What Trading Costs Can an MT5 Backtest Miss?

Check what the tester already includes

• Spread is the gap between the bid and ask. Real-tick testing can capture changes in spread within a minute. Generated ticks use the spread recorded for the relevant minute bar when available. If your test already uses the correct bid and ask fills, their spread is reflected in the P&L.
• Commission may be charged at entry, exit, or both. Compare the tester’s account settings with the fee schedule you are trying to represent.
• Slippage is the difference between a reference execution price and the fill price. MT5’s tester can model a delay on EA trade requests, during which price can change. Compare a no-delay run with a delay scenario. Neither measures how a live broker will fill every order.
• Overnight financing is a separate debit or credit on positions held through the relevant rollover. Verify the symbol’s long and short swap settings, including the day with a multiple-day charge. MT5’s profit-in-pips test mode omits commission and swap.

A 50-trade cost ledger

Suppose a mid-price model shows $180 gross across 50 trades. It has no costs included, so the following are additional hypothetical adjustments. For a run built from simulated bid and ask fills, omit any spread or slippage adjustment already reflected in those prices.

Net outcome = $180 − $60 − $40 − $70 − $18 = −$8.

Reconcile the run before changing the strategy

Keep the symbol, dates, sizing and EA inputs fixed while you test one cost assumption at a time. Record the tick mode, delay, commission and swap settings. Compare gross and net P&L, trade count, and representative entry and exit prices. If changing the execution model changes which orders filled, treat the runs as different trade sets. Compare individual fills before attributing the whole P&L difference to one cost.

Use the trading costs calculator for a round-trip estimate, then check how the resulting cost changes the break-even win rate.