Skip to main content
|View as Markdown||

▶️ Running Backtests

Executing Your First Backtest​

Now that you've set up everything properly, it's time to run your backtest and see how BananaEA would have performed historically. This guide walks you through the execution process step-by-step.

tip

Goal: Execute a complete backtest and generate reliable performance data for analysis.

Quick Start: Running a Basic Backtest​

5-Step Process​

  1. Open Strategy Tester → Press Ctrl+R
  2. Configure Settings → EA, symbol, period, model, spread, dates
  3. Set Parameters → Load preset or use defaults
  4. Click Start → Begin backtest execution
  5. Review Results → Analyze performance metrics

Detailed Backtest Execution​

Step 1: Final Configuration Check​

Before clicking [Start], verify these settings one last time:

Strategy Tester Settings Tab:

Expert Advisor: BananaEA.ex4 ✅
Symbol: GER40.r ✅
Period: H1 ✅
Model: Every Tick ✅
Spread: 20 points ✅
From: 2023.01.01 ✅
To: 2025.10.20 ✅
Visual mode: Unchecked (for speed) ✅
Optimization: Unchecked (unless optimizing) ✅

Inputs Tab:

  • Parameters loaded correctly ✅
  • Preset applied (if using) ✅
  • Magic number unique ✅

Step 2: Start the Backtest​

Click [Start] Button

What happens next:

1. MT4 loads historical data for symbol/timeframe
   ↓
2. EA initializes (OnInit() executes)
   ↓
3. MT4 simulates each bar sequentially
   ↓
4. EA analyzes market conditions per bar (OnTick())
   ↓
5. Trades executed when signals detected
   ↓
6. Results compiled and statistics generated
   ↓
7. "Tester: stop testing" message appears

Step 3: Monitor Progress​

Progress Bar (Bottom of Strategy Tester):

Testing: 2024.03.15 14:00  [████████████░░░] 78%

Information Displayed:

  • Current Date: Date/time being tested
  • Progress Percentage: How much completed
  • Bars Processed: Number of bars analyzed
  • Ticks Simulated: Tick count (if "Every Tick" mode)

Estimated Time:

  • Quick Test (6 months, Open Prices): 30 seconds - 2 minutes
  • Standard Test (2 years, Every Tick): 5 - 15 minutes
  • Comprehensive Test (5 years, Every Tick): 20 - 45 minutes
info

Duration varies by: Computer speed, data quality, EA complexity, timeframe, and modeling quality.

Step 4: During Backtest Execution​

What You Can Do:

✅ Monitor Journal Tab - Check for errors or warnings ✅ Watch Progress - Observe completion percentage ✅ Minimize MT4 - Let it run in background ✅ Work on Other Tasks - No need to watch continuously

What You Should NOT Do:

❌ Don't Close MT4 - Backtest will abort ❌ Don't Change Settings - Wait for completion ❌ Don't Open Heavy Programs - May slow testing ❌ Don't Disconnect Internet - If using live data feed

Step 5: Backtest Completion​

Completion Message:

Tester: stop testing BananaEA-v3.13.21-meanReversion.ex4 (GER40.r,H1)

Results Available:

  • Results Tab: Trade-by-trade list
  • Graph Tab: Equity curve visualization
  • Report Tab: Comprehensive statistics

Visual Mode Backtesting​

When to Use Visual Mode​

Recommended For:

  • ✅ Verifying signal detection accuracy
  • ✅ Debugging entry/exit logic
  • ✅ Checking SL/TP placement
  • ✅ Understanding EA behavior
  • ✅ Confirming indicator display
  • ✅ Educational purposes (learning how EA trades)

NOT Recommended For:

  • ❌ Full 2-year backtests (too slow)
  • ❌ Optimization runs (disabled automatically)
  • ❌ Final performance validation

Enabling Visual Mode​

Step-by-Step:

  1. Strategy Tester → Settings tab
  2. Check "Visual mode" checkbox
  3. Configure other settings normally
  4. Click [Start]

What Appears:

┌─────────────────────────────────────────────┐
│ Chart Window (Visual Backtest)              │
├─────────────────────────────────────────────┤
│ Price Chart with:                           │
│ - Price bars progressing in real-time      │
│ - EA indicators overlaying                 │
│ - Trade arrows when signals fire          │
│ - SL/TP lines on open positions           │
│                                            │
│ [<< Prev] [Pause] [Play] [Next >>]        │
│ Speed: [______|_________] (slider)         │
└─────────────────────────────────────────────┘

Visual Mode Controls​

Playback Controls:

[Play] Button: Resume/continue backtest at current speed
[Pause] Button: Stop execution temporarily
[<< Prev]: Step backward one bar (review previous)
[Next >>]: Step forward one bar (advance manually)

Speed Slider:

Slowest [|_____________________] Fastest
         ← Detailed observation  Fast screening →
  • Slowest: See every tick, perfect for debugging
  • Medium: Balanced view, catch signals and trades
  • Fastest: Quick visual confirmation, less detail

What to Look For in Visual Mode​

Signal Verification:

  • Does EA detect signals when YOU see pattern?
  • Are trade arrows appearing at correct times?
  • Is entry logic working as expected?

