Brokerage houses have many shades of "buy" and "sell" ratings. Here's what they really mean. By David Landis, Contributing Editor July 31, 2008 In his new book, Full of Bull: Do What Wall Street Does, Not What it Says, to Make Money in the Market (FT Press), former analyst Stephen McClellan says most top analysts do more marketing than research. Fill us in on how to interpret "buy" and "sell" calls. It's hilarious. You need a code book. When a stock goes from a very rare sell rating up to a neutral, that is a strong buy. When a rating is lowered from a buy to a neutral or a hold, that's a very strong sell. Obviously, there's a reticence to ever use the big, bad s word because institutional investors and corporate executives, the two biggest audiences a brokerage has, would object vociferously. In fact, only 5% of all ratings on Wall Street are sells. Analysts also run in packs. If there is an outlier (or two) whose opinion is different than the bulk of opinions, pay close attention because it's likely to be insightful, original, creative and early. Do brokers who advise individual investors understand the code? Not much. They're almost like individual investors themselves, and they adhere to the firm's party line in recommending stocks. But brokers who have been around a long time have a better sense of which analysts they can trust, which ones have some credibility, who is best at stock picking. They will gradually focus on those analysts and disregard others. Advertisement You were on Institutional Investor magazine's All-American Research Team for 19 straight years. How good is that list at identifying good analysts? The ones who are on the Institutional Investor team probably know their industry in more depth than anybody, and they are probably at least average at stock picking. The list is a reasonably decent screen of the better analysts on Wall Street, but it is a popularity ranking. If you're good at marketing, you can make a big impact. I always thought you could take an actor from central casting, give him the basics on how to do research, and he'd do pretty well. You advise investors to own a portfolio of no more than five to ten stocks. Is that enough? I don't think an analyst on Wall Street whose job is to cover a sector in depth can really track 15 or 20 companies, and for an individual it's impossible. I own fewer than six stocks, and I pay a lot of attention to them.