#MARKETS & INVESTING
Between 1927 and 2015, the excess return on US equities averaged 10.7% a year under Democratic presidents, against −0.2% under Republican ones.
Published on August 27, 2026
Pedro Santa-Clara and Ross Valkanov documented the gap in the Journal of Finance in 2003: over 1927-1998, the excess return on US equities is 9 percentage points a year higher under Democratic presidents on a value-weighted basis, 16 points on an equal-weighted one. Finding no explanation in risk or the business cycle, they named it the 'presidential puzzle'. Lubos Pástor and Pietro Veronesi extended the measurement to 1927-2015: 10.7% a year under Democrats, −0.2% under Republicans. The gap does not fade out of sample — 17.4 points a year over 1999-2015. Their explanation is not political: Democrats tend to be elected when risk aversion, and therefore the required risk premium, is high.
Does the US president's party change stock market returns?
Over 1927-2015, the average excess return is +10.7% a year under Democrats and −0.2% under Republicans. The authors attribute the gap to when presidents are elected, not to the policies they run.
In other words: it is not what presidents do, it is when they are elected.
Four caveats, because they matter
- The sample is counted in presidential terms over roughly a century, not in thousands of independent observations.
- The figure measures an excess return, above the risk-free rate, not the market's gross return.
- The result is American: none of the three studies replicates it on other markets.
- Sy and Al Zaman (2020) rule out spurious regression and data mining, but none of this work predicts the return of a future term.
Why it still matters: a century-old regularity, robust out of sample, can still be useless as a decision rule.
For you
- Is an 11-point annual gap, measured across a few dozen presidential terms, enough to make a rule?
- Did you know US equities returned −0.2% a year in excess terms under Republican presidents between 1927 and 2015?
Source: Santa-Clara and Valkanov, 'The Presidential Puzzle', The Journal of Finance, 2003; Pástor and Veronesi, 'Political Cycles and Stock Returns', NBER Working Paper no. 23184, 2019; Sy and Al Zaman, Journal of Empirical Finance, 2020.
View the source →