Performance differs depending on rebalancing frequencies in screener

I am getting results which vary significantly depending on the rebalancing frequency selected in the screener tool. I was hoping someone could shed light onto the reason.

Basically, I am starting from the universe “All Fundamentals”, then I am winnowing it down based on two rules:

  1. belong to GICS(10102010, 10102020)
  2. have MktCap > $10 MM
    The result from a simple back-test strategy (zeroed out cost parameters) show a long-term CAGR of 6.96% for 4-week rebalancing and 12.22% for weekly rebalancing. I would think that based on the low turnover rate from such simple rules would result in no significant difference in performance expectancy. However, the difference is obviously quite large.

I’ve already ruled out the possibility that a few outliers are responsible. See the chart below for relative frequency bar chart of monthly logarithmic returns from both rebalancing frequencies:

The expected outperformance of the higher rebalancing frequency is robust. But why?

The unexpected difference seems like due mainly to the presence of small caps – when I up the size constraint for mktcap to > $100 MM, then most of the delta goes away. But again, why would this be the case?

Anyway, here’s a link to the public screen: The https://www.portfolio123.com/app/screen/summary/185462?st=0&mt=1

Thank you for your consideration!



MktCap > 10 includes a lot of sub $1 stocks. When a stock falls below $10M MktCap because of falling price, rebalancing weekly gets out more quickly.

Adding a liquidity constraint like;

AvgDailyTot(30)>200000

makes the backtest more reasonable.

Walter