One of the most thankless jobs on Wall Street has to be the securities analyst. You have to dig up fresh information on a company that may not want it to be known, and often the company is a client of your employer. You have to sift through a lot of metrics, but too often at the end of the day people just want to know about one — the earnings per share.

Then you have to deal with blog posts like this one from Investor Insight (via Big Picture). It’s a piece written by James Montier, head of equity research at Societe Generale based in London, who offers some compelling proof that, when it comes to forecasting earnings, analysts are very much behind the curve. In fact, they make an excellent rear view mirror.

Montier sums up his peers’ predictive skills pretty harshly,

The chart makes is transparently obvious that analysts lag reality. They only change their minds when there is irrefutable proof they were wrong, and then only change their minds very slowly.

But he faults the practice of calling up CFOs and asking them for their thoughts. The real blame belongs at the office doors of those CFOs. A Duke University survey showing CFOs as a lot consistently more pessimistic about the economy than their own companies. They are more often right about the economy.

Analysts often get a bad rap for being bullish about companies that are their clients. That has happened a lot less since the early 2000s. The real problem is more subtle: They are too reluctant to challenge the (natural) optimism of corporate execs. It’s a classic exception to the rule that you should act as if the customer is always right.