Client Bias Lingers, Ratings Sober

In the disclosures buried at the end of most analyst reports, you'll often find that a disproportionate number of the firm's "buy" ratings are affixed to the shares of companies that are also clients of the brokerage's investment banking arm. Likewise, investment banking clients tend to be less represented among stocks rated "sell."

At Merrill Lynch & Co., whose reputation took a beating in the analyst scandal, only 8 percent of the 2,767 companies covered by the firm's analysts at the end of 2005 were rated sell, and banking clients accounted for just a fifth of those stocks. By contrast, about 40 percent of Merrill's ratings were a buy, and a third of those went to the firm's investment banking clients.

At UBS Investment Research, the overall percentage of sell-rated stocks was 11 percent, and just 30 percent of those were investment banking clients. Bear Stearns & Co. Inc. rated only 10 percent of the companies a sell and just 6.4 percent of those were investment banking clients.

By comparison, it appears that one of the firms that drew some of the sharpest accusations of tainted stock ratings may have less of a slant in its research than others.

Citigroup Inc.'s research operation, known as Salomon Smith Barney until scandal tarnished that name, reported that 42 percent of the nearly 2,800 companies rated by the firm's stock analysts at year-end were deemed a buy and 17 percent a sell. While investment banking clients accounted for nearly half of the buy-rated companies, but they also accounted for 37 percent, or 175, of the sell-rated stocks.

There are some compelling arguments to explain at least some of the lingering bias in favor of banking clients. For example, a prospective client might shun a brokerage if its research analyst already has a negative stance on that company's stock.

But such explanations only go so far, dampening any encouragement one might want to draw from what appears to be a more lasting shift in the stock research business: despite the passage of time since the scandal, analyst ratings remain far less rosy than during the bubble days.

At the end of 2005, slightly less than a third of all companies followed by Wall Street analysts were deemed a "buy" or "strong buy," while nearly 10 percent carried a negative rating, according to data compiled for The Associated Press by Zacks Equity Research.

That's in sharp contrast to the height of the market bubble six years ago this month. Back then, more than 2,800 companies, or nearly two-thirds of the roughly 4,400 firms covered by analysts, carried a positive recommendation. And less than 1 percent _ a mere 50 companies or so _ carried "sell" ratings.

Roughly 1,500 carried a neutral rating in March 2000, which in theory meant those stocks were worth holding. But in the twisted mentality of that era, neutral was often the not-so-secret code word for sell, as analysts were pressured by employers to avoid negative opinions on investment banking clients or even willing collaborators in the pitch for lucrative fees from mergers and initial public offerings of stock.

Perhaps more telling than a comparison with the market peak in 2000 is that the buy-sell-hold ratio has held fairly steady since late 2002.

It was then that Wall Street's biggest firms, chastened by charges from New York Attorney General Eliot Spitzer and the Securities and Exchange Commission, agreed to sever all interaction between their research and banking operations. It was also right about then that the stock market hit bottom in the post-bubble plunge.

That stock ratings have not reverted to an exuberant state could be taken to suggest that Wall Street's contrition has had substance beyond the $1.4 billion it paid to make the scandal go away. Even as stocks rallied in 2003, and merger activity picked up in 2004 and 2005, the breakdown of analyst ratings has held somewhat steady.

But while a welcome dose of sobriety seems to persist in the overall scheme of stock research, individual investors should still be mindful of who they listen to for advice. The stock ratings at the major brokerages still demonstrate a pattern that's hard to justify despite efforts to prevent conflicts of interest.

At least there's more transparency to the process, and any investor with the patience to read their advisor's disclosures will be better armed in deciding whose research to follow.