Competition scoring deep dive: how Sharpe is calculated
For those wondering about the exact Sharpe calculation:
Important notes:
- Annualization uses 252 trading days
- Returns are computed at EOD, not intraday
- The risk-free rate is updated monthly
31 Replies
This is a common pitfall. Make sure your features are computed before the prediction date, not on it. That's subtle look-ahead bias.
Market regimes are the elephant in the room. A strategy that works in a trending market will fail in mean-reverting conditions.
I think the platform should add a paper trading mode so we can test strategies in a more realistic setting between competitions.
The biggest mistake I see newcomers make: optimizing for the wrong metric. Sharpe != best trading strategy. Consider Calmar, Sortino, and max drawdown.
For factor models, I'd strongly recommend the Fama-French 5-factor model as a starting point. It captures most systematic risk.
The key insight is that most alpha signals decay rapidly. You need to focus on signals with a half-life of at least 5-10 days.
The key insight is that most alpha signals decay rapidly. You need to focus on signals with a half-life of at least 5-10 days.