Beginner12–18 min

Lesson 6. Honest Backtests in MT5 — A Realistic Foundation

Learn the Strategy Tester modes, include real costs (spread/commission/swap/slippage), read PF/Max DD/Expectancy/Trades, and understand why backtest ≠ live. Finish with a hands-on test on a demo EA.

You installed and configured an EA. Now validate it with a realistic test before any forward demo.

6.1. Strategy Tester modes (when to use what)

Modeling modes change speed vs realism

Every tick based on real ticks

Uses broker tick history; highest realism; slower. Good for intraday/scalpers and entries sensitive to intrabar moves.

Best for: Scalpers, intrabar logic

Speed: Slowest

Every tick (synthetic ticks)

Realistic enough for most intraday systems; faster than real ticks if tick history unavailable.

Best for: Most intraday strategies

Speed: Medium

1-minute OHLC

Fast, approximate; acceptable for H1+ systems that don't rely on intrabar logic; not for scalpers.

Best for: H1+ swing systems

Speed: Fast

Tips

If your rules use stop/trigger inside the bar → use a tick mode.
Keep the same mode across comparisons for consistency.

6.2. Costs matter: spread, commission, swap, slippage

Add costs — or you're testing a fantasy

Spread

Set a fixed value (e.g., 1.0 pip) or "current" (less reproducible).

Commission

Set per round-trip (e.g., $7 per 1.00 lot for raw/ECN).

Swap (overnight)

Holding overnight incurs/earns swap; varies by symbol.

Slippage

Simulate typical slippage (e.g., 0.2 pips).

Cost Calculator

Spread (pips)

Commission ($/lot RT)

Slippage (pips)

Stop distance (pips)

Cost per trade (pips) ≈ spread + slippage + commission/10 (for EURUSD-like $10/pip/lot)

Cost per trade: 1.90 pips

Cost as R (with 30-pip SL): 0.063R

Example: 1.0 + 0.2 + 0.7 = 1.9 pips; with 30-pip SL → cost ≈ 0.063R

For JPY pairs pip value ≈ $9/lot; for XAUUSD ≈ $1/'pip' (check symbol specification).

6.3. Read the report: PF, Max DD, Expectancy, Trades

What the key metrics tell you

Profit Factor (PF)

Gross profit / Gross loss

Above 1.2 is a common minimum threshold for beginners.

Max Drawdown (absolute/relative)

Largest peak-to-trough loss

Keep it within your tolerance (e.g., ≤ 15% to start).

Expectancy (per trade)

Average R or cash per trade

Positive and stable matters.

Trades count

Total number of trades

More trades → more confidence. Aim for ≥ 200 trades before strong conclusions (for intraday).

Win rate & Avg trade

% winners / Average profit per trade

Context for expectancy; both can vary with strategy.

Optional metrics to note: Monthly breakdowns, equity stability, exposure.

6.4. Backtest ≠ live (and the role of forward demo)

Tests filter ideas; demo validates behavior in the wild

Execution: Live spreads widen; slippage increases at rollover/news; partial fills happen.
Data: Live tick stream ≠ historical reconstruction; gaps and feeds differ.
Latency/infra: Home PC/VPS and network affect timing vs Tester's idealized fills.
Settings drift: Wrong server/GMT offsets can invalidate time filters.

Forward Demo Plan

After a promising backtest, run 2–4 weeks on demo with the same risk and filters; track PF/Max DD/Expectancy, compare with backtest.

6.6. Sensitivity: see impact of costs and modeling

Small cost changes can flip your edge

Spread / Commission	$0/lot	$7/lot	$14/lot
0.5 pip	0.7 pips	1.4 pips	2.1 pips
1 pip	1.2 pips	1.9 pips	2.6 pips
1.5 pip	1.7 pips	2.4 pips	3.1 pips

* Slippage constant: 0.2 pips added to all values.

If your average target is 6–8 pips and costs are 2.1–2.7 pips, the margin for error is thin — be careful with scalping.

Lesson 6 Quiz

Test your understanding with 3 questions. Pass with 2/3 correct.

What's Next?

You've learned to run honest backtests with realistic costs and read key metrics. Next, we'll explore optimization — finding good parameters without overfitting to historical noise.

Next: Optimization Without Overfitting

Educational content only. Not financial advice. Trading involves risk of capital loss. Backtests and past performance do not guarantee future results. Start on demo and keep risk small.

