Sell In May, from Kim Zussman
Here is a check on the evolution of "Sell in May". SP500 Nov1-Apr30 mean returns were tested against zero, by decade ("00" = 2000's, "90"'s, etc):
Test of mu = 0 vs not = 0
Variable N Mean StDev SE Mean 95% CI T P
N-A 00 10 0.0105 0.0912 0.0288 (-0.0547, 0.0757) 0.36 0.724
N-A 90 10 0.1154 0.0871 0.0275 ( 0.0531, 0.1777) 4.19 0.002
N-A 80 10 0.0830 0.1055 0.0333 ( 0.0075, 0.1585) 2.49 0.034
N-A 70 10 0.0648 0.1233 0.0390 (-0.0234, 0.1531) 1.66 0.131
N-A 60 10 0.0576 0.1246 0.0394 (-0.0314, 0.1468) 1.46 0.177
N-A 50 10 0.0790 0.0841 0.0266 ( 0.0188, 0.1392) 2.97 0.016
All positive, 3/6 decades significantly greater than zero. Here is the same test for May1-Oct31:
Test of mu = 0 vs not = 0
Variable N Mean StDev SE Mean 95% CI T P
M-O 00 10 -0.0152 0.1514 0.0478 (-0.1235, 0.0930) -0.32 0.758
M-O 90 10 0.0442 0.0657 0.0207 (-0.0027, 0.0912) 2.13 0.062
M-O 80 10 0.0429 0.0979 0.0309 (-0.0271, 0.1130) 1.39 0.199
M-O 70 10 -0.0272 0.0683 0.0216 (-0.0761, 0.0216) -1.26 0.240
M-O 60 10 -0.0064 0.0760 0.0240 (-0.0608, 0.0479) -0.27 0.794
M-O 50 10 0.0414 0.0871 0.0275 (-0.0208, 0.1038) 1.51 0.166
3/6 decades were negative (however not significantly different than zero), 3/6 were positive (only the roaring 90's significant).
Russ Sears writes:
Having heard "sell in May" for eleven years, it is time to put paper to pencil. For the S&P index,
First day, whole mnth, month, year
1.30%, ? 5 2,010
0.54%, 5.17% 5 2,009
1.70%, 1.06% 5 2,008
0.26%, 3.20% 5 2,007
-0.41%, -3.14% 5 2,006
0.46%, 2.95% 5 2,005
0.92%, 1.20% 5 2,004
-0.07%, 4.96% 5 2,003
0.88%, -0.91% 5 2,002
1.35%, 0.51% 5 2,001
1.08%, -2.22% 5 2,000
First day 9 positives, average 0.73%, worst -0.41% 2006, best 1.70% 2008, Cumm. Binomial Dist 99.4%
Whole month 7 Positives, average +1.28%, best 5.17% 2009, worst -3.14% 2006, Cumm. Binomial Dist 94.5%
No wonder this dogma seemed annoying.