More Senseless Data Mining

One of the themes of my book, Invest Like a Fox…Not Like a Hedgehog, is that investors waste a lot of time and energy trying to find the silver bullet of investing. Specifically, they comb through data looking for a key that will tell them when to buy and sell stocks or take other actions. Here’s another example from a financial writer I usually admire. Joe Wiesenthal believes that the stock market moves up and down with the weekly filings of new unemployment claims, at least for the last few years. The first part of the post actually does a good job of explaining why the charts don’t necessarily prove anything for a variety of reasons, including that there are a number of other moving parts. But that doesn’t stop him from presenting a final chart that he believes clinches his case.

Count me as skeptical, for all the reasons Wiesenthal mentions earlier in the piece. But if you buy his argument, keep this point in mind. When you and Wiesenthal’s other readers start following this signal, it will stop working. You’re better off using a combination of the long-term buy-and-hold portfolio we recommend in Retirement Watch and the tactical margin-of-safety portfolios  we present.

In this year, the connection in claims and the stock market isn’t quite as pretty, but a lot of that has to do with the fact that stocks crashed in the aftermath of the debt downgrade and the debt ceiling fight. The economy never really deteriorated to match the event-driven freakout in stock prices. But even still, you can see that in the final three months of the year, there was a nice gain in stock prices (blue line), and nice improvement in initial claims (red).

And yet here again we have a problem! That’s because on September 21, 2011, the Fed announced Operation Twist (which is essentially QE3) so once again, it becomes hard to figure out, timing-wise, whether the rebound in the stock market is related to Fed action or the improvement in jobless claims.