Bloomberg has a solid look this morning at stock-research and the conflicts that still pervade the Wall Street research business seven years after Eliot Spitzer’s campaign to clean that corrupt area up.

The big problem with Wall Street analysts, of course, is that they’re biased toward the upside. They very rarely put a “sell” on something, in large part because of stuff like this:

When Credit Suisse Group analyst Ivy Zelman refused to turn bullish on homebuilding stocks during a rally in the fourth quarter of 2006, the blowback was intense.

She says investors told her that some housing industry executives were ridiculing her analysis as a “jihad,” and several of the bank’s sales representatives pressed her to upgrade “hold” ratings to “buys” on companies to appease bullish institutional-investor clients. One sales manager even sent her an e-mail warning that analysts who stayed bearish too long often lost their jobs.

Is that even legal?

Bloomberg’s story is on the trend of analysts quitting Wall Street shops and forming their own research firms. It reports that independent research firms has soared to nearly 2,700 from just 1,000 three years ago.

I can tell you from covering public companies and dealing with analysts—though I knew some fine analysts at the banks—the best research (in commercial real estate, anyway) invariably came from an independent research firm.

But:

Today, Bank of America Corp., JPMorgan Chase & Co. and other giant securities firms receive almost 70 percent of the commissions institutional investors dole out for research. That compares with 3 percent for independents and the rest for mid- size firms, according to Greenwich.

Still the Street has changed from the bad old days. This number has been long reported, but with the benefit of a few years’ hindsight it’s just stunning: “Rock star analysts like Jack Grubman, the telecom specialist at Citigroup Inc. who earned $67 million from 1999 to 2002…” Now top analysts make about $600,000d—no pittance, of course—but orders of magnitude smaller than during the Tech Bubble.

But here’s another example of Wall Street analysts’ cluelessness or worse (emphasis mine):

In the fourth quarter of 2008, Betsy Graseck at Morgan Stanley recommended investors sell shares in Bank of America after concluding its credit card business and takeover of Countrywide Financial Corp. would saddle it with huge losses. Nineteen of the 21 analysts who covered Bank of America at that time had “buy” and “hold” calls on its stock, which nose-dived 80 percent from Sept. 30, 2008, to March 31, according to data compiled by Bloomberg.

That is just stunning. Presumably these include analysts at BofA and Merrill Lynch who surely knew better. There’s possibly a nice little story there, enterprising business journalists. Go ask Multex for those two’s research reports on BAC from the fourth quarter of 2008.

Bloomberg does well to quantify the upside bias:

In October 2008, as the global financial system teetered on the brink of collapse, “sell” calls in U.S. markets constituted 6 percent of the total recommendations by analysts, with “buys” comprising 36 percent and “holds,” 58 percent, according to Bloomberg data.

And today it’s even worse:

Almost a year later, amid a stock market rally, the percentage of “buy” calls dropped: They made up 32 percent, with “holds” comprising 63 percent and “sells,” 5 percent, as of Oct. 8.

Bloomberg covers a lot of bases here, bringing up the possibility that the Chinese wall on Wall Street isn’t exactly non-porous:

Money managers are concerned that proprietary trading desks at the largest securities firms are benefiting from research reports at the expense of clients. On Aug. 25, William Galvin, Massachusetts’s top financial regulator, subpoenaed Goldman Sachs for information on possible weekly “trading huddles” between its analysts, traders and investors.

Galvin wants to know whether Goldman’s analysts previewed imminent changes in their stock recommendations for select clients and whether Goldman traded on these tips for its own account before disseminating the information.

If you’re a reporter covering companies (or an investor) and you’re looking at analyst reports, it’s best to ignore the calls like “buy” or “sell” and just look for new information the analyst may have dug up.

So what happened to Zelman?

On Dec. 7, 2006, she slapped a “sell” call on the entire group, and during the next 12 months, the Standard & Poor’s Supercomposite Homebuilding Index plunged 53 percent as the real estate market collapsed.

Ryan Chittum is a former Wall Street Journal reporter, and deputy editor of The Audit, CJR's business section. If you see notable business journalism, give him a heads-up at rc2538@columbia.edu.