The herd of day traders is debating whether to buy Apple before Steve Jobs's keynote at Macworld Expo. But following the herd is a strategy that generally leads to getting trampled. Eric Savitz of Barron's spots a smarter strategy: Buy, and sell — or at least avoid — Barry Diller's IAC. Citigroup analyst Mark Mahaney says IAC has "few countercyclical hedges to protect against a potentially material economic slowdown in the U.S." What does that mean?

It means that Diller, who prides himself on spreading his bets, has done a bad job. The company's diversification into all fields of Internet business has proven a weakness rather than a strength — a fact even Diller now recognizes. That's why he's breaking the company up into bits, leaving an IAC that's focused on online content and advertising. The remainder, however, will still be too vulnerable to the U.S. economy.

Bezos, meanwhile, has proved the opposite of Diller. He's made one bet and stuck to it. And while e-commerce is no longer sexy on Sand Hill Road or on Wall Street, it continues to be a growth industry. The longer users have been on the Web, the more they tend to shop online.

His lone divagation from e-commerce — Amazon's panoply of Web services, popular with gullible startups — is an exercise in renting out a computing infrastructure he's built at great expense. It's unlikely to be a better business than e-commerce, but it scores Bezos valuable press at little cost.

Mahaney's stock analysis is not a shocker: The stock market undervalues persistence and overvalues change for change's sake. What's surprising is that so few people have taken this insight to the bank.