The stock market proved to be a poor predictor of this year’s U.S. presidential election, and it’s important to explore why.
In several recent columns, I reported on a simple model that correlated the incumbent political party’s chances of retaining the White House with the Dow Jones Industrial Average’s DJIA year-to-date return. Right before the election, that model was giving U.S. Vice President Kamala Harris, the Democratic candidate, a strong 70% chance of beating former U.S. President Donald Trump.
When a model fails, investors should use the occasion to explore what can be learned. Was this just one of those times when a model turns out to be wrong — something that inevitably happens sooner or later because no model works all the time? Or have there been more fundamental changes in the U.S. economy and the financial markets that make the model less useful?
With the election now over, my sense is that model’s breakdown has to do with the growing disconnect between Wall Street and the broader economy — what many call the Wall Street-Main Street disconnect. The reason that my simple model worked so well in decades past is that voters tend to vote their pocketbooks and the Dow used to be a decent barometer of overall economic health. This may no longer be true, or at least as true as it once was.
To show this, I analyzed quarterly changes in U.S. GDP with quarterly changes in the S&P 500’s SPX earnings per share (EPS) back to 1947. Specifically, I calculated the trailing 20-year correlation coefficient of these quarterly changes. (That coefficient ranges from a theoretical maximum of 100%, which would indicate that GDP and EPS move in perfect lockstep with each other, to a theoretical minimum of minus 100%, which would mean that the two move inversely to each other. A coefficient of zero would mean that the two have no detectable relationship with each other.)
The chart above plots what I found. Except for a brief blip in the wake of the 2008-09 global financial crisis, when both the GDP and EPS fell in unison, the correlation between the economy and the stock market has been steadily declining for several decades and is currently only marginally higher than zero.
