What Is Profit Factor (And When Is It Actually Good?)

Profit factor looks like the clean edge number you've always wanted. The catch: it hides sample size, volatility and tail risk. Here's how to read it like a grown‑up.

Will Simpson · 29 Sept 2026 · 9 min read

You’ve finally found it. A backtest with a profit factor of 3.2 and a pretty equity curve. You stare at it like it’s a winning lottery ticket.

Then you notice it’s based on 37 trades.

The question on the screen is simple: what is profit factor, really, and is that number actually good or just statistical fan fiction?

Profit factor: the one-number edge that lies by omission

Profit factor is simple: total gross profit divided by total gross loss over a set of trades.

If you made £10,000 on your winners and lost £5,000 on your losers, your profit factor is 10,000 / 5,000 = 2.0.

Above 1.0, you made more than you lost. Below 1.0, the system lost money. The rest of the conversation is about how much you trust that number.

And that’s where most people stop thinking and start dreaming.

What is profit factor actually measuring?

Under the bonnet, profit factor is just an expectancy ratio stripped of trade count.

Formalising it helps:

Let GP = sum of all profitable trades (gross profit)
Let GL = absolute sum of all losing trades (gross loss)
Profit factor (PF) = GP / GL

So a PF of 1.5 means “over this sample, for every £1 lost, the system made £1.50”.

Sounds clean. It isn’t.

Because that sentence is missing three words: “in this sample only”.

Why the win rate fools you, and profit factor only half-fixes it

A lot of traders cling to win rate. 70%, 80%, 90%. Feels safe.

Then they realise you can have a 30% win rate and still make money if the winners are big enough. So they move to profit factor as the “grown-up” metric.

If you haven’t read it already, the proper way to tie this together is in expectancy: /blog/what-is-trading-expectancy

But here’s the key link:

Expectancy is average per trade
Profit factor is total winners vs total losers, ignoring count

Both measure edge. Expectancy tells you how much. Profit factor just tells you “winners bigger than losers, yes or no, and by how much, in this window”.

It’s the trading equivalent of asking how good a football team is by adding all their goals and all the goals against, then dividing. Useful. Incomplete.

Profit factor and sample size: your first serious problem

Here’s where most “what is profit factor” explanations quietly skip the uncomfortable bit: sample size.

Two systems:

System	Trades	Gross Profit	Gross Loss	Profit Factor
A	60	£3,000	£1,000	3.0
B	2,000	£80,000	£60,000	1.33

Most people point at System A and start calculating what car they’ll buy.

Which is odd, because from a risk perspective, B is usually far more interesting.

Why?

60 trades can be fluke, overfitting, or a lucky regime
2,000 trades span more conditions, more pain, more reality
In the real world, keeping a 3.0 PF over thousands of trades is extremely rare for non-extreme position sizing

With a small sample, any single big winner or loser can swing the profit factor wildly. One trend trade you accidentally held through a central bank meeting and you’re suddenly a genius.

Until you aren’t.

If you want numbers that have some statistical dignity, you also need to ask: how many trades do I need before I treat this profit factor as more than a suggestion? That’s a separate topic, covered properly here: /blog/how-many-trades-do-you-need-to-test-a-strategy

Good profit factor vs durable profit factor

So what counts as a “good” profit factor?

The uncomfortable but honest answer: it depends where it came from.

As a guide — hypothetical, not a promise:

PF 1.0 or less: losing or breakeven edge. Fix the system, not the position sizing.
PF 1.1–1.3: thin but possibly real edge. Often what robust, diversified, professionally-run strategies look like before leverage.
PF 1.3–1.8: decent edge if it holds over hundreds or thousands of trades.
PF > 2.0: exciting on paper; often a sign of concentration, curve-fitting, or very specific market conditions.

Notice the missing column: “recommended car colour for your new supercar”.

A “good” profit factor is one that survives more data, more volatility, and more boredom than you think you can stand.

