She only uses data from 1999-2005. That is a ridiculously insufficient amount of data to prove her hypothesis. That is 6 datapoints per month.
I suggest she look at how researchers proved the January effect and WHY it actually occured (tax selling in Dec). Then she analyze JPY data for a...