Risk-adjusted returns: Sharpe vs Sortino in competitions
I noticed the scoring uses Sharpe ratio as a key metric. Has anyone considered whether optimizing for Sortino would yield different rankings?
Sortino only penalizes downside volatility, which seems more aligned with real-world portfolio management. Thoughts?
40 Replies
For time-series cross-validation, I've found that 5 expanding windows with a 21-day embargo works well for daily data.
Good point about overfitting. My rule of thumb: never trust a backtest with fewer than 500 observations in the out-of-sample period.
Be careful with the Kelly criterion for position sizing. Full Kelly is way too aggressive. I use quarter-Kelly in practice.
I've found that sector neutrality is a key factor in the scoring. Strategies that are long one sector and short another tend to underperform.
For time-series cross-validation, I've found that 5 expanding windows with a 21-day embargo works well for daily data.
The competition scoring docs could definitely be clearer. I spent 2 hours debugging what turned out to be a normalization issue.
I ran a quick backtest on this idea and got a Sharpe of about 1.2 before costs. Not bad for a simple strategy.
I've found that sector neutrality is a key factor in the scoring. Strategies that are long one sector and short another tend to underperform.
Market regimes are the elephant in the room. A strategy that works in a trending market will fail in mean-reverting conditions.
I've found that sector neutrality is a key factor in the scoring. Strategies that are long one sector and short another tend to underperform.
I disagree about the GARCH approach. In my experience, realized volatility estimators (like the Rogers-Satchell estimator) outperform parametric models.