Research methods
Trading Strategy Comparison
A fair strategy comparison uses the same market period, capital rules, cost model, data frequency, and benchmark. Return differences matter only alongside drawdown, volatility, exposure, turnover, and sample size.
Align the test conditions
Different date ranges or cost assumptions can explain apparent performance gaps. Normalize the inputs before attributing differences to strategy logic.
Compare behavior, not just rank
Examine when each strategy gains or loses, how correlated their returns are, and whether differences persist across market regimes.
Treat small differences cautiously
Backtests are estimates under assumptions. Minor metric differences can be noise, especially when trade samples are small or many alternatives were tested.
Research checklist
- Identical test periods
- Consistent execution assumptions
- Risk and exposure
- Multiple-testing bias
Questions and answers
How do I compare two trading strategies?
Use identical data and assumptions, then compare return, drawdown, volatility, exposure, turnover, costs, and consistency across subperiods.