There is an interesting theory out there saying that Harvard grads (and in particular HBS grads) collectively serve as a good leading indicator of where the US equity market is going. In a nutshell, the theory goes that if lots of HBS grads are headed to Wall St., then the markets are overvalued and headed south. Conversely, if fewer HBS grads are going into the world of high finance, then the markets have goodness in store for all of us. Basically, HBS students don’t like to show up to the party until everyone else has already arrived, and more often than not, they arrive just as everyone else has decided to leave and go home.

As for the 2007 HBS graduating class, it appears that 40% are headed for “market sensitive” careers, up from 37% in 2006 and 30% in 2005. So if you believe the theory, then you better be selling! Or rather you better have been selling already because as you see in the graph below, all of the major indices are down in the last 12 months (i.e., basically since the time the 2007 class graduated).

But enough with the intro. On to my point. Fortune magazine just came out with its annual list of the top 100 most desirable MBA employers. And while it doesn’t allow you to break out the list to just HBS grads, let’s assume for a moment that all MBA grads – not just the poor folks at HBS – are bad at picking their career paths. If the theory works for all MBA students, then perhaps this list enables you to infer not only which way the stock market as a whole is headed, but also where specific companies are headed?

Who’s at the top of the list for the class of 2008? Google. Does this mean GOOG is ready for a fall? Well, they were #2 in 2006 and the stock rocketed upwards during the year that followed that class’s graduation. By 2007 they had got the #1 spot, but the stock is still slightly up since the 2007 MBA grads joined the workforce.

Could this finally be the year the theory works and GOOG drops? Should we sell? Should we sell AAPL (#4 on the 2008 list)?

My own hypothesis: the theory might work fairly well when guessing the direction of the market as a whole, but it will break down and have less success when applied to individual companies. Get a big enough sample size and you will be okay. Focus in on a single data point (or company) and there are too many external factors that could change the outcome. In statistical terms, pick a single point in a normal distribution and you may be well off the mean. But take the full sample and you’ll zero right in on the mean.

Or perhaps the MBA grads are not as quite as useless as every engineer in the valley suspects!