The main reason for this lack of statistical significance is the large standard deviation of daily returns, about 1 percent, which swamps in magnitude the difference in returns. We also investigate whether the identified difference in returns between new moon and full moon windows is persistent. As a measure of persistence, we calculate the percentage of years in which mean new moon daily returns are higher than mean full moon returns. For the DJIA, this number is percent, which is moderately above the 50 percent that would be expected by chance. This percentage also signifies.
