Another Reason Not to Trust the Analysts

Behind-the-scenes incentives breed bias among Wall Street analysts. And that probably helps explain why these fine folks, who have industry contacts on speed dial and reams of supply-chain data at their fingertips, are, almost without exception, consistently wrong.

You heard me right. The people paid to predict the future of company earnings routinely fail to get it right.

Rose-colored glasses and a tall glass of Kool-Aid
Almost 10 years ago, global consulting shop McKinsey & Co. set out to study the accuracy of analysts' forecasts. At the time, the results were far from flattering. And even though various investor-friendly controls have since been enacted -- Regulation FD and Sarbanes-Oxley, for instance -- analyst track records have not improved.

Consider the following McKinsey findings, according to a recently completed update to prior research:

• During the past 25 years, analysts have estimated 10% to 12% annual earnings growth (this figure excludes the heightened giddiness of 1998-2001). Actual earnings growth, by stark contrast, came in at 6%.
• Which means that, on average, "analysts' forecasts have been almost 100% too high" (emphasis mine).
• Furthermore, McKinsey discovered that analysts are slow to revise forecasts when the economy begins to sputter, and that slowness translates into increasingly inaccurate estimates at the exact moment investors are theoretically most in need of sober forecasts.

The next big miss?
With these unsettling facts in mind, I thought we might take a look at recent analyst earnings-per-share upgrades for a handful of popular stocks.

Data from Yahoo! Finance; estimates represent analyst average.
* Estimates are for the company's current fiscal 2011 year.

Now, I'm not suggesting that analysts are necessarily off the mark on any of the above outlooks. In fact, Caterpillar management recently affirmed its 2012 target of $8 to $10 profit per share, versus the average analyst estimate of $3.15 for the current year.

But because all of the above-listed companies operate in economically sensitive industries, the risk of an EPS miss in coming quarters is particularly large, especially given evidence that we remain mired in a deflationary environment.

Notably, McKinsey's research found that there are select times when analysts don't overshoot and instead nail a bulls-eye, and that's during economic recoveries. Essentially, strong economic rebounds have historically produced a profit environment in which actual earnings catch up with earlier Wall Street estimates. As such, analysts looked unusually savvy in 1988 and the 1994-97 and 2003-06 periods.

As for today, mortgage applications for new purchases are at a 13-year low, and key metrics such as real organic income growth and consumer confidence are still deep in the toilet. I'd keep a doubtful eye trained on the Wall Street soothsayers.