6.1. Strategy Tester modes (when to use what)

Modeling modes change speed vs realism

Every tick based on real ticks

Uses broker tick history; highest realism; slower. Good for intraday/scalpers and entries sensitive to intrabar moves.

Best for: Scalpers, intrabar logic

Speed: Slowest

Every tick (synthetic ticks)

Realistic enough for most intraday systems; faster than real ticks if tick history unavailable.

Best for: Most intraday strategies

Speed: Medium

1-minute OHLC

Fast, approximate; acceptable for H1+ systems that don't rely on intrabar logic; not for scalpers.

Best for: H1+ swing systems

Speed: Fast

Tips

If your rules use stop/trigger inside the bar → use a tick mode.
Keep the same mode across comparisons for consistency.

6.2. Costs matter: spread, commission, swap, slippage

Add costs — or you're testing a fantasy

Spread

Set a fixed value (e.g., 1.0 pip) or "current" (less reproducible).

Commission

Set per round-trip (e.g., $7 per 1.00 lot for raw/ECN).

Swap (overnight)

Holding overnight incurs/earns swap; varies by symbol.

Slippage

Simulate typical slippage (e.g., 0.2 pips).

Cost Calculator

Spread (pips)

Commission ($/lot RT)

Slippage (pips)

Stop distance (pips)

Cost per trade (pips) ≈ spread + slippage + commission/10 (for EURUSD-like $10/pip/lot)

Cost per trade: 1.90 pips

Cost as R (with 30-pip SL): 0.063R

Example: 1.0 + 0.2 + 0.7 = 1.9 pips; with 30-pip SL → cost ≈ 0.063R

For JPY pairs pip value ≈ $9/lot; for XAUUSD ≈ $1/'pip' (check symbol specification).

6.3. Read the report: PF, Max DD, Expectancy, Trades

What the key metrics tell you

Profit Factor (PF)

Gross profit / Gross loss

Above 1.2 is a common minimum threshold for beginners.

Max Drawdown (absolute/relative)

Largest peak-to-trough loss

Keep it within your tolerance (e.g., ≤ 15% to start).

Expectancy (per trade)

Average R or cash per trade

Positive and stable matters.

Trades count

Total number of trades

More trades → more confidence. Aim for ≥ 200 trades before strong conclusions (for intraday).

Win rate & Avg trade

% winners / Average profit per trade

Context for expectancy; both can vary with strategy.

Optional metrics to note: Monthly breakdowns, equity stability, exposure.

6.4. Backtest ≠ live (and the role of forward demo)

Tests filter ideas; demo validates behavior in the wild

Execution: Live spreads widen; slippage increases at rollover/news; partial fills happen.
Data: Live tick stream ≠ historical reconstruction; gaps and feeds differ.
Latency/infra: Home PC/VPS and network affect timing vs Tester's idealized fills.
Settings drift: Wrong server/GMT offsets can invalidate time filters.

Forward Demo Plan

After a promising backtest, run 2–4 weeks on demo with the same risk and filters; track PF/Max DD/Expectancy, compare with backtest.

6.6. Sensitivity: see impact of costs and modeling

Small cost changes can flip your edge

Spread / Commission	$0/lot	$7/lot	$14/lot
0.5 pip	0.7 pips	1.4 pips	2.1 pips
1 pip	1.2 pips	1.9 pips	2.6 pips
1.5 pip	1.7 pips	2.4 pips	3.1 pips

* Slippage constant: 0.2 pips added to all values.

If your average target is 6–8 pips and costs are 2.1–2.7 pips, the margin for error is thin — be careful with scalping.

Lesson 6. Honest Backtests in MT5 — A Realistic Foundation

6.1. Strategy Tester modes (when to use what)

6.2. Costs matter: spread, commission, swap, slippage

6.3. Read the report: PF, Max DD, Expectancy, Trades

6.4. Backtest ≠ live (and the role of forward demo)

6.5. Practice: run a realistic backtest and read the report (10–15 min)

6.6. Sensitivity: see impact of costs and modeling

Lesson 6 Quiz

What's Next?

Lesson 6. Honest Backtests in MT5 — A Realistic Foundation

6.1. Strategy Tester modes (when to use what)

6.2. Costs matter: spread, commission, swap, slippage

6.3. Read the report: PF, Max DD, Expectancy, Trades

6.4. Backtest ≠ live (and the role of forward demo)

6.5. Practice: run a realistic backtest and read the report (10–15 min)

6.6. Sensitivity: see impact of costs and modeling

Lesson 6 Quiz

What's Next?