Election Pop, by Dr. Kim Zussman
One theory is that incumbent parties grease markets before elections. In that September and (thus far) October has been good for stocks, how do these two months compare for years with elections vs. not?
Checking SP500 index monthly closes (with dividends) since 1980, two-month non-overlapping returns were compared with two-month Sep-Oct returns, counting back every two years from 2006:
Two-sample T for election vs all 2 months
N Mean StDev SE Mean
election 14 0.025 0.0637 0.017
all 2mo 156 0.017 0.0646 0.0052
T-Test of difference = 0 (vs not =): T-Value = 0.44 P-Value = 0.667 DF = 15
Though Sep-Oct returns for election years are better than average two-month returns, the difference is not significant. But perhaps the levers are longer term. The same test on July-Oct returns of election years vs. all four-month returns shows election years to be lower:
Two-sample T for elect vs all 4 months
N Mean StDev SE Mean
elect 13 0.0160 0.0926 0.026
all 4m 77 0.0332 0.0913 0.010
T-Test of difference = 0 (vs not =): T-Value = -0.62 P-Value = 0.544 DF = 16
For two- and four-month moves prior to elections, it seems the greater powers which are working are not political parties.