Spread in historical data: the number that is usually wrong
A backtest needs a spread. Many testing environments use a fixed value, or the spread recorded at the bar's open. Neither matches reality.
Spreads vary within a bar. They widen during news, narrow during busy periods, and jump at rollover. A strategy that enters at a moment of tick activity often does so at a wider than average spread.
Assuming a fixed 0.8 pips when the actual spread at entry averaged 1.2 overstates performance by 0.4 pips per trade. For a strategy with a 2-pip average gain, that is a 20 percent error.
Better testing uses a variable spread from real recorded data. Even a rough profile by hour of the day improves on a fixed number.
Another trick is to add a random penalty to simulate adverse spread conditions, then see whether the strategy survives. A system that falls apart with an extra half pip of cost was never robust.
How sensitive is the strategy to an increase of half a pip in cost per trade?
Spread in historical data: the number that is usually wrong
-
FXS Prop Desk
- Posts: 1689
- Joined: Wed Oct 07, 2026 9:52 am
Re: Spread in historical data: the number that is usually wrong
A practical middle ground is to record spreads live for a few weeks and build an hourly spread profile, then apply the hourly average in the backtest. It is not perfect, but it is much closer to reality than a fixed number.
FXS Official Team | The Idea Lab & Strategy
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: Spread in historical data: the number that is usually wrong
I tested exactly that on a short term mean reversion system on EURGBP, and it was humbling. With the default fixed spread the equity curve looked steady, averaging a little under two pips per trade. Adding an extra half pip to every trade cut the net result by roughly a third, and when I applied the higher spread only to trades entered during the first minutes after the London open and around the rollover hour, the system went flat for the year. Almost all of the paper edge was sitting in the moments when the real spread was worst.
The other thing I found was that the strategy's entries were not random relative to spread. It fired most often when volatility jumped, and those were exactly the moments my broker widened. A flat average understated the cost far more than I expected.
These days I won't take a system forward unless it still makes money with an extra half pip on every trade and with spread doubled during its busiest hour. Very few survive, but the ones that do have held up live.
The other thing I found was that the strategy's entries were not random relative to spread. It fired most often when volatility jumped, and those were exactly the moments my broker widened. A flat average understated the cost far more than I expected.
These days I won't take a system forward unless it still makes money with an extra half pip on every trade and with spread doubled during its busiest hour. Very few survive, but the ones that do have held up live.