Some data in various hold times on a Jim screen.
"Member of Nasdaq 100
price / (52 week high + 52 week low) top 10"

GTR1 gives the average CAGR for all the variations of start date, but to me that is optimistic rather than realistic. Average is great, but the worst is more important because that's where the risk is. It's better for surprises to be happy rather than bad.

I would like to see the lowest 1/5th (20%) of the returns by start date. From a risk standpoint that's where I would like to count on.

So here are some CAGR stats and the lowest 5 returns for hold periods of 1 week, 2 weeks, 1 month, 2 months, and 1 year.
(20th percentile means 20% are below this and 80% are above this. Median is 50th percentile.)

# days    5         10        21        42        252
Avg       21.74     21.88     21.84     21.23     15.81
Median    21.47     22.02     22.07     21.09     15.73
Avg 5 low 21.74     21.19     20.31     18.57     12.00

20th percentile:
          20.26     21.45     20.79     19.53     14.30
                                                  
Lowest 5:
          19.72     19.46     19.66     18.00     11.48
          20.39     21.30     19.94     18.67     12.10
          21.47     21.48     20.49     18.70     12.11
          23.06     21.68     20.66     18.74     12.13
          24.07     22.02     20.79     18.75     12.16

One year (252 days) hold for gains to be long term. This is clearly suboptimal for this screen.
Hold from 1 week to 2 months have roughly similar return.