Traders often notice that a strategy behaves well on a demo account but worse on live, even with the same broker. Several reasons are possible, and not all of them suggest anything improper.
Demo accounts may fill orders at the quoted price without simulating liquidity, while live orders interact with real liquidity and may slip. Demo servers may use a slightly different price feed. Order sizes that are easy on demo may be large relative to available liquidity on live, especially during news.
A method to measure the difference is to run both accounts at the same time with the same EA and log requested and filled prices. Comparing average slippage over a few hundred trades gives a concrete number.
If live slippage is consistently negative while demo shows none, that difference belongs in any backtest assumptions.
Asking the broker about their execution model and liquidity providers can also clarify what to expect.
Have you measured the gap between your demo and live fills?
Demo and live accounts on the same broker: why fills can differ
-
FXS London Scalper
- Posts: 976
- Joined: Sun Oct 18, 2026 12:25 pm
Demo and live accounts on the same broker: why fills can differ
FXS Official Team | London Session Desk
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: Demo and live accounts on the same broker: why fills can differ
I did, side by side for six weeks, with the same expert advisor submitting the same EURUSD London entries on a demo account and a live account at the same broker. Across three hundred and twelve paired trades the demo filled with a tiny improvement, about four hundredths of a pip on average, while live slipped against me by thirty-one hundredths. During one payroll week the live average was minus one point four pips and the demo was almost flat.
Away from news the gap shrank. On quiet London mornings live was only about twelve hundredths worse than demo, which I can live with. The gap jumped again once I tried live size above point four lots, when average slippage moved out to about half a pip.
I now add three tenths of a pip to every backtest and demo fill before I trust an expectancy number, and I keep live orders at or under that point-four-lot line on this name. Support said the demo uses a synthetic book and live goes to their liquidity stack, which matches the two logs.
Away from news the gap shrank. On quiet London mornings live was only about twelve hundredths worse than demo, which I can live with. The gap jumped again once I tried live size above point four lots, when average slippage moved out to about half a pip.
I now add three tenths of a pip to every backtest and demo fill before I trust an expectancy number, and I keep live orders at or under that point-four-lot line on this name. Support said the demo uses a synthetic book and live goes to their liquidity stack, which matches the two logs.