Plenty of info out there if you’re looking for ideas on calculating or tinkering with momentum strategies. Here’s one I stumbled across today:
Quote:
QMOM FACTSET ANALYTICS INSIGHT
QMOM targets the 10% of stocks with the highest total return over the last 12 months, excluding the most recent month. The fund also screens for consistency of momentum by excluding stocks with too many negative-return days during the 12-month period. This methodology produces a few dozen holdings which get weighted equally. QMOM launched as an actively managed fund but began tracking a new homegrown index in January 2017, a change which did not significantly alter the fund’s strategy or holdings. Lukewarm trading increases round-trip costs for QMOM.
So I tested this idea, and used parameters of between 120 and 180 negative days per year, so roughly half the days, and it made very little effect to my rotational systems. Too few negative days and the CAGR was impacted dramatically. Too many and there was no effect compared to my existing system.
I feel that optimising for negative days is counter to robustness, especially if there is a different number for each system. Basically you need the down days to give you the up days it would seem.