Where tail risk hides inside a shiny profit factor

Profit factor also has a blind spot: tail risk.

Imagine a system that sells volatility. It collects lots of small, neat wins and occasionally explodes.

A hypothetical example:

900 winning trades, average win £100 → £90,000 gross profit
90 losing trades, average loss £600 → £54,000 gross loss
Profit factor = 90,000 / 54,000 ≈ 1.67

1.67 looks respectable. You’d see that and relax.

Then you realise those 90 losers are not evenly spaced; they cluster in three or four panics where the system loses half its equity each time.

Profit factor does not tell you:

Maximum drawdown
How losses are distributed in time
Whether one bad day can wipe out a year

It is perfectly possible to have a PF above 1.5 and still blow up. That’s the nature of tail events and leverage.

If you care about staying in the game (you should), drawdown and streaks matter just as much as the headline ratio: /blog/normal-maximum-drawdown-trading-system and /blog/how-long-a-losing-streak-should-you-actually-expect

Profit factor and drawdown: the unromantic pairing

Profit factor tells you “edge per unit of loss”. Drawdown tells you “pain per unit of edge”.

Put them together and you start to see whether a system can actually be traded by a human with a pulse.

Two hypothetical systems with the same profit factor:

System C: PF 1.4, max drawdown 8%, smooth-ish equity
System D: PF 1.4, max drawdown 45%, violent swings

Same PF, totally different experience.

System D will empty more accounts, because most traders will nuke it at -30% and never reach the “on paper” long-term outcome.

So a better question than “is this profit factor good?” is “is this profit factor good relative to its drawdown and my sizing?”

Position sizing makes profit factor feel very different

Profit factor is calculated in cash terms, so it’s independent of position size in theory. Double the size, you double both GP and GL, PF stays the same.

In practice, your sizing choice determines whether that nice clean 1.4 PF feels like a balanced system or a daily heart attack.

If you size aggressively — think high fixed fraction or aggressive Kelly-style bets — the same profit factor produces higher drawdown and nastier losing streaks. That’s covered in more depth here: /blog/what-is-the-kelly-criterion-and-why-full-kelly-ruins-you and /blog/fixed-fractional-vs-fixed-lot-sizing-that-compounds

On instruments like gold (XAUUSD), you also hit a practical problem: minimum lot sizes.

With a 0.01 lot minimum on some brokers, even a small account can see big equity swings per trade. Profit factor on the statement might be fine, but emotionally you’re trading a rollercoaster.

Same PF. Different heart rate monitor.

How automation changes profit factor (and how it doesn’t)

Automation tempts people into thinking profit factor is somehow more “real”. As if a robot trade makes a 1.3 PF more reliable than a discretionary one.

It doesn’t. The maths is the same. The trades are just executed by code instead of by hand.

What automation does help with:

Actually achieving the recorded profit factor, because you don’t skip trades or move stops
Running multiple systems across FX, indices and gold so you’re not hostage to one setup
Tracking gross profit and gross loss cleanly, without “I’ll just ignore that news spike” editing

What automation does not fix:

Overfitting that produced a fantasy PF in backtest
Tail risk — the code will calmly keep trading while your equity is cut in half
The emotional impact of drawdown; watching an automated system take five losers in a row feels just as bad

Automation is an execution tool, not a magic amulet that makes a 2.5 PF sustainable.

If anything, it makes bad systems blow up faster.

What is profit factor without context? Dangerous

So when you see a backtest or an account with a headline profit factor, your checklist should be boringly methodical.

Ask these, every time:

How many trades? A PF of 1.4 over 2,000 trades impresses me more than 3.0 over 60.
What’s the max drawdown? High PF with deep drawdowns is more casino than business.
How are losses clustered? Do the big hits come in regimes you can survive?
What instruments and timeframes? A 1.3 PF on a diversified FX basket can be much saner than 2.0 from selling options into quiet markets.
What position sizing? If they needed 10x leverage to get those numbers, your lived experience will be different.
Is there any obvious overfitting? Ten optimised parameters, perfect equity, and PF 4.0 should scream “curve-fitted”. See /blog/what-is-overfitting-in-trading-and-how-to-spot-it

On their own, profit factor and a smooth equity curve are just a nicely formatted story.

