Mid price is a fiction that charts love
The midpoint between bid and ask is a tidy number. Many analysis tools use it, and it makes charts look smooth. Nobody can trade at it.
A buyer pays the ask. A seller receives the bid. The mid is a reference, not a price anyone gets, and in wide spread conditions the gap between the mid and the real fill is large.
Backtests that fill at the mid understate costs by half the spread on each side. For a scalper, that is the difference between a profitable strategy and a losing one.
The honest approach is simulating buys at the ask and sells at the bid, with a penalty for slippage. Anything else is giving the strategy a gift it will not receive in live trading.
Mid prices do have a place, such as measuring true market movement without the noise of spread changes. They just do not belong in the part of the test that decides profit.
One honest check: re-run a backtest with entries at the ask and exits at the bid, and compare the profit factor. The drop is often startling, and it is the first real information about whether the strategy has an edge at all.
Does the backtest ever fill at the mid price, intentionally or not?
Mid price is a fiction that charts love
Mid price is a fiction that charts love
It’s Fairman 
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: Mid price is a fiction that charts love
My charting package used mid prices by default, and I did not realise it until I compared a backtest trade by trade against my live fills. The setups lined up perfectly, but every live entry was half a spread worse and every exit was half a spread worse again. On a system aiming for five pips, that alone took a big bite out of each trade.
The worst cases were the trades near the Asian session close and the rollover hour, when the spread can widen a lot. On the mid-price chart those entries looked clean, yet in reality the ask was several pips above where the chart said I bought. Several of my best backtest trades turned out to be ones that could never have been filled as shown.
I now keep a bid chart and an ask chart side by side when designing anything for scalping. The mid still helps for spotting structure, but every entry and exit I test uses the real side of the quote.
The worst cases were the trades near the Asian session close and the rollover hour, when the spread can widen a lot. On the mid-price chart those entries looked clean, yet in reality the ask was several pips above where the chart said I bought. Several of my best backtest trades turned out to be ones that could never have been filled as shown.
I now keep a bid chart and an ask chart side by side when designing anything for scalping. The mid still helps for spotting structure, but every entry and exit I test uses the real side of the quote.