Backtesting

Test your strategy on historical data before risking real money.

---

Backtesting Process

Step 1: Define Rules

Write clear rules:
- Entry condition
- Exit condition
- Stop location
- Position sizing

Step 2: Choose Period

Minimum:
- 1+ year of data
- 100+ trades

Avoid:
- Only recent data

Step 3: Execute Tests

Method 1: Manual
- Scroll charts
- Apply rules
- Record trades

Method 2: Automated
- Code strategy
- Run backtest
- Get metrics

Step 4: Analyze Results

Metrics to track:
- Win rate
- Profit factor
- Max drawdown
- Average trade

---

Backtesting Metrics

Primary Metrics

Metric	What to Calculate	Target
Win Rate	Wins ÷ Total trades	>40%
Profit Factor	Gross profit ÷ Gross loss	>1.5
Max Drawdown	Peak to trough	<20%
Expectancy	Avg win × Win% - Avg loss × Loss%	>0

Secondary Metrics

Metric	What to Calculate
Avg Winner	Average win size
Avg Loser	Average loss size
Avg Bars	Holding time
Max Winners	Consecutive wins
Max Losers	Consecutive losses

---

Backtesting Mistakes

1. Over-Fitting

Problem: Curve-fit to historical data.

Solution:

• Save 20% for out-of-sample
• Use realistic parameters

2. Ignoring Costs

Problem: No spread/commission.

Solution:

• Include costs in results

3. Survivorship Bias

Problem: Only testing survivors.

Solution:

• Include delisted stocks

4. No Walk-Forward

Problem: Optimizing on all data.

Solution:

• Walk-forward testing

---

Backtesting Checklist

☐Rules written
☐Period defined
☐Data reliable
☐Costs included
☐Metrics calculated
☐Out-of-sample done
☐Results acceptable

---

Key Takeaways

1. Test before trade — Validate strategy
2. Save data — For out-of-sample
3. Include costs — Spread, commission
4. Metrics matter — Win rate, PF, DD
5. Forward test — After backtesting

---

Related

demo-trading.md >>>

trade-setups.md >>>

exercises.md >>>

Backtest to validate. Forward test to confirm.