Your job is to find out if it’s non-fiction.

So what is a good profit factor for a trading system?

Let’s answer the question people really type into search: “what is a good profit factor?”

Stripped of marketing and fantasies, a “good” profit factor for a mature, diversified, reasonably sized system is often in the 1.2–1.6 range across a decent sample.

Why that low?

Markets are competitive; clean, high edges get arbitraged away
Diversification across systems and markets naturally dilutes the headline PF
Conservative sizing keeps you alive long enough to let the edge matter

If someone waves a PF of 3–5 with tiny drawdown at you, it might be real. It’s just more likely to be curve-fitting, lucky data, or some version of martingale or grid hiding under the surface.

The fact that some bloke on YouTube can always find such a system in hindsight doesn’t mean you can trade it in real time.

How to actually use profit factor in your system development

Profit factor becomes genuinely useful when you treat it as one voice in a committee.

Here’s a concrete way to use it when you’re building or evaluating systems:

Stage 1: Sanity filter – exclude systems with PF < 1.1 over any realistic sample. The edge is too thin or non-existent.
Stage 2: Robustness check – for anything with PF > 1.5, be extra suspicious. Stress test across years, markets, parameter sets.
Stage 3: Risk alignment – pair PF with maximum drawdown, volatility of returns, and your own risk tolerance.
Stage 4: Portfolio view – several systems with PF 1.2–1.4 that are uncorrelated can be far healthier than one superstar with PF 2.5.

And then, crucially, you automate the execution so the live trades actually resemble the test.

Staring at a lovely profit factor on a backtest and then freelancing every live entry is how traders turn maths into fiction.

Risk first, always

Everything above boils down to one principle: treat profit factor as an edge summary, not a promise.

Ask what had to be true to get that number. Ask what would have to happen for it to fall apart.

And never forget the dull line that should be written on every account opening form: trading involves risk; you can lose money, including your initial capital.

If you want to see how different systems behave in the wild – winners, losers, drawdowns and all – watching them run in real time is often more educational than another spreadsheet.

Start your free 14-day ArcisTrade demo →

Watch multiple automated strategies trade FX, indices and gold into your own broker demo. No card, no martingale, no grid. Just real trades, including the ugly ones.

P.S. Don’t just look at the profit factor on the screen. Watch how it behaves when the market has a tantrum.

Common questions

How do you calculate profit factor in trading?

Profit factor is calculated as total gross profit divided by total gross loss over a set of trades. You add up all winning trades to get gross profit, add up the absolute value of all losing trades to get gross loss, then divide gross profit by gross loss. Values above 1.0 indicate net profit for that sample; below 1.0 indicate net loss.

What is a good profit factor for a trading system?

A realistic, robust profit factor for a mature system is often between 1.2 and 1.6 over hundreds or thousands of trades. Much higher values can occur but are often linked to small samples, curve-fitting, or concentrated risk. Always judge profit factor together with sample size, maximum drawdown and position sizing, not in isolation.

Is a higher profit factor always better?

Not necessarily. A very high profit factor over a small number of trades can be the result of luck or overfitting. Some high profit factor systems also hide significant tail risk, where rare but large losses can cause deep drawdowns. A slightly lower but stable profit factor over a large, varied sample is usually more dependable.

Can two systems have the same profit factor but different risk?

Yes. Profit factor only compares total profit to total loss; it ignores how those losses are distributed. Two systems with the same profit factor can have very different maximum drawdowns, volatility and streak behaviour. One might have shallow, frequent losses; the other rare but severe equity crashes. That’s why you must pair profit factor with drawdown and streak analysis.