1. Strategy input
- Original prompt
- Universe and market scope
- Strategy assumptions and constraints
Template
Quant strategy backtest report template
A reviewable backtest report shows more than a return number: it explains the strategy input, rules, data conventions, risk metrics, diagnosis and the next validation steps.
2. Strategy spec
- Factor and signal definitions
- Entry / exit rules
- Risk constraints and tunable parameters
3. Backtest setup
- Historical window
- Transaction cost
- Rebalance frequency and holding count
4. Metric results
- Annual return
- Max drawdown
- Sharpe
- Win rate
- Turnover
5. Diagnosis
- Return source
- Drawdown source
- Parameter sensitivity
- Failure scenarios
6. Next steps
- Parameter optimization
- Out-of-sample validation
- Transaction-cost sensitivity
- Human review
How FiClaw produces the report
FiClaw starts from a natural-language prompt, generates the strategy spec and Python code, submits a real backtest and then captures the metrics, diagnosis and next optimization steps into a reviewable report.