✅ Quick Checklist (TL;DR) #

🎯 Core Rules #

Keep These Principles in Mind:

• Simple is better: Fewer settings = more reliable results. Aim for 3 or fewer optimized parameters.
• More trades = more confidence: Try for 200+ trades minimum, 500+ is better.
• Test different markets: Your strategy should work in up, down, and flat markets.
• Include real costs: Always add spreads, commissions, and slippage to your tests.
• Never test and validate on the same data: Keep a separate period for final checks.
• Stable settings win: Good parameters work across different pairs and timeframes.

🔬 Backtesting Basics #

Data & Trade Count Guide #

Simple Tips:

• Use tick or M1 data for strategies under H1 timeframe
• Remove big news events if your strategy doesn't trade news
• Test on multiple pairs to see if your strategy is truly robust
• Always include realistic trading costs

⚙️ Optimization Tips #

The 3-Setting Rule #

Simple Optimization Steps:

📈 Advanced Optimization Techniques #

🔄 Walk-Forward Analysis Framework #

Enhanced Step-by-Step Process #

Implementation Example: #

• Total Data: 2012-2024 (12 years)
• Stage 1: Optimize 2012-2017 (5y IS) → Validate 2017-2018 (1.25y OOS)
• Stage 2: Optimize 2012.25-2017.25 → Validate 2018-2019.25
• Repeat until final OOS period ends at present
• Final Decision: Use parameters from last optimization ONLY if aggregated OOS results meet criteria:
  • Profit Factor > 1.3
  • Max Drawdown < 1.5x backtested DD
  • Sharpe Ratio > 0.8
  • Parameter values stable across stages

🔄 Single-Stage with Validation Process #

What Is This Process? #

This is a simple, practical method to find robust trading settings without over-complicating things. Instead of multiple walk-forward stages, we use one long optimization period followed by a dedicated validation period before going live.

Why Use This Approach? #

• ✅ Simpler to set up and manage
• ✅ Longer optimization period = more reliable results
• ✅ Clear separation between testing and validation
• ✅ Easy to spot over-optimized settings
• ✅ Works well with MT5's built-in forward testing

The Timeline Example (8-Year Total) #

What We Optimize #

In this single stage, we optimize these key components:

• Market context filters (when to trade based on market conditions)
• Strategy filters (entry/exit rules and conditions)
• Take Profit / Stop Loss ratios (R:R settings)
• Stop Loss methods (fixed, ATR-based, volatility-adjusted)
• Trailing Stop Loss settings (when and how to trail)
• Weekend / trading day rules (close trades before weekends or hold)

This focused approach lets us test meaningful combinations without exploding the number of tests.

Step-by-Step Workflow #

Step 1: Set Your Periods in MT5

• Total backtest period: 2019-2026
• In-Sample (Optimization): 2019-2024
• Forward/OOS (Validation): 2025-2026
• Enable “Forward Testing” in MT5 Strategy Tester

Step 2: Run Optimization on In-Sample Period

• Test your parameter combinations on 2019-2024 data
• Use your custom metrics: CAGR / Blended DD or CAGR / Avg DD
• Save the top 20-50 performing settings (not just #1)

Step 3: Validate on Forward Period (2025)

• Check how those top settings performed in 2025 (unseen data)
• Look for:
  • Similar equity curve shape (IS vs OOS)
  • Reasonable drawdown (not 2-3x worse than IS)
  • Consistent trade frequency and win rate
  • No single huge wins masking poor consistency

Step 4: Select Settings That Pass Validation

• Keep settings that are “good and stable”, not necessarily the absolute best
• Discard any setting where OOS looks materially different from IS
• Typical outcome: 2-10 settings per symbol pass this filter
• If nothing passes → discard the symbol entirely, move on

Step 5: Final Manual Check (“Every Tick” to Today)

• Run each passing setting with “Every Tick” data quality up to today
• Visually inspect equity curves for smoothness and consistency
• Prune down to 0-3 final settings per symbol for your portfolio

Step 6: Portfolio Building

• Combine your selected settings across symbols
• Check overall portfolio correlation and risk
• Start live trading with small size, monitor closely

Simple Selection Criteria #

Visual Example: Good vs Bad Validation #

🛡️ Validation Checklist #

Before Going Live: Quick Checks #

✅ Performance Check

• ☐ OOS results are within 20-30% of IS results (not identical, but close)
• ☐ Drawdown in validation is acceptable for your risk tolerance
• ☐ Strategy makes money in at least 2 different market conditions

✅ Equity Curve Check

• ☐ IS and OOS curves look similar in shape and smoothness
• ☐ No huge gaps or spikes that suggest data issues or overfitting
• ☐ Growth is steady, not dependent on one lucky period

✅ Trade Quality Check

• ☐ Minimum 50 trades in validation period (more is better)
• ☐ Win rate and profit factor are reasonable, not extreme
• ☐ Average trade duration matches your strategy design

✅ Final Prep

• ☐ Run “Every Tick” test to today for final settings
• ☐ Visually review each equity curve one last time
• ☐ Document your final settings and why you chose them

⚠️ Common Mistakes #

Simple List of What to Avoid #

Red Flags to Watch For:

• Profit Factor above 3.0 in backtest (probably too good to be true)
• Fewer than 50 trades in your validation period
• Best settings are at the very edge of your tested range
• One or two trades make up most of the profit

📚 Resources #

Helpful Tools & Learning #

🎥 Video Series

• Darwinex: Backtesting & Optimization
  https://www.youtube.com/watch?v=WBZ_Vv-iMv4&list=PLv-cA-4O3y95J6xmwSaCILL4FlGJZO0PJ

• Clear explanations of complex topics. Great for visual learners.

📖 Simple Reads

• Quantitative Trading by Ernest Chan – Practical and readable
• The Evaluation and Optimization of Trading Strategies by Robert Pardo – Focus on walk-forward and validation

🛠️ Tools