WHAT A REALISTIC
PROFIT FACTOR
ACTUALLY LOOKS LIKE.

Profit factor is gross profit divided by gross loss. A strategy that made £11,000 and lost £10,000 has a profit factor of 1.10. Above 1.0 it made money; below, it did not.

That is the entire definition, and it is why the number gets quoted so freely. What it does not tell you is how much you had to risk to get it, how long you sat underwater on the way, or whether the figure would survive being measured again. Here are the seven we ship, plus the two we retired in July 2026, so the question can be answered with numbers rather than adjectives.

Strategy	Profit factor	Sample	Status
Gap Fade	2.82 in pips	1,572 gaps	Opt-in
Reversal Capture	~2.3 out-of-sample	3.6 years, few trades	Opt-in
Mean Reversion	1.73	2,727 positions	The flagship
Aggressive Survivable	1.66	4,007 positions	Opt-in · Pro and above
Trend Follower	1.54	1,280 positions	Core
Swing Rider	1.27	1,343 positions	Opt-in · Pro and above
Momentum Rejoin	1.23 full span, 1.02 in 2026	675 trades	Retired 29 Jul 2026
Stretch Fade	1.14	8,307 trades	Demo only
News Momentum	1.05	best variant	Retired 29 Jul 2026

Measured Feb 2023 – Aug 2026 at shipped settings, except Reversal Capture (3.6 years to 2022) and Gap Fade (1,572 weekend gaps). Figures corrected 28 July 2026 — see below. Trend Follower and Aggressive were re-measured on 7 August after USDJPY was dropped from Trend’s basket on 31 July; both rose, because the pair had been losing money.

We published four of these wrong

An earlier version of this article gave the flagship as 1.11 and Trend Follower as 1.00, and drew a tidy moral from it: that a number you measure properly is almost always lower than the one you were quoting before.

That moral was wrong, and so were the numbers. The tool that produced them worked from a hand-copied list of settings which had drifted out of step with the engine, and four results ended up attached to the wrong strategies. The figure we published as the flagship’s belonged to a configuration the engine has never run. The one we published as Trend Follower’s came from a setup that takes the opposite side of the same signal — on the strength of which we announced the strategy was breakeven and switched it off for everyone.

Measured against what actually ships, the flagship is 1.73 and Trend Follower is 1.54 (it read 1.44 on the 11-pair basket it ran until 31 July). All four errors ran in the same direction: they understated the product.

The direction of a measurement error is not something you get to assume. Ours flattered our humility rather than our results, which made it far harder to spot.

That last point is the one worth taking away. A figure that makes you look bad feels like evidence of rigour, so nobody interrogates it. We had built an entire public argument on top of numbers that were never checked against the thing they claimed to describe — and the checking is the only part that was ever load-bearing. The measuring tool now reads the engine’s settings directly and refuses to emit a figure if it cannot; a second check compares those settings against a fingerprint taken when the current figures were measured, and has to pass before any number reaches a page like this one.

What follows is the part of the original article that survives contact with the corrected data — because none of it ever depended on our own numbers being low.

Higher is not automatically better

The highest number in the table belongs to Gap Fade at 2.82, and it is also the strategy with the risk profile most likely to hurt you.

Gap Fade trades the Sunday reopen, fading the weekend gap back toward Friday’s close. It wins 81.8% of the time. But gaps appear on many pairs at the same reopen — typically seven at once, sometimes sixteen — so at 1% risk per trade a bad weekend puts far more than 1% at stake. The worst weekend in the sample cost 6.1% of equity.

A profit factor is computed across trades. It has no opinion about how many of those trades were open simultaneously, which is exactly the dimension that decides whether a drawdown is survivable.

And check the units

Gap Fade’s 2.82 is measured in pips. The same result expressed in risk units comes out around 3.5. Both are real; they are different scales, and quoting the flattering one without saying which is how a lot of marketing gets made. We publish the pip figure and explain the other.

What the number needs before it means anything

A sample worth the name

Stretch Fade’s 1.14 comes from 8,307 trades. Reversal Capture’s ~2.3 comes from a strategy that deliberately trades rarely — few trades, small edge on each. Both figures are honest; they are not equally certain. A profit factor over forty trades is a rumour.

A confidence range, not a point

Stretch Fade is the only strategy here we can put a range around: resampling whole trading days puts its 90% confidence interval at 1.07–1.19. That interval sits entirely above 1.0, which is a stronger statement than the point estimate. It is also narrower than most people would guess, because it took 8,307 trades to earn it.

Consistency across windows

Momentum Rejoin is the cautionary one, and it is the reason this section exists. It measured 1.28 on its training window and 1.36 on validation. Against random entries held for the same duration it scored p = 0.102 — then p = 0.158 a month later, moving away from significance as data arrived. It was retired on 29 July 2026.

Two good windows and one bad one is not a 1.3 strategy; it is an unproven one. A significance test that drifts the wrong way on fresh data is the clearest signal you will get that a result was luck — and the right response is to stop, not to keep waiting for the number to come good.

A year-by-year breakdown

Three of the four trend-and-reversion strategies were positive in every year tested: Mean Reversion at 1.77, 1.47, 2.16 and 1.47; Aggressive Survivable at 1.62, 1.39, 2.02 and 1.62; Trend Follower at 1.42, 1.23, 1.79 and 2.18. Swing Rider is the exception, reading 1.39, 0.92, 1.39 and 1.42 — a losing year sitting inside a positive overall figure. You live through years, not averages.

So what should you expect?

• 1.2 to 1.8 is a realistic band for a diversified, out-of-sample-validated strategy trading liquid FX at retail costs. Five of the seven we ship sit there. We would have written “1.0 to 1.3” a week ago, on the strength of figures that turned out to be measuring the wrong thing — which is its own lesson about how confidently these bands get quoted.
• Above 2 usually means a narrow, event-driven edge — Gap Fade only exists at the Sunday reopen; Reversal Capture only trades a three-hour window. Both are real. Neither runs all day, and both carry concentration risk the ratio does not show.
• Above 3, sustained, across a large sample, with a stated drawdown — treat with suspicion. Ask what the sample was, what the units are, and whether it was measured on data the strategy had already seen.
• Around 1.05 is a real result too — and the honest response to one is to stop shipping it. We carried a 1.05 strategy for weeks behind a “do not fund it” label before accepting that the label was not an answer. It was retired on 29 July, along with a second that had never reached significance.

The most useful thing a profit factor does is sit beside a drawdown, a sample size and a per-year breakdown. On its own it is one number describing a distribution it cannot possibly summarise — which is the same criticism we made of win rate, for the same reason.

See all seven, in full