My crypto backtest had a sharpe of 13.4. Then I added trading costs

I was not trying to prove a grand theory about crypto. I was testing simple, falsifiable baselines and looking for one result worth a second experiment. Cross-sectional reversal looked almost too clea...