Window Dressing at the end of the month

It is commonly believed that a lot of funds buy additional stocks of the positions they are already holding in order to window dress their performance on the reporting day, which is usually the last day of each month.

I investigate whether we can profit in buying only at the second last day of a month and selling our position again at the closing price on the last day of the month.

This graph shows by how much the last trading day of a month outperforms other trading days on average.

Since 1930 only in 43% of the months we were able to discern an outperformance of the last trading day of the month versus the other days. We outperform on the last day of the month the other days by 0.06% on average.

When looking at the statistical test, we can find that there is indeed statistical significance.

Average last day of month compounded return = 0.08%
Average other day of month compounded return = 0.02%

What happens if we look at the period of January 2006 to October 2008? Unfortunately the result is no longer statistical significant. While we did have some months where we clearly outperformed on the last day of the month, the outperformance in relation to the variance of the outperformance is not enough to conclude that window dressing leads to an outperformance on the last day of the month.

Below the histogram for the result for the period of January 2006 to October 2008.