When it comes to curbing analysts' penchant for bullishness on the companies they cover, the horse is arguably out of the barn. But nearly two years after the stock market's top, Morgan Stanley ( MWD) is swinging the barn door shut.Monday, the firm announced that where it once had four stock ratings -- strong buy, outperform, neutral and underperform -- beginning March 18 in the U.S., Europe and Japan, it will have only three (which it will term overweight, equal-weight and underweight). More important, analysts will assign ratings based not on what they think of the companies they cover relative to the overall market, but what they think of them relative to the other companies they cover. In effect, this will result in a sharp rise in companies under coverage that garner the firm's lowest rating. "We expect to see the full spectrum on the ratings," says Dennis Shea, director of global equity research at Morgan Stanley. "Yes, we're going to have underweights." The firm will also publish percentages of how many companies fall into each rating category. That, too, will be a change -- Wall Street firms have shied away from publishing such percentages (and have often not given them in response to reporters' inquiries) because they would show how stunningly bullish analysts are. Shea wouldn't give the percentage of underperform ratings among companies Morgan Stanley has under coverage, but he did allow that "like everyone else on the Street, it's quite minimal." Putting that in context, only 1.8% of Wall Street analyst ratings are sell equivalents, according to Thomson Financial/First Call. Such optimism in the face of the stock market's long slide has made sell-side analysis something of a joke on Wall Street. Too often, research departments appear to pander to their firms' investment banking divisions, despite barriers that are supposed to keep the two separate. Buy ratings are also popular with trading departments, because it is easier to drum up business by telling clients about a cool new thing to buy, rather than telling them it's time to sell. This isn't just psychology at work, but math -- all of your clients can act on a recommendation to buy a stock, but only a portion can act on a recommendation to unload.