The Financial Times is reporting about a possible pre-election bump in U.S. stock prices if historical election trends remain in place. For those of you who are interested in such data (snooped or not), it appears that since 1926 U.S. stocks have beaten the world market in 75% of the presidential election years by an average spread of 13.4% from June to October. Analyzing that data more, it appears that when global stocks are rising, U.S. stocks lead 77% of those periods. When stock markets decline, U.S. stocks led in 71% of those periods. It also appears that it does not matter whether you begin the data analysis in July, August, or September, at least as far as having returns outperform.

In the past I did some research on how the stock market performed when there was a change in power in Congress (from one party to another). This research is similar to the analysis of returns for the stock market during different presidential cycles - data that always gets thrown around about this point in the election cycle. For our research we speculated that Congress, with its ability to create tax law, had more of an impact on stock returns, even with a Presidential veto threat. For those interested, our research found that divided government produced the highest returns (probably since Congress and the President were unable to get anything done, but that is just conjecture). Anyway, I point this out since it is worthwhile to remember that for any of these studies, the data sets that are used are often very small, with most results outside the acceptable levels of statistical error. So while they are fun, and sometimes worth paying attention to, always remember what type of data set you are dealing with so you can properly adjust your own confidence in the data and risk levels for your trading.