little crazy idea , could it work ?

This is all theory, and I'm posting it to see what others think

During each trading day , we have the best preformer and the worst preformer on the 500 stocks of the sp500, that's a fact , we have best performer and worst performer of the year , that's a fact , we have best performer and worst performer of the decade , that's a fact too .

Now

Let's say I can buy 10$ in each of these stocks , that will equal to 5000$( assuming they are 500 stocks ) , Now for sure I will have the best performer in my list and the worst performer , cool.

Now I'll close the stocks that gained 3% , leave the rest

So for each stock that gained 3% I'll have 13$

I'll take that 13$ and buy again in those stocks , each 13$

Next day again we will have the best and worst performer .

Basically if I buy 10$ in a stock I'll leave it until it gains 3%

Or let it set there indefinitely

If it gains 3% I'll sell it and buy again with 13$ , basically compounding my winings.

For example in 2010 to 2022 , netflix gained 1425%

If I bought 10$ and sold at 3% profit , I could have done that 475 times , which totals to 1.5 million

What's wrong here ?

Remember I'm deploying 10$ at each and every stock of the sp500