Analyst Whores; Qualcomm Soars

If there’s one thing to hope for after this day on which the Nasdaq leapt over 4,000 (hardly a month after it crossed 3,000), it’s that three months from now no one remembers Walt Piecyk’s name.

Piecyk is the Paine Webber telecom analyst who this morning put a 12-month price target of $1,000-a-share on Qualcomm, which closed yesterday at $503 a share. Piecyk’s call sent Qualcomm’s stock back off to the races. It opened at $560, and ended the day at $659, which is a leap of 31 percent. That’s respectable by any standard, but it’s especially remarkable when you consider that Qualcomm, before today, was already the best-performing stock in the S&P 500 Index, up about 1,800 percent for the year. (It’s now up better than 2,300 percent.)

I know, I know. These numbers are effectively gibberish, especially since I could fill this column every day with stories about stocks making incredible moves upward. So why does it matter or not matter whether in May the mention of Walt Piecyk’s name will be greeted with anything other than “Who?”

The simplest answer is that the success of calls like Piecyk’s–success here meaning just that investors buy on the news of your price target–leads to more calls just like it. When Henry Blodget, who was then at CIBC Oppenheimer, said that Amazon’s shares would hit $400 within the year (he was right), it catapulted him into national attention and, in some sense, probably got him his new job at Merrill Lynch. And other analysts noticed. When FreeMarkets, a hot new business-to-business company, went public a few weeks ago, an analyst at a firm that wasn’t underwriting the deal initiated coverage on the day of the IPO and set a price target of $300 a share, even though the company was going public at $48 a share. It’s become almost de rigueur, in fact, for analysts to make audacious calls. Otherwise they won’t even be noticed.

You can see the same phenomenon at work in analysts’ estimates of things like that fabled Internet business-to-business market. For a while now, the number everyone’s been tossing around is $1.3 trillion in 2003. But that means that if you’re an analyst and you cite that number, no one will notice you. So now the numbers are getting ratcheted up, to the point where a recent Goldman, Sachs study said the market could be as large as $2.5 trillion. (And, of course, here I am mentioning the report.)

The phenomenon of number creep is bad because we want analysts to reach conclusions they think are true, and not conclusions that they think are going to get them on CNBC. (I realize this is incredibly naive, but there it is.) And the truth is that there still isn’t enough accountability for these kinds of forecasts to make the rewards and punishments for success and failure commensurate. But the phenomenon is also bad because these kinds of calls have become market-moving events, which means that it’s no longer irrational to trade as a result of them, at least if you’re a trader.

According to a traditional definition of “information” about a stock (as in, a stock’s price reflects all the available information the market has about it), Piecyk’s report was not new. Moreover, when you consider that Qualcomm was already up 1,800 percent for the year, the appropriate question to ask is probably: “Why are you just starting to cover this stock now?” But if we broaden the definition of “information” to include everything that might affect that stock’s price, even in the short term, then the report has to be considered new information, which means that it has to affect the stock price. And that means that ignoring it is not necessarily rational, even if there’s nothing valuable about Piecyk’s analysis.

The best thing for everyone, then, will be for Qualcomm to settle back in the next couple of months, or even just rise slightly, and for Piecyk’s forecast to be forgotten. In the long run, even in this stock market, companies’ stock prices will find their true level. But what has become incredibly clear in 1999 is that the signal-to-noise ratio in the stock market is degenerating (with something like Piecyk’s report being noise), and that means that it will take longer for those true levels to be reached.