Fixed Fractional vs Fixed Lot: Which Sizing Really Compounds?

You and a friend take the same trades for a year. Same entries. Same exits. Same broker. At the end, one account has doubled and the other is flat.

This is a plain-English guide to position sizing.

The only difference was fixed fractional vs fixed lot.

One of you changed size every trade. The other didn't. The market didn't care. The maths did.

Position sizing: Why this argument even exists

Most retail traders obsess over entries and win rates. Position sizing gets left to whatever the platform default happens to be, or whatever some bloke on YouTube shouted loudest about.

"I always trade 0.10 lots" sounds sensible. Until the account changes.

The core question: do you size every trade off your current equity (fixed fractional), or do you hold the lot size constant and hope for the best (fixed lot)?

Fixed fractional vs fixed lot: the simple definitions

Let's get the labels done quickly so we can move to why they behave so differently.

Fixed fractional (or percentage risk) means you risk a set percentage of your current equity per trade. Example: 1% risk per trade, with a 50 pip stop on EURUSD, position sized so a full stop-out equals 1% of whatever the account is now.

Fixed lot means you trade the same lot size every time. Example: always 0.10 lots on EURUSD with a 50 pip stop, regardless of whether the account is £1,000 or £10,000.

Both feel sensible when you start. Only one respects compounding and drawdown arithmetic.

What actually compounds (and what doesn't)

Compounding isn't mystical. It's just applying returns to a changing base.

With fixed fractional sizing, if you make +2% on trade one, trade two risks 1% of a slightly larger number. If you then lose 1%, you lose 1% of that slightly larger number, not the original.

The sequence matters, but the mechanism is simple: every gain or loss sets the new stake for the next bet.

With fixed lot sizing, you disconnect stake from stake size. The pip value stays the same while the account wanders off somewhere else.

That feels like stability. It's not.

One table that explains the whole thing

Take a very basic hypothetical system:

• Win rate: 50%
• Average win: +2R
• Average loss: -1R
• R is the initial 1% risk per trade for the fixed fractional example

We'll run 10 trades from a £10,000 start.

Scenario A – Fixed fractional at 1% risk per trade

After 10 trades you've made roughly +5%. Not exciting. But every bet respected the same percentage risk, so the profile scales.

Scenario B – Fixed lot sized from the start

Day one, 1% risk equals £100. Let's say that, for the chosen stop distance, that means 0.20 lots on your pair of choice. You then keep 0.20 lots for all 10 trades, ignoring equity.

That means every full loss is -£100. Every average win is +£200.

The equity result is almost the same here. About +5%.

So what's the problem?

Where fixed lot quietly breaks your risk

The problem appears when you stretch that table out to years, not 10 trades.

Fixed lot risk as a percentage floats. After a deep drawdown, your fixed lot now risks more of a smaller account. After a big winning run, the same fixed lot becomes a tiny scratch relative to the new equity.

You are unintentionally doing the opposite of what you should want: betting bigger when you're weak and smaller when you're strong.

Take the same example but fast-forward it:

• If equity halves to £5,000 and you keep trading that 0.20 lot, your £100 risk is now 2%.
• If equity grows to £20,000 and you keep 0.20 lots, your £100 risk is now 0.5%.

The system's statistical edge hasn't changed. Your implementation of it has.

Fixed fractional keeps the risk per trade in line with your actual state of health.

Why fixed fractional sizing is the only one that truly compounds

When people say a strategy has, say, 20% expected annual return, they are almost always assuming fixed fractional position sizing in the background.

They're assuming you always risk 1% or 2% of whatever the account is at that time, and that the return is expressed on that rolling base.

If you run the same win/loss path but freeze position size, you no longer have percentage returns compounding on themselves. You just have a series of fixed cash gains and losses attached to a drifting account value.

Mathematically, fixed fractional turns a sequence of per-trade percentage returns into a geometric growth process. Each trade multiplies equity by a factor: 1 + r.

So ten trades with returns [ +2%, -1%, +2%, -1%, ... ] becomes:

Equity = Start × (1 + 0.02) × (1 - 0.01) × (1 + 0.02) ...

That's geometric compounding. The order matters, but every trade updates the base.

Fixed lot breaks that link. The percentage effect on equity is no longer stable per R, which makes long-term performance stats misleading if you backtested on one and trade with the other.

Why traders cling to fixed lot anyway

There are three main reasons people defend fixed lot sizing.

First, simplicity: "I just click 0.10 lots and move on". Second, emotional anchoring: they remember what 0.10 "feels like" on their account. Third, minimum lot constraints: especially on instruments like gold (XAUUSD) where the smallest position, say 0.01 lots, can already be quite a big swing for a £500 account.

On small accounts the last point bites hard.

If 0.01 lots on XAUUSD with your stop size implies 5% risk on a £500 account, you can't run clean fixed fractional sizing at 1% without changing the system or enlarging the account.

So people settle for a fixed lot they can live with and call it a day.

That's understandable. It's still dangerous if you then talk about expectancy, risk of ruin and drawdown as if you were running a percentage-based model.

Drawdown behaviour: the part everyone feels in their stomach

If you've read our piece on how much to risk per trade without blowing up, you'll know drawdown depth and duration are more important than the headline return curve.

That article assumes fixed fractional sizing by default. Most risk of ruin maths does.

Why? Because it ties maximum loss on each trade to your remaining capital. That keeps theoretical ruin probabilities stable.

With fixed lot, risk of ruin is a moving target.

Imagine a nasty losing streak: 10 losses in a row from £10,000 with fixed fractional 1% risk.

• Drawdown: roughly 9.56% (because each 1% is on a slightly smaller base).
• You're annoyed, but not finished.

Same streak with fixed lot, risking £100 every time, is a straight -£1,000, i.e. -10%.

