A historical good Sharpe is usually a prerequisite, but not a guarantee of havin... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		maest on Dec 30, 2020 \| parent \| context \| favorite \| on: Ask HN: What's a side project you built to make mo... A historical good Sharpe is usually a prerequisite, but not a guarantee of having good live performance. I'm sure you're aware of overfitting backtestes and it sounds like that's what you were doing.

cg94301 on Dec 31, 2020 [–]

You can find patterns without necessarily overfitting to them. It's more a question of correlation vs causation. Looking for patterns reported in research papers would be more promising. But cannot automate that.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact