With the presidential election just about a week away, and media narratives shifting with each poll that is released, we thought it may be good time to look back at the previous sixteen presidential elections and see how equity markets behaved around them. Equity markets are clearly jittery, though it is hard to reconcile correlation and causation amid news from the presidential race and perhaps more importantly, a Federal Reserve that seems poised to hike interest rates once again before the year is out.
In this post we look at how equity markets behaved around presidential elections.
Note that we have data only for the past sixteen presidential elections, which is a really small sample from which to make many conclusions. More so because each presidential year has its own idiosyncratic characteristics, starting with the candidates themselves, the economic situation or other forces (war, etc). Nevertheless, we do believe it useful to do this exercise by treating it more like an event study. As Noah Smith points out , event studies don’t predict markets and shouldn’t be treated as if they do. The goal here is not to find a market-beating forecast, but to study the behavior of markets around presidential elections, and whether or not the election had any discernible impact on equity markets.
We start by looking at equity markets around election day and whether returns in the month after are higher than the month before. One would think that the month prior to election day would be a period of high uncertainty, which should reduce after election day (unless you get a contested result like in 2000).
The table below shows returns for the S&P 500 price index (excluding dividends) in the month prior to and post election day. The table lists each of the past sixteen presidential elections, along with the President elect and whether or not they were an incumbent.
Ten out of the sixteen years saw post-election day returns higher than in the month prior (positive difference in the last column). However, the average return in the month after election day is actually 0.3 percentage points lower than in the month before. Since this is a small sample and averages can be skewed by outliers (like 2008) it is useful to look at the median as well. In this case, the median difference between returns in the month after election day and the month prior to it is only +0.1%, which is hardly significant. The difference is insignificant even if you start isolating the years by whether or not an incumbent won the election.
Is equity market volatility higher in the month prior to election day than in the month after?
The next table shows annualized volatility (of daily returns) in the month prior to and post election day.
On average, volatility has dropped 1.1 percentage points in the month after election day from the prior month but, as you can see from the table, this number is skewed significantly by some outlier years (again, 2008). The median difference between post election day volatility and the month before is only -0.1 percentage points. Volatility fell in the month after election day (compared to the month earlier) in eight out of the sixteen presidential election year.
In other words, based on what happened in previous presidential election years, it’s anyone’s guess as to whether equity market returns after the election will be higher than the period before, let alone volatility. The fourth quarter of the year is typically the strongest period for equity markets and even in presidential election years, investors can miss out by sitting out this period based on election uncertainty.
Aren’t equity markets positive in presidential election years?
Presidential election years have in fact been good ones for an investor in the equity market, with just three exceptions over the past sixteen cycles. The table below shows returns for the S&P 500 index (excluding dividends) over the entire presidential election year. Also shown are year to date (YTD) returns until election day and in the last column, returns from election day through the end of the year.
The average annual return in a presidential election year was 6.6% but this is skewed significantly by 2008, when the S&P 500 fell 38.5%. The median annual return across presidential years was considerably higher, at 10.4%. The average return over the approximately ten months prior to election day is 4.9% whereas the average return in just the two month period after election day is 1.3%.
The S&P 500 was positive between election day and year-end in ten out of the sixteen years, once again indicating that it is not easy to predict whether or not equity markets will run up during this period.
What about the exception years, namely 1960, 2000 and 2008, when the annual return was negative?
Well, it turns out that two out of these three years (1960 and 2008) coincided with a recession and rising unemployment, as the next table illustrates.
All but two of the sixteen years, namely 1960 and 2008, saw recessions across most of the year, and coincided with negative returns for the S&P 500. Unemployment also trended higher in these two years whereas it was steady or falling in all other years. The year 2000 was an exception, with no recession and a fairly steady unemployment rate across the entire year, but the S&P 500 had a negative year thanks to the tech bubble bursting. Note that the early 2000s recession was officially declared to have started only in March 2001.
Overall, the point is that a better approach to deciding whether or not to invest in the equity market is to focus on the fundamental situation of the economy. For example, determining whether or not the economy in a recession or close to one, as opposed to the vicissitudes of an election year. As of the end of October 2016, the unemployment rate has been fairly steady and it is very unlikely that the U.S. economy is in a recession or close to one – the latest third quarter GDP growth came in at 2.9%, the highest rate in two years.
This is also a good time to remember that the U.S. economy is a very complex engine and a President by themselves has very little control over it. If at all they have an impact, it is likely to come only as part of a concerted effort with Congress, and typically occurs in the middle of a crisis/recession in the form of a stimulus package.