Stock Market Returns by Presidential Party
October 16, 2008 — Theodore Gray, Co-founder, Wolfram Research, Inc; Founder, Touch Press; Proprietor, periodictable.com
25 Comments 3991 213 6
The New York Times recently published an “Op-Chart” by Tommy McCall on its Opinion page showing what your returns would have been had you started with $10,000 in 1929 and invested it in the stock market, but only during the administrations of either Democratic or Republican presidents. His calculations showed that if you had invested only during Republican administrations you would now have $11,733 while if you had invested only during Democratic administrations you would now have $300,671. Twenty-five times as much!
That’s a pretty dramatic difference, but does it stand up to a closer look? Is it even mathematically plausible that you could have essentially no return on your investment at all over nearly 80 years, just by choosing to invest only during Republican administrations?
To answer these questions, I of course turned to Mathematica.
And the answer is that yes, it is mathematically plausible, using the assumptions made by McCall. My analysis using historical Dow Jones Industrial Average data available in Mathematica‘s FinancialData function roughly matches the figures in the Times, which used Standard & Poor’s data. (I used the Dow because it’s more convenient, not because I think it’s a better measure.)
But the fact that they are correct doesn’t mean the figures are even remotely meaningful. Here are some problems with the New York Times‘ Op-Chart:
First, it gives each president "credit" for stock market performance from the first day of his administration. That’s not reasonable: it surely takes at least a few months if not years before a president’s actions can start to affect the performance of the market.
Second, it completely ignores dividends: in earlier financial eras dividends were much more important than capital gains, and ignoring them distorts the picture.
Third, it ignores inflation: If the stock market was shooting up during a given administration but inflation was also high, the stock market gains may not count for much. Anyone can print money, that’s easy.
Fourth, there is the problem of the great stock market crash of 1929. While it occurred during a Republican administration, and while there were in fact a full eight years of Republican presidents preceding it, there were a lot of other factors involved as well. (McCall acknowledges this problem by including the observation that if you ignore the Great Depression a Republicans-only investment would now be worth $51,211 instead of $11,733.)
The question is, how representative, or how fragile, is the dramatic result shown in the Times? To help explore this I developed an interactive Demonstration that lets you play with the parameters and assumptions. Here’s what it looks like with the parameters set up to roughly mirror what the Times‘ Op-Chart showed. The differences are somewhat less dramatic (I’m using Dow Jones data while McCall used S&P data), but still very substantial, 10 to 1 in favor of Democrats.
October 16, 2008 — Theodore Gray, Co-founder, Wolfram Research, Inc; Founder, Touch Press; Proprietor, periodictable.com
25 Comments 3991 213 6
The New York Times recently published an “Op-Chart” by Tommy McCall on its Opinion page showing what your returns would have been had you started with $10,000 in 1929 and invested it in the stock market, but only during the administrations of either Democratic or Republican presidents. His calculations showed that if you had invested only during Republican administrations you would now have $11,733 while if you had invested only during Democratic administrations you would now have $300,671. Twenty-five times as much!
That’s a pretty dramatic difference, but does it stand up to a closer look? Is it even mathematically plausible that you could have essentially no return on your investment at all over nearly 80 years, just by choosing to invest only during Republican administrations?
To answer these questions, I of course turned to Mathematica.
And the answer is that yes, it is mathematically plausible, using the assumptions made by McCall. My analysis using historical Dow Jones Industrial Average data available in Mathematica‘s FinancialData function roughly matches the figures in the Times, which used Standard & Poor’s data. (I used the Dow because it’s more convenient, not because I think it’s a better measure.)
But the fact that they are correct doesn’t mean the figures are even remotely meaningful. Here are some problems with the New York Times‘ Op-Chart:
First, it gives each president "credit" for stock market performance from the first day of his administration. That’s not reasonable: it surely takes at least a few months if not years before a president’s actions can start to affect the performance of the market.
Second, it completely ignores dividends: in earlier financial eras dividends were much more important than capital gains, and ignoring them distorts the picture.
Third, it ignores inflation: If the stock market was shooting up during a given administration but inflation was also high, the stock market gains may not count for much. Anyone can print money, that’s easy.
Fourth, there is the problem of the great stock market crash of 1929. While it occurred during a Republican administration, and while there were in fact a full eight years of Republican presidents preceding it, there were a lot of other factors involved as well. (McCall acknowledges this problem by including the observation that if you ignore the Great Depression a Republicans-only investment would now be worth $51,211 instead of $11,733.)
The question is, how representative, or how fragile, is the dramatic result shown in the Times? To help explore this I developed an interactive Demonstration that lets you play with the parameters and assumptions. Here’s what it looks like with the parameters set up to roughly mirror what the Times‘ Op-Chart showed. The differences are somewhat less dramatic (I’m using Dow Jones data while McCall used S&P data), but still very substantial, 10 to 1 in favor of Democrats.
