Despite the legitimate questions about analysts' forecasting abilities, professional investors still rely heavily on Wall Street's picks and pans. More than ever, though, investors want to feel confident that they're listening to the right analysts.
But how to tell? That's where analyst-rating service StarMine comes in. The four-year-old company takes the database of analysts' predictions culled by earnings tracker First Call and crunches the numbers to determine which analysts are the most accurate.
"DIFFERENT SKILLS." How does StarMine separate the chaff from the wheat? First, it takes into consideration that some analysts are better at predicting earnings, while others are better at picking stocks. "When you think about it, there are really different skills involved," says Joe Gatto, StarMine's CEO and founder. Coming up with accurate earnings estimates requires being detail-oriented, watching numbers closely, and updating models regularly, he points out, while picking stocks requires more of a big-picture focus on the direction of the market and coming trends.
StarMine has separate rankings for both skills. In 2001, the top three analysts for stock-picking (judged on the basis of the total returns earned on a hypothetical portfolio StarMine managed according to their recommendations) were A.G. Edwards' Joel Houck, who covers financial services; Credit Lyonnais' Bryan Maher, who covers lodging and real estate; and A.G. Edwards' Gregory Gieber, who covers industrial and tech services.
For accuracy of earnings estimates, the top three are Lehman Brothers oil analyst Paul Cheng, Goldman Sachs' paper and packaging analyst Mark Weintraub, and J.P. Morgan autos analyst David Bradley. (StarMine's free site also lists the top three analysts for stock-picking and earnings estimates in each industry.)
GROWING CUSTOMER BASE. Knowing the best analyst in an industry often isn't enough, says Gatto, who believes that investors really need to know the best analyst on each stock. StarMine determines that by crunching numbers for each public company. The disparity in skill level the same analyst can show depending on the stock in question is likely due to the fact that analysts typically devote most of their energy to five or so companies, while they might officially cover 20.
Given all the recent controversy surrounding research, it's not surprising that this year has been a very good one for the still-unprofitable StarMine. The number of institutional investors that pay for its Web-based service has nearly doubled, from 56 to 106. Fees start at $1,300 a month, making the service too pricey for all but the wealthiest individual investors.
Gatto's own experience as an individual investor convinced him of the need for an analyst-rating service, so for the average investor, a portion of StarMine's site (www.starmine.com) is free. Registered users can put in a ticker symbol and get a list of the top analysts on that stock as well as the current quarter's "SmartEstimate" -- SmartMine's earnings forecast derived from the estimates of the highest-rated analysts. The number of registered users to StarMine's free site has climbed more than 40% this year, from about 15,000 to 21,000.
CHECKING UP. Last October StarMine debuted a new tool that allows brokerage firms to monitor the skill of their own research teams. Six firms have signed on so far, including U.S. Bancorp Piper Jaffray, Credit Lyonnais Securities, and Merrill Lynch. (The latter was the focus of New York State Attorney General Eliot Spitzer's probe into analyst conflicts of interest, and it agreed in May to a pay a $100 million penalty.)
Robert Peterson, Piper Jaffray's head of equity research, says the service would help the firm "continuously monitor and improve upon" the quality of research. But David Lichtblau, StarMine's vice-president for marketing, believes these firms will eventually use the independent service to help determine analyst compensation. "It's definitely the direction people expect to go," he says.
One might think First Call would be threatened by StarMine's encroaching on its analyst-tracking territory. But First Call's director of research, Chuck Hill, says he considers StarMine's service "complementary," not competition, since First Calls aims to be objective. "Their approach has been sound, and to my knowledge their results have been pretty good," he says.
HIGH SCORERS. Hill says some large investment firms no doubt already replicate StarMine's service in-house. But Piper Jaffray's Peterson says StarMine was the only independent source for this kind of information and was "far more powerful and cost-effective than building a system internally."
Along with helping investors pick stocks and brokerage firms monitor analysts' performance, StarMine's own analysis also sheds light on how all investors should use Wall Street research. Certain analysts consistently score better in StarMine's rankings, says Gatto.
According to StarMine, analysts who earn a five-star ranking for their earnings estimates (the most accurate get five stars, the least get one star), are three times more likely to stay in the top 10% than fall to the bottom 10%. Plus, when StarMine's "SmartEstimate" varies from consensus estimate by more than 2% (either above or below), it gets the direction right 71% of the time, says Gatto.
TOO OPTIMISTIC. What about investment-banking relationships -- one of the key issues that helped to put research's credibility in question? Gatto says they're important, but not to such a degree that you should ignore the calls made by analysts at firms with a banking affiliation. StarMine's research shows that those analysts are often tops at coming up with earnings estimates (perhaps because they have closer ties to management).
But not surprisingly, they tend to be too optimistic in their recommendations -- throwing off their scores for stock-picking ability. StarMine's Lichtblau says it's best to avoid stocks where the bank-affiliated analyst has the highest recommendation.
As controversy over research continues to rage, StarMine's promise to help investors -- pros and individuals -- determine who they should listen to puts the company in the eye of the storm. That's not a bad place to be, given the 41-person outfit's goal of being profitable by yearend. If current trends continue, it seems well on its way. By Amey Stone in New York