Last Updated Nov 3, 2010 12:32 AM EDT
Mid-term elections predict a bull stock market for the next three months, according to an article on Slate.com. That would be one good thing to come out of a year of enduring negative, mud-slinging TV slur campaigns. But should you buy stocks based on this research?
Markets in Mid-Term elections
This article gets right to the point on the love affair between mid-term elections and the market.
One thing is certain: Markets love midterm elections. Brian Gendreau, a market strategist for Financial Network, found that the Dow has risen after 19 of the 22 most recent midterms. From 1922 to 2006, the Dow jumped 8.5 percent in the 90 trading days following the midterms, versus just 3.6 percent in non-midterm-election years.
This implies that there is an 86 percent chance (19 of 22 years) that the market will rally and that the expected rally will be 8.5 percent in the next three months.
Explanations for this rally
Gendreau offers two theories for this rally. First, mid-term elections usually mean that the incumbent party loses some of its power, which leads to more legislative gridlock. Markets tend to like gridlock. The second explanation is that markets hate uncertainty, and elections tend to bring some certainty.
I'll actually buy both explanations to a degree. Personally, I'm finding it nearly impossible to do tax-planning for clients, as we still don't know income tax rates, capital gains rates, or dividend rates for 2011, which is less than 60 days away.
Why I'm selling stocks
As compelling as the data and the explanations are, I will not be hopping on this rally bandwagon. You'll find me on the other, less crowded, bandwagon that will be selling stocks.
I see this prediction as a textbook example of data mining. Trying to find the secret to what makes the market tick is an all too common quest. As it happens, you can compare the stock market with 1,000 different random variable sets like traffic, the super bowl, or butter production in Bangladesh, and you'll get, on average, one series that give a 99.9 percent probability of correlation. Unfortunately, this correlation will offer no predictive knowledge whatsoever. Behavioral finance indicates that humans hate randomness, and seek out patterns to predict the future.
Even if my explanation above is wrong, think through the logical conclusion. Since this pattern is pretty well known, investors would have bought stocks in the last few weeks, getting ready for this 8.5 percent bull market. That buying would have already caused the market to move upward, which could explain why markets are up about four percent for the last month.
But I'm not selling stocks because I think this prediction is going to cause markets to decline over the next 90 days. I'm selling because I'm a believer in rebalancing, and stocks are up about 80 percent from the bottom of the bear on March 9, 2009. I'm taking the far less traveled path than that of advisors who time the market poorly.
Ignore these silly predictions, no matter how compelling the logic. It's difficult to accept that we don't know the future. It's more difficult, however, to make up your losses that timing the market is almost certainly going to cause.