Entry/Exit Accuracy:

  • Buy/Sell orders placed at correct prices?
  • Stop Loss levels appropriate?
  • Take Profit targets reasonable?

Indicator Behavior:

  • Indicators plotting correctly?
  • Signal arrows appearing on chart?
  • Multi-timeframe data syncing?

Trade Management:

  • Break-even adjustments triggering correctly?
  • Partial closes executing as expected?
  • Trailing stop following price appropriately?
warning

Performance Note: Visual mode is 10-50x slower than normal backtesting. Use for verification only, not full tests.

Handling Backtest Interruptions​

Stopping a Backtest Mid-Execution​

How to Stop: Click [Stop] button in Strategy Tester

What Happens:

  • Backtest aborts immediately
  • Partial results may be available
  • No report generated (incomplete data)
  • Must restart from beginning

When to Stop:

  • 🛑 Errors flooding Journal tab
  • 🛑 Realized wrong settings configured
  • 🛑 Need to change parameters
  • 🛑 Testing taking unexpectedly long

Common Errors During Execution​

Error: "No History Data"​

Cause: Insufficient historical data for test period

Solution:

  1. Tools → History Center (F2)
  2. Download data for symbol/timeframe
  3. Restart backtest

Error: "Testing Failed"​

Cause: EA compilation error or resource missing

Solution:

  1. Check Experts tab for error details
  2. Open MetaEditor → Compile EA (F7)
  3. Fix compilation errors
  4. Verify all includes (.mqh files) exist
  5. Restart backtest

Error: "Array Out of Range"​

Cause: EA trying to access unavailable bars

Solution:

  1. Check EA code for array access
  2. Verify sufficient bars available (e.g., Bars > 50)
  3. Add data validation in EA code
  4. Use longer test period with more data

Warning: "Not Enough Money"​

Cause: Initial deposit too small for EA's lot calculation

Solution:

  1. Increase initial deposit in Strategy Tester
  2. Reduce risk percentage in EA parameters
  3. Use smaller FixedLotSize if applicable
  4. Check LotMethod setting

Warning: "Invalid Stops"​

Cause: Stop Loss/Take Profit too close to entry price

Solution:

  1. Check symbol's minimum stop level (MODE_STOPLEVEL)
  2. Increase SL/TP distances in EA parameters
  3. Verify broker's stop level requirements
  4. Use larger pip buffers

Post-Backtest Actions​

Immediately After Completion​

Step 1: Quick Visual Check

Click [Open Chart] button
→ Look at equity curve shape
→ Check if results make sense
→ Verify trades appear on chart

Step 2: Review Key Metrics

Results tab → Report
→ Total Net Profit: Positive or negative?
→ Max Drawdown: Acceptable level?
→ Profit Factor: Review with drawdown and trade count?
→ Total Trades: Sufficient sample size?

Step 3: Save Results

Right-click in Report tab
→ Save as Detailed Report (includes graphs)
→ Choose location and filename
→ Example: "BananaEA_DAX_H1_2023-2025_EveryTick.html"

Organizing Backtest Results​

Recommended File Naming Convention:

[EA-Name]_[Symbol]_[Timeframe]_[DateRange]_[Model]_[Version].html

Examples:
BananaEA_DAX_H1_2023-2025_EveryTick_v3.13.21.html
BananaEA_NAS_M5_2024Q1_OpenPrices_v3.13.21.html
BananaEA_DAX_H1_Conservative_2023-2025_v3.13.21.html

Folder Structure:

Backtests/
├── 2023/
│   ├── January/
│   ├── February/
│   └── ...
├── 2024/
│   └── ...
├── 2025/
│   ├── October/
│   │   ├── BananaEA_DAX_H1_2023-2025_EveryTick_v3.13.21.html
│   │   ├── BananaEA_DAX_H1_Aggressive_2023-2025_v3.13.21.html
│   │   └── ...

Comparing Multiple Backtests​

Testing Different Parameters​

Scenario: Compare Conservative vs. Aggressive presets

Process:

  1. Test 1: Load Conservative preset → Run backtest → Save report
  2. Test 2: Load Aggressive preset → Run backtest → Save report
  3. Compare: Open both HTML reports side-by-side

Key Comparisons:

  • Profit: Which made more?
  • Drawdown: Which risked less?
  • Win Rate: Which had higher success rate?
  • Profit Factor: Which was more efficient?
  • Recovery: Which recovered from losses faster?

