What’s The Most Accurate Presidential Election Predictor?

The Daily Beast reports on some of the offbeat U.S. presidential election outcome predictors, including 7-Eleven Coffee
cups, Halloween masks, the Family Circle Recipe Contest, the Washington Redskins Home-Field Record, and others. But Elizabeth Wine writes in September’s On Wall Street about the most historically accurate predictor of the outcome of U.S. presidential elections. The pattern, according to market historians,. is as follows:

  • If the S&P 500 rises in he three months before the election, the President will most likely win a second term, So far this year, it’s up 10.6%.

Sam Stovall, Chief Equity Strategist at S&P Capital IQ, writes in a research note,

Investors will not have suddenly decided to embrace the current administration and their seemingly anti-Wall Street policies, but will have become sufficiently encouraged by the improving economic conditions to stay the course.

The Evidence

The evidence, according to Stovall, is that, in 28 presidential elections since 1900:

  • The incumbent was reelected 80% of the time when the S&P 500 increased the three full months before the election.
  • If the market fell, the incumbent lost 88% of the time.
  • On average, the S&P predictor works 82% of the time.

Stovall insists that, of the many indicators from the Superbowl Effect to the Hemline Theory, the S&P performance has greater accuracy, and is more intuitive:

That’s why investors look to stock price movements; to see whether they point to a strengthening or a weakening in the economic data.

Forget about ideology, including the old saw that Republicans are good for stock portfolios. Jeff Hirsch, editor of the Stock Trader’s Almanac and Chief Market Strategist at Magnet Æ Fund, agrees, asserting:

It troubles me that everyone still believes that just because Republicans are said to be good for the market that they are. It’s not about the party; it’s about the incumbency issue.

Why The President’s Reelection Chances Are Good

Stovall also compared stock prices to the poll numbers. His findings: when the market has gone up, so have President Obama’s poll numbers. When the markets went down, so did Republican challenger Mitt Romney’s poll numbers. Do good poll numbers drive stock rallies, or  do rallies create better feeling toward the president?  He can’t explain causality:

This is a chicken or egg thing. All I know is Obama’s number went up, as did the market’s numbers. I’m not willing to say which drove which, but I am willing to say they move pretty much in lockstep.

Are There Exceptions?

In only four elections since 1900, the incumbent lost despite S&P rallies. Stovall believes that in three of those four times -1912, 1968 and 1980- this happened because a third party candidate siphoned off votes from the incumbent:

  • Theodore Roosevelt in 1912.
  • George Wallace in 1968.
  • John Anderson in 1980.

So, only once, in 1932, when Franklin Delano Roosevelt beat President Herbert Hoover, was the predictor completely wrong.

Relief Always Follows The Election

History also shows that, regardless of the winner, investors have been relieved or optimistic enough to cause the S&P 500 up an average of 0.6% in November, and 1.7% in December. After an August-October gain, the S&P typically rose an average of 1.4% in November and 1.8% in December.

When markets were down ahead of the election, the index rebounded an average of 1.3% in November, and broke even in December.