January 11th, 2012, 4:52 pm
Hi all,In literature, I often read papers claiming there are alphas in simple momentum strategies, for example, look back for 250 days and compute the 250-day-returns and then follow the momentum ...The problem is that for publishing papers it's okay to use results from such static/data-mined/over-fitted numbers: eg. 250.For real-trading, this parameter needs to be dynamic ... So I wanted to consult with your expertise: what are the good ways to make a momentum strategy dynamic?Could you please shed some lights on this?Thanks a lot!