suburb, and got solid grades in college. My record as a forecaster is far from perfect, but I'm very good at explaining my mistakes. Just send me $10, and I'll tell you how I supplement my academic salary.

PAUL KRUGMAN is a professor of economics at the Massachusetts Institute of Technology.