Cobra,
I think we need more result to see if this data are biased.
This data show that gain/loss around Fed meeting days. We need to have a non Fed meeting day to compare. However, it is very hard to get that. Fed meeting usually happens every month. There are only about 22 trading days in a month. This test already take out at least 10 days. We couldn't tell what did the market react to if the Fed impact is longer than that. Or the 2 day result after Fed meeting will be chosen for trading if the Fed impact is so short lived.
I'm not against you but just play the devil's advocate.
[ 本帖最后由 palm 于 2008-4-30 12:37 编辑 ] |