I am so much used to use data from the past, that I only did the analysis on the day a f t e r full moon. now here it is o n full moon. datastart 04.01.1984 dataend 06.05.2003 asset SP1 Comdty dailyGain to mean 0.02% dailyGain to median 0.04% range to mean 1.39% range to median 1.18% criteria full moon observations 167 dailyGain to mean 0.25% dailyGain to median 0.05% range to mean 1.41% range to median 1.24% ttest on dailyGain #DIV/0! could not figure out why ttest failed.
now here is finally the day before full moon. quite average. datastart 04.01.1984 dataend 06.05.2003 asset SP1 Comdty dailyGain t+1 mean 0.02% dailyGain t+1 median 0.04% range t+1 mean 1.39% range t+1 median 1.18% criteria full moon observations 167 dailyGain t+1 mean 0.03% dailyGain t+1 median 0.06% range t+1 mean 1.41% range t+1 median 1.24% ttest on dailyGain 0.10
backtesting to buy the open on full moon and get out MOC: 1984 -6.1% 1985 4.1% 1986 9.4% 1987 33.6% 1988 39.9% 1989 8.5% 1990 -6.6% 1991 13.2% 1992 -2.9% 1993 5.0% 1994 9.4% 1995 6.7% 1996 17.2% 1997 17.5% 1998 -3.4% 1999 -30.5% 2000 24.6% 2001 -11.7% 2002 75.5% 2003 -8.4% (no fees)
selling the open at the day after full moon and get out MOC: 1984 20.2% 1985 3.3% 1986 -20.5% 1987 13.7% 1988 7.0% 1989 -2.6% 1990 14.5% 1991 -5.8% 1992 -1.0% 1993 4.4% 1994 20.7% 1995 -5.0% 1996 8.2% 1997 17.4% 1998 -15.1% 1999 -14.9% 2000 14.1% 2001 3.3% 2002 26.9% 2003 -10.6% (no fees)
combining the two: 1984 7.1% 1985 3.7% 1986 -5.6% 1987 23.6% 1988 23.4% 1989 3.0% 1990 3.9% 1991 3.7% 1992 -1.9% 1993 4.7% 1994 15.1% 1995 0.8% 1996 12.7% 1997 17.4% 1998 -9.3% 1999 -22.7% 2000 19.3% 2001 -4.2% 2002 51.2% 2003 -9.5% (no fees)