When execution tricks are really the strategy: separating edge from friction
Some scalping systems seem to make money from price prediction, yet the profit comes mainly from execution quirks, like capturing the spread with limit orders in a stable range or exploiting a feed that lags slightly. That is not a problem in itself, but it is worth understanding.
The danger arises when the quirk disappears. A broker changes its routing, a new risk filter is installed, or the spread structure shifts, and the edge vanishes overnight without any change in the market.
To diagnose, decompose results. Take the trade log and compute what the profit would have been using mid prices with no costs, then add costs back in gradually. If the paper edge is tiny and the entire profit sits in a favorable fill pattern, the strategy is fragile.
A system whose edge survives at mid price with pessimistic fills is far more robust.
Knowing the source of returns helps decide how much to risk and when to stop.
Has any system's profit turned out to be mostly execution quirk rather than market prediction?
When execution tricks are really the strategy: separating edge from friction
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: When execution tricks are really the strategy: separating edge from friction
One of my Asian limit systems was an execution trick dressed up as a forecast, and I only saw it when I stripped the fills back to mid price. It bought and sold EURUSD in the late Tokyo range with resting orders. On my live fills it showed about twenty-eight hundredths of a pip expectancy over eleven weeks. Recalculated at mid price with no costs, that figure shrank to four hundredths, which is basically noise. Almost the entire profit came from being queued at the inside and capturing a stable spread.
When the broker tightened the top of the book and I stopped getting those queue-priority fills, the same rules went to minus six hundredths live within three weeks. Nothing in the market regime had changed in a way I could point to. The routing had.
I retired it. Anything I test now has to show at least twelve hundredths of a pip at mid price after pessimistic fills before I give it credit for predicting anything. If the money only appears in the fill column, I treat it as borrowed from the broker's book, not as an edge I own.
When the broker tightened the top of the book and I stopped getting those queue-priority fills, the same rules went to minus six hundredths live within three weeks. Nothing in the market regime had changed in a way I could point to. The routing had.
I retired it. Anything I test now has to show at least twelve hundredths of a pip at mid price after pessimistic fills before I give it credit for predicting anything. If the money only appears in the fill column, I treat it as borrowed from the broker's book, not as an edge I own.
-
FXS London Scalper
- Posts: 976
- Joined: Sun Oct 18, 2026 12:25 pm
Re: When execution tricks are really the strategy: separating edge from friction
Profits from capturing the spread disappear when the broker changes execution rules. Separating execution gains from prediction gains in the stats shows how much of the edge depends on the broker.
FXS Official Team | London Session Desk
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: When execution tricks are really the strategy: separating edge from friction
This is a really important distinction, and it's one that has caught a lot of people out with EAs that look brilliant on one broker and then fall apart when they're moved.
A common version of this is a mean-reversion scalper, often on the yen pairs, that posts great live results on the first broker for months. After a switch to a different broker with nominally similar spreads, it starts losing almost immediately. After a lot of digging, it often turns out that most of the profit on the first account came from fills during fast moves. That broker's price feed lagged slightly behind the market at certain moments, so the EA was effectively buying a slightly stale price. On a cleaner feed, that advantage vanishes and what's left is a strategy with no real edge.
The test I'd suggest to anyone is to rerun your results with a fixed penalty per trade, say an extra half pip of slippage on every fill, and see whether the strategy survives. A real edge degrades gradually. A strategy built on execution quirks often flips from profitable to losing with a tiny cost change.
Another sign is when the profits cluster in very specific moments, like the first second after news or during rollover. That's where feeds and fills behave strangely, and it's also where brokers tend to tighten things up once they notice.
A common version of this is a mean-reversion scalper, often on the yen pairs, that posts great live results on the first broker for months. After a switch to a different broker with nominally similar spreads, it starts losing almost immediately. After a lot of digging, it often turns out that most of the profit on the first account came from fills during fast moves. That broker's price feed lagged slightly behind the market at certain moments, so the EA was effectively buying a slightly stale price. On a cleaner feed, that advantage vanishes and what's left is a strategy with no real edge.
The test I'd suggest to anyone is to rerun your results with a fixed penalty per trade, say an extra half pip of slippage on every fill, and see whether the strategy survives. A real edge degrades gradually. A strategy built on execution quirks often flips from profitable to losing with a tiny cost change.
Another sign is when the profits cluster in very specific moments, like the first second after news or during rollover. That's where feeds and fills behave strangely, and it's also where brokers tend to tighten things up once they notice.
Re: When execution tricks are really the strategy: separating edge from friction
FXS London Scalper wrote: Thu Nov 12, 2026 11:42 amProfits from capturing the spread disappear when the broker changes execution rules. Separating execution gains from prediction gains in the stats shows how much of the edge depends on the broker.
That distinction is best tested with two columns rather than a single profit line: mark-to-mid result before execution costs, then realized result after spread, commission and slippage. For example, suppose 400 round trips produce +0.12 pip per trade at mid, but average costs are 0.08 spread, 0.06 commission-equivalent and 0.05 slippage. Net is -0.07 pip; a green fill summary would be hiding a negative forecast edge. I would split this by pair and session too, because rollover or release minutes can dominate an overall mean. If fills are the only source of profit, the strategy may still be usable while that execution condition lasts, but the position size and stop plan should assume it can disappear. A rolling check of fill-to-mid and rejection rate can flag a change. It cannot guarantee the old routing advantage will return, so a separate out-of-sample period with pessimistic costs is the more honest gate before scaling.
It’s Fairman 