Research Methodology
Calculation methods and scope for backtests, event studies, and cost models.
Backtest design
Backtests report benchmarks, costs, exposure, maximum drawdown, and parameter sensitivity. A single best parameter is not treated as a tradable conclusion; out-of-sample and cross-year stability take priority over peak historical return.
- Fees and slippage are applied to every fill.
- Training and test samples are split chronologically.
- Mean, median, win rate, and maximum drawdown are shown together.
Editorial and revision policy
Each study states its sources, calculations, and material limitations. Method changes are logged publicly, and automated headlines are kept separate from original research.