Dear all,
I’ve been researching ranking factors and found that many of them (mostly technical indicators) have high performance in a 16y backtest with weekly rebalance without any transaction costs (slippage and comission) but their return turns negative when backtested with variable slippage and 0.5 cent per share comission due to very high turnover.
Examples:
AvgVol(10) / AvgVol(60) - 24.87% annual return without trading costs, -4.35% with trading costs, 2,479.17% annual turnover
SIRatio / SIRatioPM - 19.28% annual return without trading costs, -2.1% with trading costs, 1791.07% annual turnover
StochK(14) vs. Industry - 25.67% annual return without trading costs, -17.75% with trading costs, 4254.72% annual turnover
UpDownRatio(20,0) vs. Industry - 18.78% annual return without trading costs, -5.32% with trading costs, 2448.73% annual turnover
The universe used in the tests is Close(0) > 1, Vol > 100K, MktCap > 100M.
Here is what I have tried so far:
- Increasing the period of the factor. It reduces the turnover and the trading costs, but most of the times it also reduces the indicator’s ability to find good stocks (the return is actually lower);
- Increasing the rebalance period from 1 Week to 4 or 8 weeks, didn’t help;
Do you have any other ideas on how to use these factors and minimize turnover so our profit doesn’t get eaten by transaction costs?
Thank you.
Bruno