Forex Trading Costs for EAs: Spreads, Commissions and Swaps

An EA pays more than the spread shown in Market Watch. The relevant figure is the total cost produced by the strategy’s position size, holding time, trade frequency and execution quality. A fair comparison puts every cost into the same unit.

The five costs an EA can face

Component	When it applies	Why it varies
Spread	At entry and effectively at exit.	Symbol, account, session, news and liquidity.
Commission	Per side or round turn, often per lot.	Account type, volume tier and currency conversion.
Swap / financing	When a position remains open across rollover.	Direction, instrument, day and broker schedule.
Slippage	When the fill differs from the requested price.	Volatility, liquidity, size, latency and routing.
Other charges	Depending on broker and account.	Inactivity, data, conversion or withdrawal policy.

For strategy analysis, separate trading costs from non-trading charges. A withdrawal fee affects the account owner, but it does not change the expectancy of each signal. Spread, commission, swap and slippage do.

Convert costs to one comparable unit

Choose account currency, pips, points or money per standard lot and convert every component consistently. Check the symbol’s contract size and tick value instead of assuming that one pip has the same monetary value on every pair and account currency.

Simple round-turn estimate

Total trade cost ≈ entry spread + exit spread or closing spread effect + round-turn commission + swap for the holding period + net slippage. Use observed values rather than advertised minimums.

If commission is quoted per side, double it for a complete open-and-close cycle. If the EA scales into several positions, count every ticket. For partial closes, verify how commission and financing are applied in the actual account history.

Match the cost model to the EA

A high-frequency or short-target EA is usually more sensitive to spread and slippage. A swing system that holds for days may be more sensitive to swap. A grid strategy can accumulate spread, commission and financing across several simultaneous tickets, so cost should be evaluated at basket level as well as per trade.

Short horizonScalping

Focus on spread distribution, fill quality and commission per round turn.

Multi-daySwing trading

Model long/short swap and the broker’s triple-swap schedule.

Multiple ticketsGrid or basket

Add costs across the full basket and peak exposure.

Cost sensitivity should be tested before optimizing entry rules. A backtest built with an unrealistically low fixed spread can make a fragile strategy appear robust.

Build a representative cost sample

Collect typical spreads during the exact hours the EA trades.
Include rollover, session open and news only if the EA can trade there.
Export commission and swap from closed live or demo trades.
Calculate requested-versus-filled price for entries and exits.
Group results by symbol, session, direction and lot size.
Stress the model with worse-but-plausible costs before risking capital.

Use median values for a normal-case model and upper-percentile adverse values for a stress case. The goal is not to predict every fill, but to learn whether the strategy still has room after realistic friction.

A fair account comparison workflow

Same broker legal entity and account currency

Same symbol contract and lot size

Same trading hours and observation dates

Spread distribution, not minimum spread

Commission converted to round-turn cost

Both long and short swaps checked

Execution errors and slippage included

Enough observations for the EA’s normal behavior

Do not choose an account on cost alone. Platform stability, regulatory protections, withdrawal reliability and the broker’s execution policy remain separate requirements.

Official reference

The MT4 Contract Specification displays spread, contract size, tick value, volume limits, swap and trading sessions supplied by the broker. Treat those fields and the account’s actual history as the starting data for your calculation.