One-page strategy specification
Course context: This lesson belongs to a practical beginner curriculum for Pakistan.
| Field | Required definition |
|---|---|
| Market context | Symbol, timeframe, Pakistan-time session and allowed regime |
| Entry | Observable conditions known before the order |
| Invalidation | Price/condition that proves the setup wrong |
| Position size | Cash risk ÷ stop distance in cash per lot |
| Exit | Stop, target, time exit and early-exit rule |
| No-trade filters | Spread, news/HMR, volatility and daily-loss limit |
| Costs | Spread, commission, swap, conversion and slippage model |
Backtest without hindsight
- Freeze the rules and parameter values.
- Select historical periods before viewing outcomes.
- Include trending, ranging and volatile regimes.
- Reveal bars sequentially or use chart replay.
- Record every qualifying signal, including losses and skipped trades.
- Deduct realistic costs.
- Reserve an untouched out-of-sample period.
Metrics that answer the real question
| Metric | Formula / interpretation |
|---|---|
| Expectancy | Win rate × average win − loss rate × average loss |
| Profit factor | Gross wins ÷ gross losses |
| Maximum drawdown | Largest peak-to-trough equity decline |
| Average R | Outcome divided by initial planned risk |
| Rule compliance | Fully compliant trades ÷ all decisions |
| Cost sensitivity | Result after base, stressed spread and slippage |
Forward-test bridge
Run the frozen strategy on demo in current market conditions. Record requested and executed prices, live spread, commission, swap and rejected orders. A backtest can validate historical logic; forward testing exposes timing, platform and discipline failures. Neither guarantees live results.
Rejection criteria
- Expectancy turns negative after realistic costs.
- Performance comes from one unusual period or a few outliers.
- Small parameter changes destroy the result.
- Drawdown exceeds the predefined tolerance.
- Rules require information unavailable at decision time.
- The learner cannot follow the process consistently on demo.
Research references
Exness Insights: backtesting guide · Demo forward testing · Trading rules and risk controls
Questions from a first-time learner
What is the minimum information a strategy needs?
Exact entry, invalidation, size, exit, costs, market/session and stop-trading rules.
How many backtest trades are enough?
There is no universal number; use a meaningful sample across regimes and preserve out-of-sample data.
Does profitable backtesting prove future profit?
No. It can be affected by overfitting, data bias and execution differences.
Why forward-test on demo?
It tests current execution, timing and rule compliance without risking real funds.