BusinessWeek e.biz -- Clicks & Misses
What's an Investor to Do?
Web sites can help you pick stock analysts to follow--and then get their reports
These are times that try online investors' souls. The estimated 13 million people who trade online--almost all of them through limited-service, deep-discount brokers--have spent the past six months or so standing beneath a torrent of cascading stocks. The natural answer: Turn to Wall Street's mandarins for help. But of the sharp suits who spend so much time on TV, which ones know what they're doing, and who's just part of the Hype Machine?
Naturally, some enterprising folks are trying to make businesses out of rating the experts. The idea is that when investors reach out for help, they should know who can be trusted. BulldogResearch.com and StarMine.com are two leading sites that give investors a sort of meta-tool: They analyze the analysts. They tell you which ones are likeliest to pick the best stocks--in up to 90 industries--and whose earnings estimates are the most accurate.
By themselves, these sites don't provide enough information upon which to base trading decisions. Most people should follow up by getting detailed analyst research from a site such as MultexInvestor.com, which distributes full reports from 250 brokerages and other providers. Some Multex reports are free, others cost up to $50. The idea: Use Bulldog or StarMine to identify experts worth listening to, then use Multex to get the full scoop on what the creme de la creme consider the cream of the crop.
An instructive example of how to use these sites together is DuPont's poor fourth quarter. StarMine's earnings-surprise alerts would have tipped you that DuPont's numbers would come in a little short. It also would have told you that the best DuPont (DD) analyst on the Street is J.P. Morgan Chase & Co.'s Don Carson (Bulldog rates Carson No. 2). But you wouldn't know why the quarter was short, or how long any dip in the stock would likely last. That could best be gleaned by reading Morgan's three-page follow-up note on Multex (where it was free if you signed up for J.P. Morgan's research trial). The report warns that the first quarter looks tough, too: High natural gas prices and a soft economy are giving the Delaware contingent a tough winter. So use a Bulldog or StarMine to get ideas, but use a Multex to get context.Innovative. Bulldog and StarMine use similar approaches to identify and highlight top analysts. Both start with data on recommendations and earnings estimates that they get from the Institutional Brokerage Estimate System (I/B/E/S), an established service owned by Canadian publisher Thomson Corp. But both massage the I/B/E/S data to use it more as a starting point for their own calculations.
San Francisco's StarMine is the more innovative. It rates analysts on a scale of one to five stars and tracks the movements of the best analysts to warn investors of short-term surprises. The key is StarMine's "SmartEstimate," a recalculation of Wall Street's consensus that gives more weight to numbers that have been revised more recently--and more weight to the highest-rated analysts.
StarMine claims these estimates are more accurate than Wall Street's consensus almost three-quarters of the time, predicting earnings surprises that make stocks fly up or down in volatile markets such as this one. (Now here's my question: Since StarMine is backed by Hummer Winblad Venture Partners, backer of such gems as Pets.com (IPET), HomeGrocer.com (HOMG), and TheKnot.com, why couldn't StarMine have shared its investment smarts at home? Fellas, where's the love? But I digress.)
StarMine called 16 earnings surprises correctly in a second-quarter test in which they supplied 18 early calls to the San Francisco Chronicle. To do my own test, in January I asked StarMine for some predictions for the fourth quarter. Of the 10 they forwarded, all but one of those that pertained to companies that reported quarterly results in time for my deadline were on or near the mark. StarMine was correct in predicting that DuPont would disappoint. They also spotted a big upside surprise from Phillips Petroleum (P), and they were correct on El Paso Energy (EPG) and Tyson Foods (TSN).
The big miss: Merrill Lynch & Co. (MER) The SmartEstimate said Merrill would earn 84 cents a share instead of the consensus estimate of 88 cents. The real number was 93 cents. For those who plan to keep score at home, StarMine says Harrah's Entertainment (HET) and Wal-Mart Stores (WMT) will post disappointing results after my deadline, and Williams-Sonoma (WSM) will beat estimates.
The other good thing about StarMine is that it puts its best ideas where you can find them. Go to the site, and there's a list of predicted earnings surprises and new "bold calls by five-star analysts" smack-dab on the first page. Predicted surprises are stocks for which StarMine's SmartEstimate is higher or lower than the Street consensus. The "bold calls" feature highlights situations where a single analyst whose recent estimates are in the top 10% according to StarMine's accuracy ratings is making a call different from the consensus.
Bulldog starts with the same logic but doesn't follow it as far. Instead of having SmartEstimates, for example, it offers Flash Consensus estimates, a number that gives extra weight to recent estimate changes but doesn't appear to weight the opinions of the top stock-pickers any more highly than the rest of the pack.
More disappointing: The way Bulldog presents information is distressingly backward-looking. Its front page is dominated by a "Stocks to Watch" feature that, when I looked at it, was mostly composed of stocks that had just announced quarterly results or other big news. Comparing this to StarMine's front page is like comparing a smoke alarm to a volunteer fire department. Bulldog would be more compelling if it got ahead of the news more often, as StarMine is trying to do.
The biggest flaw at Bulldog, however, is one it shares with StarMine: Too many analysts are missing, thanks to their unwillingness to cooperate with the sites' ratings. Neither site rates Merrill Lynch, Morgan Stanley Dean Witter (MWD), or Credit Suisse First Boston analysts, among others. Powerful firms like these have to be in the picture for these sites to be truly helpful.
Bulldog and StarMine are interesting and useful but hardly sufficient to guide investors through a volatile market. They're headline services that generate ideas, not fully formed analysis (which isn't infallible itself). Bulldog appears to recognize this: It provides prominent links to Multex on its site. The key to using these sites is to keep your expectations reasonable. Hmmm. That may be tough. Unreasonable expectations on the part of investors are what made the market bounce like a ping-pong ball for the past two years.By Timothy J. Mullaney, Tim_mullaney@ebiz.Businessweek.comReturn to top