Why Analysts Might Be Worth a Listen

Resisting cynicism is always a challenge. Nowhere is it harder than in dealing with Wall Street. And nothing on the Street inspires more cynicism than the terabytes of ''research''--stock reports, analyst opinions, profit forecasts--that firms flash to grab attention. Long ago, I concluded that research from firms with stock to sell is so riddled with conflicts of interest as to be worth zero of my time.

Now, I'm finding I may be wrong. A growing pile of academic studies strongly suggests that the Street's stock recommendations do hold value. Nothing says that the familiar conflicts, such as firms praising stocks they've underwritten or giving big clients first word of a downgrade, have vanished. Yet consensus recommendations ''are actually very informative,'' Stanford University accounting professor Maureen McNichols told me. ''Given the amount of money Wall Street spends on research, at some level it just makes sense that it's not all marketing for investment banking.''

McNichols isn't just using common sense. With co-authors from the University of California at Davis and Berkeley, she makes the case in a paper (downloadable at set to be published in the Journal of Finance. They used a huge database from Zacks Investment Research, covering more than 360,000 stock recommendations from 4,340 analysts at 269 firms from 1986 to 1996. Zacks, whose reports many investors read at such popular Web sites as Yahoo! Finance, assigns a rating of 1 through 5 to distill every analyst's opinion on a stock. A rating of 1 means ''Strong Buy,'' while a 5 means ''Strong Sell.'' Then, Zacks averages all the ratings to find the consensus.

Next, the researchers measured how well portfolios made up of stocks in each of those five groups performed. The results are startling (chart): Strong Buys returned an average of 18.8% a year, while Strong Sells returned 5.8%. The broad market's yearly average was 14.5%--some four percentage points below that of the top-rated stocks.

Before you rush off to invest in a list of Strong Buys, hang on for what the researchers found next: After subtracting trading costs, the gap between analysts' buys and the market's performance disappeared. Recommendations on the most liquid big-cap stocks also held no advantage. The study assumed swift action on any rating change, capturing the stock's closing price that same day.

STALE? That's important. Despite the Web's aura of instant information, the truth is this stuff can be a week old, or older, at lots of popular sites., for example, updates data from Zacks only once a week, as do America Online (AOL) and (TSCM), which get their numbers from Zacks's rival, First Call. Some other spots, including Yahoo! (YHOO), Standard & Poor's Personal Wealth (like Business Week, part of The McGraw-Hill Companies), and Zacks's own site (, refresh data daily.

So what good is any of this to you? If you plan on trading and paying a commission anyway, ''this is another variable that makes sense to look at,'' said one of the researchers, UC-Davis finance professor Brad Barber. If you're eyeing a stock to buy, for example, you might gain confidence if it has recently been upgraded. Or if you aim to sell one, all else being equal, choose a stock that has recently been downgraded.

Now I'm curious to see what these researchers discover next. ''The magnitude of our results, based only on a naive trading strategy, is provocative,'' Barber told me. ''They suggest there might be even better strategies.'' A second study found no evidence that the brokerage firm with last year's best stock recommendations will predictably prevail this year. But what about individual analysts? Will the researchers find that yesterday's winning analysts reliably pick tomorrow's winning stocks? That's the question they're asking now. To their answer, I'm keeping an open mind.

Questions? Comments? Send an e-mail to or fax (321) 728-1711


_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Why Analysts Might Be Worth a Listen

CHART: Smart Calls

E-Mail to Business Week Online

Copyright 2000-2009, Bloomberg L.P.
Terms of Use   Privacy Notice