On October 14th 2008, three weeks before the presidential election, The New York Times published the eye-catching chart below: (follow link for full size version)
Full sized NYT Chart
They say that if you torture data enough you can eventually get a confession, and this chart proves it - Khalid Sheikh Mohammad got better treatment! And yet there is not a single number I dispute on the chart. No, the problem is not one of accuracy, but rather of timing and the very premises on which the chart is based. The fact is, by appropriately correcting just the timing problems I was able to reverse the conclusion of The Times, but unlike the Grey Lady I would not suggest that you invest a single dime based on these more relevant results.
First to the premises: In order for there to be some relevance to this chart, one would have to assume a minimum of five dubious premises:
1. Party affiliation is the determining factor in a President’s economic policies.
2. Presidential economic policies are isolated from congressional policies.
3. Political parties have had consistent economic policies throughout history.
4. Markets live in the moment and are not in the habit of discounting the future.
5. Markets were either irrelevant or not measured before 1929.
Not a single one of these premises is strong enough to hang your hat on, let alone bet $10,000!
To wit: Were JFK’s economic policies anything like Barack Obama’s? Were Nixon’s anything like Reagan’s? How about the idea that Presidential policies are isolated from congressional policies? Was Bill Clinton the same President before ’94 as he was after? And how about parties being consistent over time. Is that valid? Hoover, a Republican, was a protectionist and Clinton, a Democrat, was a free-trader; what about Reagan vs. Obama on the same issue?
As for the 4th and 5th premises, they are just laughable but correctable, and to that end I reworked the chart. To account for the fact that markets are forward-looking I used equity values from 6 months prior to Presidential elections when markets are poised to react to the political calendar and when Presidents are referred to as “lame-ducks”. The Times chart uses values from the exact date of inauguration which assumes that the market has made no adjustment for an incoming President's expected policies. Anyone who follows the market knows that it is concerned with little else but the future! Finally, to account for the fact that there is nothing magic about only going back to 1929, I went back to 1900 which coincides with the dawn of statistical movements in the Standard and Poors Index for which there is data back to 1871.
Here’s what I found:
S&P Data Source: Yale Professor, Robert J Shiller
1. From 1900 to today, Republican Presidents yielded average annual returns of 4.66% and Democrats 4.05%, and that includes Hoover!
2. Even since 1929, excluding Hoover, Republicans beat Democrats by 7.13% to 5.03%!
3. The only way to get Democrats to beat Republicans is to start with Hoover (as the Times did) and then it’s Democrats, 5.03%, to 3.51% for the Republicans. Whew!
4. Many like to use the post-war period as an investment benchmark: Since WWII Republicans have whomped Democrats by 7.13% to 3.66%!
So, what does this prove? Absolutely nothing - except that you really shouldn’t pay too much attention to The New York Times for investment advice! For one, the correlation between parties and markets is too weak. Second, there is too much “noise” for a reliable correlation. And third, there is still the problem of the first three premises!
In short, vote for the guy or gal you prefer, and base your investments on other factors.