Comparing Trading Rule Contributions Across Market Regimes
Summary
The post explains how to examine performance by trading rule within a dynamically optimized strategy. Because positions depend on optimization, instrument selection, capital, and contract rounding, the author uses static-portfolio estimates as a proxy for each rule’s contribution. These estimates may overstate achievable results for a smaller account. The analysis separates weighted results, which reflect forecast weights in the aggregate strategy, from unweighted results, which show each rule at equal risk.
The author groups rules into sub-strategies to make comparisons manageable, then considers year-to-date, live versus backtested, ten-year, and full backtest periods. The reported year-to-date patterns favor divergent rules such as momentum and skew, while several convergent rules perform poorly; similar momentum variants also appear to offer limited diversification. The post notes that short-period Sharpe ratios are unstable, execution affected live results, and the figures omit some recent losses. The charts and underlying numerical detail are absent from the supplied text, so the comparisons cannot be independently assessed here.
Key ideas
- Static portfolio estimates can approximate rule contributions, but optimization and position path dependencies limit their accuracy.
- Weighted results show aggregate impact, while equal-risk unweighted results make individual rules easier to compare.
- Grouping rules into sub-strategies can make a large rule set easier to analyze.
- The reported year-to-date results favor divergent rules, though the sample is short and execution influenced live performance.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.