RSS feedSuper Tuesday is coming up in a week and this year’s presidential elections are heating up. With all sorts of worries about the markets and the greater economy, can history enlighten us on what to expect? Maybe it can even tell us who to vote for!
Keynesian economics was first postulated in 1936 and became widely accepted by the 50s and 60s. Since then, the government either through the Federal Reserve and monetary policy or the President/Legislature and fiscal policy have been very active in trying to influence the direction of the economy.
Furthermore, we all know how important it is for an outgoing President to attempt to leave office while the economy is chugging along. It gives him the best shot at winning another term or being able to leave a favorable legacy and pass control to his party.
So, does this usually work out? Is the government really so able to create cycles in our economy centered around election dates? Well, an article written by a Pepperdine Professor, Marshall D. Nickles, claims that there is some truth to this idea. He analyzed the returns of the S&P 500 on a monthly basis from 1942 until 2004 and found that there does seem to be a roughly 4 year cycle of peaks and troughs in the S&P returns. Of the 16 four-year terms there have been from 1942 to 2004, it was found that 15 of the 16 market troughs fell in the first or second year of a presidency and one (the 1985-1988 term) fell at the end of the third year of a presidency.
If one were to define bear markets as a draw down of 15% or more in the S&P 500 in a one year to three year period, it was found that most of these bear markets occurred within the first two years of a new presidency. On average, the market has reached its worst point 1.87 years into any new presidency. It was also found that investing in the market during an election year more often than not proved to be quite lucrative.
So, here’s the question, can we really depend on this data? Truth be told, the analysis done in this paper is not rigorously statistical. There are some interesting patterns, but it’s hard to say whether or not 15 trials is sufficient enough to declare victory on the subject of investing based on election cycles. After all, look at the economy this year? Could this be another “anomaly” like that of 1987 where the market crashed in one of the latter two years of a presidency?
In fact, a little more rigorous paper on the subject examining market returns using the Fama/French factors (a modification of CAPM which takes into account the risk/return anomalies found in low P/E and low P/B stocks) found that there is a statistically insignificant change in expected returns of the market over the 12-month period before and after an election. Though, they did find some correlation to presidential elections in the quarter prior to and after an election. In fact, they went as far as to attempt to use market performance to predict potential winners of presidential elections (albeit only in years where an incumbent is up for re-election… i.e. not this year). Check out the paper here if you’d like to try a little heavier stuff.
