Backtest & Optimize EAs Like a Pro

Professional EA validation demands rigorous backtesting beyond basic MT5 runs—aim for real-market replication via tick data, variable spreads, and forward-testing. Setup Phase: Download 99% quality tick data (Dukascopy/Tick Data Suite). MT5 Strategy Tester > Expert tab > Visual mode off > Spread=1.2 pips (model current). Period: 5+ years (2020-2025), deposit $10K, leverage 1:100.

Core Metrics to Target:

• Profit Factor (PF) >1.5 (gross profit/loss)

• Sharpe Ratio >1.2 (risk-adjusted)

• Max Drawdown (MDD) <15%

• Recovery Factor >3 (profit/MDD)

• Expectancy >0.5 pips/trade

Walk-Forward Analysis (Gold Standard): Split data 70/30 (in-sample/out-sample). Optimize on 2020-2023 (genetic algorithm, 5000 generations), test 2024-2025 forward. Roll monthly: Retrain Q1 on new data. Example: EMA crossover EA optimizes Fast=12/Slow=26 in-sample, validates 14/21 out-sample (PF drops <10% = robust).

Optimization Best Practices:

1. Parameter Ranges: EMAs 8-50, RSI 10-21, ATR multipliers 1-3. Avoid extremes.

2. Genetic vs. Exhaustive: Genetic for 10+ params (faster, 95% accuracy).

3. Multi-Currency: Optimize EUR/USD, validate GBP/USD (no overfitting).

4. Out-of-Sample Test: >80% in-sample performance retained.

5. Monte Carlo: Reshuffle trades 1000x—expectancy holds?

Common Pitfalls & Fixes:

• Curve-Fitting: Solution—Walk-forward + smaller datasets.

• Look-Ahead Bias: Use shift=1 for indicators.

• Spread Slippage: Model commissions ($7/lot round-turn).

• News Gaps: Skip high-impact events or add FFCal filter.

Sample MQL5 Walk-Forward EA Code:

#property strict

extern double Lots = 0.01;

extern int FastEMA = 14;

extern int SlowEMA = 21;

extern double ATRMult = 2.0;

extern int Magic = 123456;

// OnTick logic

double fast = iMA(NULL,0,FastEMA,0,MODE_EMA,PRICE_CLOSE,1);

double slow = iMA(NULL,0,SlowEMA,0,MODE_EMA,PRICE_CLOSE,1);

double atr = iATR(NULL,0,14,1) * ATRMult;

if(fast > slow && OrdersTotal()==0) {

OrderSend(Symbol(),OP_BUY,Lots,Ask,3,Ask-atr,Ask+atr*1.5,"WF Buy",Magic);

}

Source / reference to read on and on.

Here we share the references and sources that inspire and support our blog content, ensuring you get well-researched and reliable information.

https://blog.quantinsti.com/algorithmic-trading-top-blogs/

https://www.maheshwariandco.com/blog/sebi-algo-trading-regulations-2025/