Third Presidential Years: Coincidence or Causation?

Third Presidential Years: Coincidence or Causation?

December 1, 2010

Beginning in the 1970’s, and after seeing some work by my good friend Jon Lovelace of Capital Research and Management, I have tracked U.S. stock total returns (S&P 500) by Presidential Years.

Now I know that many others have done similar work, but I think my studies have been unique in an interesting way. I have used Jon Lovelace’s idea of looking at years as politicians see them, that is, twelve month periods bookended by election days. For convenience I have used November 1st to October 31st.

Each November I have presented a forecast for the following twelve months based on the historical data for that presidential year. The forecasts have been in a range, using the standard deviation, plus and minus, around the historical mean return.  The range forecasts have scored pretty well, but often only because they were quite wide, and therefore of limited practical use.

In November 2009, for example, the forecast for the year just ended on October 31, 2010 was for a return of 7.2%, plus or minus a whopping 21%, or a range of 28.2% to negative 13.8%, hardly a courageous prediction! The actual return was 16.5%.

However, over time one presidential year, the Third has been very different:

  • It has been, by far, the highest return year, with an average return of over twice any other year. The median return has also been more than twice Years One and Two and 90% higher than Year Four.
  • There has never been a negative Third Year.
  • Of the fifteen Third Years beginning in 1951, only one (1987, Reagan’s Third Year) had a return below 14.6%.
  • The variability of returns in Third Years has been the narrowest by considerable measure. The standard deviation has been only 8.8%, a bit more than half the level of other years.
  • After considering this data it may not be surprising that there have been no business cycle peaks in Third Years, while there have been five cyclical troughs.
  • The Third Years’ lows in stock prices have only once (1975, Ford’s presidency) been more than 10% below the previous October’s close, and in only five years have the lows been more than 5% below the October close.

Based solely on this history, the forecast for the current year ending next October 31 is for a total return of 23.4%, plus or minus 8.8%, or a range of 14.6% to 32.2%. The low for the year, if historical average holds, will be only 2.6% below the October close.

I hope you find all this entertaining, but I’m sure you are asking whether there is causation here or simply coincidence.  Well, I can at least advance a hypothesis, one that I doubt can be either proved or disproved.

The hypothesis is that the party in power (in the presidency) should be expected to try  to take the bad economic knocks in Years One and Two, then gear up the economy in Year Three for the election push in Year Four. Some confirmation of this shows up in the timing of business cycle peaks and troughs. Eight of the eleven cyclical peaks experienced since 1948 have occurred in Years One and Two. To the extent that the federal government can significantly affect the economy, the politicians will try to take as much stimulative action as possible in Year Three, knowing that such actions have an effect only with a lag.

I know, I know, I hear what you say: what can be done this year with a split Congress, a bunch of new legislators arriving with ideological obsessions and no experience, and a huge continuing deficit staring at us??

True, there certainly are obvious ways that this year is different from the past. The economy is more global in scope than in earlier years, we are still exiting the worst recession and financial meltdown since the Great Depression, there is no room to either lower interest rates or expand federal spending, state and local governments are on the financial ropes, and consumers are still paying back debt.  What can government do? Not much, in my opinion, except to get out of the way and let the private economy continue to expand slowly but surely.  Also, a credible plan to reduce the federal deficit would help confidence. Maybe that combination will be enough to give us another good Third Presidential Year.

Am I willing to make the prediction for a good market year based on this data?  Sure.  It may not be scholarly or the result of deep economic/financial analysis, and I may not bet the ranch on it,  but do you know a better way to forecast the stock market?

Here is some of the data.  If you want more I may be able to send it by email….it’s voluminous!


Years Ending Oct. 31 President S&P500 Total Return Business Cycle Peaks Business Cycle Troughs Low as % of Oct
1951 Truman 26.0     97.3
1955 Eisenhower 39.4 N   100.3
1959 Eisenhower 15.7     99.4
1963 Kennedy 35.3 O   101.1
1967 Johnson 21.0     99.3
1971 Nixon 16.9 N Nov ’70 99.4
1975 Ford 26.0   Mar ’75 88.0
1979 Carter 15.3 E 99.3
1983 Reagan 27.9   Nov ’82 99.4
1987 Reagan 6.6     92.2
1991 Bush 33.5   April ’91 100.7
1995 Clinton 26.4     94.3
1999 Clinton 25.7     101.1
2003 Bush 20.8   Nov ’02 90.4
2007 Bush 14.7     99.0

Here is a summary table covering all years:

Presidential Years ONE TWO THREE FOUR
Average Return 7.9 7.8 23.4 10.8
Median Return 7.3 11.3 25.7 13.5
Std. Deviation 15.2 20.4 8.8 15.6
Low as % of Oct 90.0 88.1 97.4 92.9
Negative Years 4 of 16 6 of 16 None 2 of 15
Cyclical Peaks 5 3 None 3
Cyclical Troughs 3 2 5 1




One Response to Third Presidential Years: Coincidence or Causation?

  1. Crawford Long says:

    I like your blog and look forward to each new entry. I agree with your data and conclusions but wonder why data further back, especially the 1930’s, was not included. You can’t have too many data points.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s