Everybody that is seasoned in the market know that analyst research and projections are designed with the purpose of providing institutional business to the underlying investment brokers. As a result, you very rarely hear analysts giving sell recommendations, unless if the firm in question has already burnt its bridges with the company in question.
At best, this research usually reflects the market consensus, plus or minus a few cents on the earnings per share projection. Analyst research is never the basis for making a trade – instead, it is a quasi-benchmark for how you think the company in question will really perform.
When an analyst is brave enough to see a company that is clearly overvalued and makes a sell recommendation, he is usually looking for an exit out of the firm. For example, a brave analyst called Brian Kennedy made a sell call on a company called CardioNet. He was basically forced out of the company, but his negative projections turned out to be correct.
Note if you cut-and-paste the URL into Google, you can view the entire article.
The moral of the story is that analyst research has a function; but the incentives that analysts have serve the investment banking arms of their companies. Anybody doing very sharp research is most likely trading on that information rather than releasing it to the public, as it is far more profitable to trade with superior information than to try to drum up investment banking business.