Presidential Elections Since 1896: Does the Market Express an Opinion?
November 11, 2012
by Doug Short
In previous commentaries this week I've been reporting on the historical patterns of pre- and post-election market behavior. Here is a wrap-up featuring the 30 presidential elections during the 116-year history of the Dow with some focus on outlier election weeks.
As an opening gambit, I'd like to share a bit of feedback on my latest World Markets Weekend Review. I commented that "The US presidential election took center stage of the world news media, and obviously the reelection of Obama was not what the market wanted. The S&P 500 dropped 2.37% the day after the election. For the week it fell 2.43% -- the worst weekly performance since late May."
I received an email suggesting that my comment "diverge[d] into new territory with a subtle, but obvious, dig at our presidential election. The writer closed the email with the speculation that "the reelection of President Obama was not what Doug Short wanted."
His email included the following observation:
|For someone who relishes statistics and multivariate analysis, attributing the primary cause of a short term market response to Obama's reelection seems like a stretch. If we go back to January 2009 when Obama was inaugurated, the S&P and Dow have shown significant gains (48% and 42%). I could characterize that longer term behavior as a positive response to fiscal and economic policies proposed by this administration. But it is probably more complex than that. Certainly the policies established by this administration have not inhibited the economy and markets to a partial, but successful recovery.|
As for policy, mine is to remain objective, all the more so in matters that touch on politics. The market's performance in the years following Obama's 2008 win was an unknown and thus irrelevant to the market's behavior in election week 2008. Similarly, the initial post-election reaction of the Dow and S&P 500 to last week's election is just that -- a reaction. It was an immediate response to an outcome, one the market didn't like. Will the market have second thoughts next week? Perhaps. But that's outside the focus of this commentary.
Presidential Elections and the Dow Sorted by Four-Day Performance
Below is an updated table illustrating the behavior of the Dow during the 30 election weeks stretching back to 1896.
Please note the "4-Days" column above (on which the table is sorted) and check out the 2008 election cited by my email correspondent. It's the second row above the averages at the bottom.
Someone who uses phrases like "multivariate analysis" will perhaps appreciate the significance of the fact that the 4.03% Dow decline in election week 2008 was nearly two standard deviations below all 29,000-plus rolling 4-day Dow performances since May 1896. The only election week with a worse Dow decline was Truman's historic upset of Dewey. There are few visitors to this website who aren't familiar with the photo of Truman holding up an early edition of the Chicago Tribune with the erroneous headline DEWEY DEFEATS TRUMAN (which, incidentally, is reported to be the most sought after newspaper of all time).
Note that I included the 4th and 5th standard deviation boundaries above the mean in the table. Why? The first row in the Dow table above is by far the most conspicuous outlier. The 10.56% four-day gain in FDR's first election week is smack in the middle between the 4th and 5th deviation boundaries.
Now check out the "Wednesday" column in the table above. Of all 30 post-election Wednesdays since the inception of the Dow, the one following Obama's 2008 win was the absolute worst: a 5.05% collapse. As we can see from the second right-hand table, that's four standard deviations below the mean Dow daily performance. Interestingly enough, the 4th standard deviation below the Dow daily mean starts at minus 4.59%, which, incidentally, is very close to what FDR got on post-election Wednesday in that amazing 1932 election week.
As for President Obama's first presidential election, the economy was indeed going through some traumatic times in the fall of 2008. The Lehmann Brothers bankruptcy occurred just seven weeks earlier. But I believe that the market's behavior immediately following the 2008 election was primarily focused on one thing: The fact that Democratic candidate Obama beat Republican McCain, thus terminating eight years of GOP control of the White House.
What about the 2012 election? Was the post-election market expressing an opinion about the outcome? Here is a 15-minute chart of the Dow for last week. What do you think?
For anyone interested in further exploration of the historical data and to help us see table numbers in the larger context, here is the chronological version of the chart above.
Presidential Elections and the Dow: Chronological Sequence
And here are some basic facts:
- Since its inception in 1896, Dow daily closes have been 52.2% positive, 47.2% negative and 0.6% unchanged.
- The average daily change, to three decimals, is 0.026%.
- The average of all four-day moving averages is 0.106%.
How do the presidential numbers in the table compare? Of the 120 days, 58.3% have been positive, 41.2% have been negative. There were no flat finishes. The average daily gain is 0.33%. Now please pay attention: That's 12.5 times larger than the mean Dow daily gain. It's no overstatement to say that election week tends toward extremes of daily volatility.