I really have trouble understanding the role of most financial analysts. Who (keeps) paying these folks? Why do journalists keep quoting them? In general, it seems like many financial analysts are nothing more than fortune tellers with horrible track records. Yet no one ever calls them on it.
This really comes to light when they try to assign percentages to future events. Using very precise numbers implies a level of scientific accuracy, so it sounds good. But these percentages are just silly.
Consider this example from a story in today’s Washington Post discussing the proposed XM/Sirius satellite radio merger.
The Post states that “in a Banc of America Securities report, analyst Jonathan Jacoby put the probability of [government] approval of the merger at about 35 percent, but noted that it was likely much lower.”
Uhh… Say what? I don’t know where to begin. Jacoby uses detailed analysis (I presume; after all, he wrote and published a “report”) to arrive at the probability of 35%. That must be a very scientific and insightful number, gleaned from years of careful analysis of the satellite radio and related industries.
But then, he says the number is likely much lower. This either means the 35% number is rubbish, or that Jacoby is trying to have it both ways. Which is it? If you arrive at 35% but believe the number is likely much lower, here’s a clue for analysts: lower your voodoo percent estimate to something lower than 35%. (Say, 27.82%. That number still sounds extremely scientific, doesn’t it?) If you’re going to publish a detailed statistic, you should at least stand behind it. Otherwise why say 35% in the first place?
This “having it both ways” syndrome afflicts too many analysts to count, and you can witness it in the regular “analyst earnings prediction parade” that precedes quarterly financial results for publicly-traded companies. Analysts go on the record predicting that a company’s earnings will be a certain amount (right down to cents per share), but then they whisper that their expectations are actually higher or lower, vastly extending the range of results that will permit them to say “I called that one!” when the results come in.
I gave an example of this in a past blog entry, when Richard Farmer of Merrill Lynch was interviewed by a reporter:
“Farmer said he expects Apple to exceed his own earnings forecast of 37 cents a share.”
As I noted then, if an analyst expects a company’s results to exceed his or her estimates, that analyst should raise his or her estimates. This isn’t rocket science. Or maybe they should simply get out of the business of providing estimates in the first place, and instead focus their analysis on higher-level aspects of an industry.
Unfortunately, companies and shareholders get hurt when the actual results don’t align with the pretend numbers and expectations analysts come up with. Don’t get me wrong; companies and shareholders can be helped, too (in the short term), because reality can clash with fantasy in positive or negative ways — what if a company’s results are dramatically better than an analyst’s predictions? The stock goes up! But this gives way too much power to analysts, who are trying to predict events with a level of certainty that is simply impossible to determine. (And the analysts never seem to get fired or removed from reporters’ Rolodexes when they get it wrong. Repeatedly.)