Theories linking stock market returns to the election cycle have not held water. (Keystone).

When this theory became popularized in the mid-1900s, it appeared to be true based on recent elections, but like other market superstitions (the January effect, Sell in May, etc.), this perceived correlation and causality turned out to be coincidence, with the data skewed by a few outliers. In many years since then, the data have gone against the theorized relationship; for example, the S&P 500 fell -37% in 2008—a presidential election year—and rose significantly in the first year of presidential terms, such as +26.5% (2009), +32.4% (2013), and +21.8% (2017).

Finding signal, even in random noise

Why did this theory arise, and why does it continue to persist despite contrary evidence? For one thing, it feels good to look for patterns, and we all secretly hope we can find a way to time markets. In addition, this is an example of a logical fallacy known as post hoc, ergo propter hoc (after this, therefore because of this), where we presume correlation is causation.

Our brains are pattern-recognition machines—it's our super power—and we love a good story, so we're hard-wired to connect dots even when we are fed random noise. This is especially easy to do when we have a small sample size, as in the case of US political cycles. There are a hundred ways to look at data, and some of them will appear genuinely related to returns or market volatility, but this is an illusion. This is something we call "data mining," and if you're interested in seeing this in action elsewhere (divorce rates and margarine consumption, for example), Google "spurious correlation examples."


2020 is "Year 4" of President Trump's first term, but the presidential cycle doesn't tell us anything about whether that means it will be an above- or below-average year for markets. As investors, it can be difficult —but important—to resist the temptation to turn political hopes, fears, and predictions into investment changes. Instead, we should evaluate the prospect of specific policy changes, and assess them as risks and opportunities for asset classes and sectors of the equity market. Please be sure to stay tuned to our ElectionWatch series and discuss things with your financial advisor (and tax professional). Today, it's too early to reach any investment conclusions, but that will change as we get closer to a Democratic nominee and gain better understanding of the shape of the likely policy outcomes that could be implemented when the next presidential term begins in 2021.

Read the original post Are election years good or bad for markets? 14 November 2019

Main contributor: Justin Waring

This content is produced by the Chief Investment Office.