Testing Different Timeframes​

Scenario: Compare H1 vs. M15 performance

Process:

  1. Test 1: Period=H1 → Run backtest → Save report
  2. Test 2: Period=M15 → Run backtest → Save report
  3. Compare: Analyze trade frequency and risk

Expected Differences:

  • M15: More trades, faster signals, higher frequency
  • H1: Fewer trades, more selective, larger moves

Testing Different Symbols​

Scenario: Test DAX vs. NASDAQ vs. DOW

Process:

  1. Test 1: Symbol=GER40.r → Run backtest
  2. Test 2: Symbol=NAS100.r → Run backtest
  3. Test 3: Symbol=US30.r → Run backtest
  4. Compare: Which symbol suits EA best?
info

Note: Different symbols have different volatility, spreads, and trading hours. Results will vary significantly.

Backtest Validation Checklist​

After completing any backtest, verify:

Data Quality​

  • Historical data complete (no gaps)
  • Test period covers 2+ years
  • Includes different market conditions (trends, ranges, volatility)

Configuration Accuracy​

  • Realistic spread used (matches broker)
  • Commission included (if applicable)
  • Swap costs considered
  • Initial deposit matches real capital
  • Modeling quality appropriate ("Every Tick" for validation)

Results Sanity Check​

  • Total trades sufficient (50+ minimum for statistics)
  • Win rate realistic (30-70%, not extremes)
  • Drawdown fits your risk tolerance and account rules
  • Profit factor reviewed together with drawdown, trade count, and costs
  • No "too good to be true" metrics (100% win rate, zero drawdown)

Documentation​

  • Results saved with descriptive filename
  • Parameters documented (preset used or manual settings)
  • Test date recorded (when backtest run)
  • Notes added (observations, anomalies, insights)

Common Backtesting Scenarios​

Scenario 1: Quick Parameter Screening​

Goal: Quickly test if parameter change improves results

Setup:

  • Period: 6 months (recent data)
  • Model: Open Prices Only (speed)
  • Visual mode: OFF

Process:

  1. Run baseline test with defaults
  2. Change one parameter
  3. Run test again
  4. Compare results
  5. If promising, validate with "Every Tick" on longer period

Time: 2-5 minutes per test

Scenario 2: Comprehensive Validation​

Goal: Thoroughly validate EA before live trading

Setup:

  • Period: 3-5 years (extensive history)
  • Model: Every Tick (accuracy)
  • Visual mode: OFF (too slow for long period)

Process:

  1. Run full backtest on primary symbol (e.g., DAX)
  2. Save detailed report
  3. Analyze all metrics carefully
  4. Validate with forward testing
  5. Only then proceed to live testing

Time: 20-45 minutes

Scenario 3: Visual Verification​

Goal: Verify EA logic and signal detection

Setup:

  • Period: 1-3 months (short, manageable)
  • Model: Every Tick
  • Visual mode: ON

Process:

  1. Watch first few days carefully
  2. Verify signals trigger correctly
  3. Check SL/TP placement
  4. Observe trade management
  5. Confirm expectations match behavior

Time: 30-60 minutes (with pauses for observation)

Tips for Efficient Backtesting​

Time-Saving Strategies​

Use Progressive Testing:

1. Quick screening (6 months, Open Prices) → 2 minutes
   ↓ (If promising)
2. Standard test (2 years, Every Tick) → 10 minutes
   ↓ (If good results)
3. Comprehensive validation (5 years, Every Tick) → 30 minutes

Batch Testing:

  • Run multiple backtests overnight
  • Queue tests with different parameters
  • Review all results in morning

Template Settings:

  • Save Strategy Tester configurations
  • Load templates for consistent testing
  • Avoid reconfiguring each time

Common Time Wasters to Avoid​

❌ Visual Mode on Long Tests - Use only for short verification
❌ "Every Tick" on Initial Screening - Use faster models first
❌ Testing Same Thing Repeatedly - Document what you tested
❌ Not Saving Results - Always save before running next test
❌ Watching Progress Bar - Let it run, do other work

Next Steps​

Now that you know how to run backtests:

  1. Analyze Your Results → Analyzing Results
  2. Learn Optimization → Optimization Fundamentals
  3. Run Optimizations → Running Optimizations
  4. Validate Results → Validation & Forward Testing
tip

Pro Tip: Keep a testing journal. Document every backtest with date, parameters, results, and observations. This becomes invaluable for tracking what works and what doesn't!