A recent New York Times story ("The Obama Years: The Best of Times to Be a Stock Investor," by Jeff Sommer, August 19, 2016) made the case that the past few years were terrific for stock investors, as the market has nearly tripled during Obama's years in office. The author also makes the point that "Fed interest rate policy may be the single most important factor behind the stock market boom."
While the New York Times article examined returns on the Dow since 1900 without considering dividends, directionally the author's conclusions are correct. My research shows that since 1929, the S&P 500 with dividends has returned an average of 15.5 percent during the calendar years under Democratic presidents and only 6.9 percent during Republican presidents.
The returns during Democratic presidents have also been much more consistent than during Republican presidents over that period. Average annual returns under Republican presidents have ranged from minus 21 percent annually under Hoover to nearly 17 percent positive under Eisenhower. In fact, the only two Presidents that experienced negative stock returns during their times in office were Hoover and George W. Bush. On the other hand, returns during Democratic administrations have all fallen in a narrow band - from a low of up 9.3 percent during the Johnson administration to up over 18.2 percent during the Clinton years.
These statistics would seem to infer that one should never vote for a Republican for the highest office in the land, as the evidence is overwhelming that the market performs worse during Republican administrations. The problem with that logic is that in this simple analysis, stock market returns are attributed to only one variable - the political party of the President. While I think we all would agree that the President has some influence on the market, there are a host of other factors - fiscal policy, monetary policy, global economic conditions, and political gridlock, to name a few - that combine to help shape stock market returns.
Data and statistics have been misused over time to support a point an advocate wishes to make, and users of empirical studies need to be very careful regarding their conclusions. Humorist Mark Twain popularized the saying that there are "lies, damned lies, and statistics." One way to hold statistics accountable is to examine the degree of relationship between multiple factors through tools such as correlation coefficients. The correlation coefficient measures how closely related two sets of numbers are - for instance, the correlation between educational attainment level and income is quite high, which makes sense intuitively. People who have graduated from college, on average, have higher incomes than people who graduated from high school and pursued no additional education. And, people who graduated from high school have higher incomes than people who dropped out of high school. It is obvious that there is a cause and effect relationship between those sets of numbers.
However, a correlation between two sets of numbers does not always represent a meaningful relationship. This is when the statistics begin to "lie," when the correlation between two recurring events is often misused and misinterpreted. Simply put, two sets of data can be highly related, yet have zero cause and effect. Empirical researchers refer to this as spurious correlation, and there are many ridiculous examples. For instance, over an 11-year time period from 1999 through 2009, there was a high correlation between the number of people who annually drowned in a swimming pool and the number of films Nicolas Cage appeared in during that year. Specifically, people drowned at a higher rate in the years Mr. Cage appeared in a larger number of movies.
While I don't believe many (I would say no one, but according to a 1997 CNN/Time poll, sixty-four percent of respondents believed that aliens had abducted humans) would believe there is a correlation between movies Nicolas Cage appeared in and swimming pool drownings, people seemingly have an insatiable appetite to connect all kinds of events and, hopefully, better predict the future. It seems that every year in the week leading up to the Super Bowl, the Super Bowl stock market indicator is discussed. That is, the "fact" that, on average, the stock market performs better when an old NFL team wins the Super Bowl than when an old AFL team prevails. In 1989 the venerable Financial Analysts Journal actually published an article entitled "Did Joe Montana Save the Stock Market?"
I am certainly not making the case that the political party of the President has no impact on stock market returns, just that there are other variables to consider. Along with Scott Beyer of the University of Wisconsin Oshkosh, Luis Garcia-Feijoo of Florida Atlantic University, and Gerald Jensen of Creighton University, I jointly analyzed stock market returns relative to the political party of the president, the Federal Reserve's monetary policy, the year of the president's term and the state of political gridlock (that is, if the Presidency, the House, and the Senate were under one party's control). The research considered all of these variables together, and the findings were published in 2015 in the academic journal, Managerial Finance.
What we found was that considering those four variables all at once, the political party of the president was actually the least important. Our specific findings are that equity investors are wise to monitor Federal Reserve monetary policy. Stock returns are much higher when the Fed is pursuing an expansive monetary policy than a restrictive policy and that factor overrides the political considerations. It seems that all of the attention paid to Fed actions and inactions is warranted. Whether under Republican or Democratic presidents, the Fed has a significant influence on market returns, and Fed policy shifts are signals of coming inflationary pressures or other economic trends that impact markets.
With respect to political considerations, the conventional wisdom is that "gridlock is good for the markets" and this mantra is often repeated in the financial press. The theory is that significant fiscal policy actions, which tend to disrupt the markets, are more likely to occur during political harmony rather than gridlock. What we find is that political harmony is actually better for the stock market.
Fourth, equity investors generally have to wait until year three of a president's term to experience the highest returns. Even after controlling for Fed policy, political gridlock and the party of the President, there is an unexplained "third year effect." Thus, regardless of the outcome in the November election and any films Nicholas Cage may make in the meantime, it appears that equity investors may have to wait until 2019 to reap the rewards of the election season.