Not much difference at that level. But stretch it.

At 40 consecutive 1% fractional losses (hypothetical, but we're making the point), the account is down about 33%.

At 40 fixed £100 losses, you're down 40%. The gap widens because fixed fractional reduces the stake with each loss, while fixed lot keeps kicking the same hole in a smaller hull.

Psychologically, both feel awful. Mathematically, one is trying to stop the bleeding, the other is still swinging.

Expectancy and sizing: they have to be married

When we talk about expectancy (the average gain or loss per trade) we normally express it in R-multiples or percent of equity.

Read that again: "percent of equity".

If your backtest says the system has a 0.3R expectancy per trade at 1% risk, that assumes 1% is 1% every time.

Switch to fixed lot in live trading and your actual per-trade expectancy in percentage terms is now a function of where you are in the equity curve. Early trades have higher percentage impact; later trades, if you've grown the account, barely move the needle.

The system edge hasn't changed. Your bet sizing edge has.

We went deep on win rate vs risk–reward in another piece, here: Win Rate vs Risk-Reward: Why Your 80% System Still Bleeds.

The summary in this context: a beautiful expectancy number won't save you if your position sizing model is fighting it.

Automation: what it fixes and what it doesn't

Automation is good at one thing in this space: actually doing the percentages without you bottling it.

Fixed fractional position sizing requires recalculating trade size off live equity every time. A human can do that with a calculator for a while. Then Friday afternoon happens.

A system will apply your 1% rule on every trade, including the ugly ones after five losses when you secretly want to halve size because it "doesn't feel right".

That discipline is the main advantage.

What automation does not fix is the underlying choice of model.

If you wire up a bot to fire 0.10 lots forever, all you've done is made fixed lot faster. You've automated the wrong idea.

Equally, if your broker or instrument minimum size makes true fixed fractional sizing impossible for your capital level, automation can't change that either. The constraint is structural.

On something like XAUUSD, where 0.01 lots already implies chunky swings on small balances, you can absolutely automate the execution. You cannot pretend you're on a smooth 1% risk per trade model when you're effectively swinging 3–5% per stop.

The numbers still count.

Applying fixed fractional sizing in the real world

So how do you move from theory to a position sizing strategy that behaves like the maths you backtested?

Step one is boring: decide your maximum percentage risk per trade and per day. If you haven't, read How Much To Risk Per Trade Without Blowing Up and sort that out first.

Step two: translate that into a mechanical rule.

• Example rule: "Risk 0.5% of current equity per trade, with stop-loss always defined in advance; position size = (0.5% of equity) / (stop size in pips × pip value)."
• Apply that to every trade, regardless of how you feel.

Step three: recognise where platform and instrument constraints force compromises.

• On FX majors, with micro-lots available, you can usually get very close to the exact fraction you chose.
• On gold or indices, minimum lot size may force you above your ideal percentage risk on small accounts. That isn't solved by clever maths. It's solved by either running a larger account, wider stops with smaller lots, or not trading those products at that size.

Fixed fractional is a principle. The implementation will always be a discretised approximation.

When fixed lot can be defensible

There is one context where fixed lot sizing isn't insane.

If you explicitly care about absolute cash outcome, not percentage return, and the account size is so large that your fixed lot is a tiny fraction of equity at all times, fixed lot is just "I trade small forever".

Someone with £5m deciding they are fine with £500 of risk per trade in perpetuity isn't "optimising" anything. They're just choosing a lifestyle.

But most people reading about fixed fractional vs fixed lot are not in that camp.

They're trying to grow an account sensibly without blowing it up. For that problem, you want your risk model to detect when the account shrinks and respond accordingly.

Fixed fractional does that automatically. Fixed lot only notices when you blow the account and have to deposit again.

Testing, decay, and why the sizing method must match

One last point that bites people moving into systematic trading.

When you test a strategy, you must test it with the same position sizing model you intend to use live.

If your backtest used fixed fractional sizing and you go live on fixed lot "until the account grows a bit", you have changed the distribution.

Metrics like maximum drawdown, Sharpe ratio, profit factor, and even system decay behaviour are all sensitive to how size changes over time.

We talked about profit factor here: What Is Profit Factor (And When Is It Actually Good?).

That number is blind to bet size if you compute it in R-multiples. But your actual cash and percentage drawdowns in the live account are very much not blind to it.

Likewise, when you assess whether a strategy is decaying or just in a normal drawdown, as in What Is System Decay In Trading (And When To Retire One), you need a stable risk per trade to judge fairly.

Changing sizing model half-way through is the fastest way to confuse yourself about whether the system or the staking is broken.

The boring answer to the exciting question

So which sizing actually compounds?

Fixed fractional, every time, as long as you can implement it sensibly with your account size and instrument minimums.

Fixed lot has its uses as a blunt risk cap when you're trading tiny relative to equity, or when platform constraints make fractional ideals impossible. But if you're serious about expectancy, drawdown, and long-term growth, it shouldn't be your default.

The adult move is simple.

Decide your fraction. Automate it if you can. Then leave it alone.

Trading involves real risk. You can lose some or all of your capital, and no sizing method removes that — it only shapes how fast the damage or the growth arrives.

Let the maths, not your mood, size the trades

If you want to see fixed fractional sizing run without flinching after three losses, watching a systematic account helps.

Seeing multiple automated strategies across FX, gold and indices trade in live time — winners and losers, tests and live systems next to each other — is a good antidote to "I'll just trade 0.10 lots and hope".

And if you care about whether a strategy quietly sneaks martingale or grid into the mix, you should be able to see that too, not find out after the fifth doubling.

Start your free 14-day ArcisTrade demo →

P.S. Watch the position sizing column, not just the equity curve. The edge starts there.