Why Your EA Does Worse Live Than in the Backtest, and Where the Money Goes

Almost every EA trader knows the feeling. The Strategy Tester shows a smooth, rising curve. You run the same EA live, on the same symbol, over the same weeks, and the result is smaller. Sometimes a little, sometimes a lot.

The usual explanations are vague: "slippage", "the broker", "the tester is too optimistic". They may all be true, but none of them tells you how much each one cost, or what to do about it. Here is a real example where every cent of the gap is accounted for, and a method you can apply to your own EAs.

A real example: 48 trades, €145.62 missing

One strategy, one live account, and a Strategy Tester report of the same EA over the same period. Strategy Ledger Pro's COMPARE view matched each of the 48 live trades to its twin in the backtest, then scaled the backtest to the lot sizes actually traded live, so the two sides are compared like for like.

The lines add up exactly to the gap. Nothing is left as "unexplained".

What the example teaches

Costs were not the problem. Commission cost 27 cents. Swap cost nothing. If you had blamed the broker's fees, you would have been looking in the wrong place.

Prices were. Three quarters of the gap came from the prices themselves: the live fills against the prices the tester assumed, at entry and at exit. That includes spread and slippage, and the difference between the prices recorded in history and the prices you actually get at the moment of execution. An EA that targets small moves feels this most, because a fraction of a pip is a large share of its profit per trade.

Missing and extra trades matter. One trade the backtest took never happened live, and it alone explains 11% of the gap. Live and backtest rarely trade exactly the same signals: a terminal restart, a slightly different price at the decision moment, or a connection drop is enough. A total-profit comparison hides this completely.

Timing is real money. Two pairs of trades opened at different moments live and in the tester. The same idea, entered a little later, costs €18.76 here.

A five-point check for your own EAs

1. Match trades, not totals. Comparing total profit tells you there is a gap, not why. Pair each live trade with its backtest twin.
2. Compare at the same size. If the backtest ran at a different lot size, scale it to your live lots first. Otherwise the gap is mostly a sizing difference.
3. Separate prices from costs. Commission and swap are easy to blame and easy to measure. Price differences are usually larger and much harder to see.
4. Count missing and extra trades. Every trade one side has and the other doesn't is a full trade's profit or loss in the gap.
5. Check the clock. A wrong time offset between the tester and your server shifts every match. Get the offset right before you trust any comparison.

Then ask: does the strategy survive real prices?

Once you know where the gap comes from, the next question is whether the EA still makes sense with it. Strategy Ledger Pro's What-if Simulator replays a backtest's trades at other sizes and with other costs: a fixed lot, equal risk, equal drawdown or a risk budget, on one balance. You can price it with your own measured slippage and commission from your live account, or double the slippage with 2× STRESS and see whether the edge is still there.

Every figure in the simulator is clearly marked as simulated, and its limits are stated on screen: it replays closed trades with hindsight and does not model margin calls. It answers "what would this have done", not "what will it do".

Try it on your own account

Strategy Ledger Pro is read-only. It never places, changes or closes a trade, and nothing leaves your terminal. COMPARE works with a standard MT5 Strategy Tester report and the history of your live account.

• Strategy Ledger Pro on the MQL5 Market
• Strategy Ledger on the MQL5 Market
• The manual
• How to evaluate your broker's execution
• Questions and discussion: AureusAI Trading Suite channel

What does the gap look like on your EAs? Share your numbers in the comments.