How to Evaluate a Trading Strategy

Start with definable rules

A strategy you cannot write down precisely cannot be tested. Each rule should be unambiguous:

• Entry: what exact condition triggers a trade?
• Exit / expiry: when does the trade end?
• Filter: when should you not trade?
• Size: how much do you risk?

If two people would apply the rules differently on the same chart, refine them.

Sample size

Ten trades tell you almost nothing. Twenty winning trades in a row can happen by chance. Aim for hundreds of trades across different market conditions before drawing conclusions.

Expectancy

Expectancy estimates the average result per trade:

Expectancy = (Win rate × Average win) − (Loss rate × Average loss)

For fixed-payout contracts with an 80% payout risking 1 unit: at 52% accuracy, expectancy = 0.52 × 0.8 − 0.48 × 1 = −0.064 per unit. That is a losing strategy despite "winning more than half the time".

Drawdown and consistency

Look at the maximum drawdown, the longest losing streak and how results vary month to month. A strategy with good average results but brutal drawdowns may be impossible to follow emotionally.

Beware of overfitting

Adding filters until historical results look perfect usually creates a strategy that fits past noise. Warning signs:

• Many parameters with very specific values
• Excellent results on one asset or period only
• Performance collapsing when a parameter changes slightly

Forward test

After backtesting, test on data the strategy has never seen, ideally in a demo environment, for weeks. Compare results honestly to the backtest.

Red flags in strategies sold online

• Claims of extremely high accuracy without verifiable records
• Screenshots instead of complete trade histories
• Pressure to deposit quickly or join through a specific link
• No discussion of losing periods

Explore our Trading Strategies educational frameworks.

• #strategy
• #backtesting
• #statistics