Streetwise: If they say `sell,’ pay attention

In a market going up for so long most Silicon Valley millionaires can hardly remember what down looks like, bad news is not always easy to come by. By bad news I’m not talking about forgetting to recharge the Palm Pilot or realizing that the Range Rover needs some work.

No. I’m referring to sell recommendations, those rare summaries by equity analysts after the most careful and diligent study of a company whose stock you might happen to own right at this very moment.

What is interesting about these calls to sell, aside from the fact that they make up only about 1% of all analyst recommendations, is that they can be highly effective portfolio management tools. Even, perhaps, more effective than buy recommendations.

study can pay off

Kent Womack, associate professor of finance at the Tuck School of Business Administration at Dartmouth College, has tracked the impact of analysts’ recommendations back to the early 1980s.

He makes a good case for why sell recommendations provide a valuable perspective on a stock’s potential performance.

For starters, he says, all sorts of reasons exist why an analyst might be biased when issuing a buy recommendation. Some of those reasons fall into the category of knowing your source, as in is the analyst linked with an investment bank that’s itching to do some underwriting ?

But, generally, in a market so heavily weighted toward good news, the analyst can be serving multiple constituencies, from nervous stockholders to hypersensitive executives. And, analysts know that a sell recommendation can mean lost or limited access to company information, making it nearly impossible to make recommendations of any kind.

“As an analyst, you don’t make a lot of friends by saying negative things too often,” Mr. Womack says.

Obviously, plenty of companies deserve sell recommendations. But in a world so rich with stocks that merit buy recommendations, why bother?

This is not to suggest that buy recommendations are not legitimate and justified; it is just that by sheer disproportion their impact is diluted.

Mr. Womack’s research shows, for example, that a buy recommendation will lift a stock by 5% on average over a concentrated six-week period.

Beyond that point, the recommendation has little impact. This follows, to some extent, the efficient-market hypothesis that new information is immediately factored into a stock’s price.

But, by comparison, when an analyst musters up the courage to single out a company with a sell recommendation, the average reaction is a 10% decline, spread out over a six-month period.

In Mr. Womack’s view, the steeper, longer-lasting effect of a sell recommendation vs. one to buy reflects some emotional attachment and some denial of news that should be treated with more respect.

In MBA parlance, this all makes sell recommendations more “predictive” than buy recommendations. Which, in regular parlance, means it takes longer for bad news to show up in the stock price.

But it will show up — if analysts have the guts to call it, that is.