Super Seven Days, shared by Ken Drees
Thackray's 2011 Investor's Guide refers to the effect as "Super Seven Days": The last four days of the month and the first three days of the following month typically yield higher returns than other periods during the month. Reasons for strength are month end "window dressing" by institutional investors and new monthly fund inflows into pensions and mutual funds that subsequently are invested into equity markets.
The month-end trade phenomenon is notable around Halloween. Gains are not related to sales of candy and spooky costumes.
Kim Zussman responds:
Checked this with SPY (2000-2010) by comparing returns of super seven days to the seven days just before (paired t-test):
Paired T for super - presuper
N Mean StDev SE Mean
super 129 0.006008 0.031161 0.002744 T=2.43
presuper 129 -0.003947 0.030088 0.002649
Difference 129 0.009955 0.046565 0.004100
Definitely worked in the whole period. However dividing data into 2005-10 and 2000-05 shows the effect was mainly from the older period and recently faded out:
Paired T for 2005-2010 - 2005-2010pre
N Mean StDev SE Mean
2005-2010 63 0.006552 0.034890 0.004396 T=1.4
2005-2010pre 63 -0.002733 0.032691 0.004119
Difference 63 0.009285 0.051677 0.006511
Paired T for 2000-2005 - 2000-2005pre
N Mean StDev SE Mean
2000-2005 66 0.005489 0.027395 0.003372 T=2.1
2000-2005pre 66 -0.005104 0.027578 0.003395
Difference 66 0.010594 0.041495 0.005108