Fixed Fractional vs Fixed Lot: The Sizing That Actually Compounds

You start with £5,000 and a neat little system on gold.

This is a plain-English guide to fixed fractional.

Risk feels tiny. 0.01 lots a trade. "It's nothing."

Fast forward a year, the balance is £500. Same 0.01 lots. Now each loss is 2% of the account. Same trade. Very different risk. That's the fixed fractional vs fixed lot problem in one chart you probably never drew.

Why fixed fractional vs fixed lot even matters

Position sizing is the boring bit people skip to get to entries.

Which is odd, as it's the bit that determines whether your equity curve bends up, drifts sideways, or implodes.

Fixed lot sizing says: "I trade 0.1 lots, always." Fixed fractional sizing says: "I trade 1% of whatever equity I have, every time." Same strategy, same entries, same exits. Very different journey.

When someone shows you backtests without specifying sizing, assume it was whatever made the chart look the nicest.

What fixed lot sizing really does to your account

Fixed lot sounds sensible.

It feels controlled. You can visualise the risk: "I lose about £50 when it goes wrong."

The catch is simple: that £50 doesn't stay £50 relative to your account. It drifts. Quietly. On day one, maybe it's 1% of equity. After a good run, it's 0.4%. After a bad run, it's suddenly 3–4%. Same lot size, different percentage of what you have left.

So fixed lot is really "random percentage risk".

Take a simplified example.

You start with £10,000, trade a system on EURUSD with a 50-pip stop, and each pip is £1 at 0.2 lots.

Each loss is £50. At the start, that's 0.5% of equity. If you draw down to £5,000 and don't adjust, that same loss is now 1%. If the account falls to £2,500, it's 2%. The worse you're doing, the more you lever up relative to your shrinking base. Without touching a setting.

It's martingale's boring cousin. No doubling. Just a slow, automatic overweighting of your worst period.

Fixed fractional: the boring thing that actually compounds

Fixed fractional says: pick a number. 0.5% per trade. 1%. 2% if you enjoy pain.

Then size every trade so the most you can lose (to your stop) is that percentage of current equity.

Mathematically, if your equity is E, your risk fraction is f, and your stop distance is S (in money), your position size is simply f × E / S. Equity goes up, position size goes up. Equity goes down, position size goes down.

You don't need a spreadsheet. Just that relationship.

That one decision does three useful things at once:

• It compounds winners (larger pound risk on higher equity).
• It defends in drawdowns (smaller pound risk as equity falls).
• It keeps risk of ruin anchored to reality, not nostalgia for what the balance used to be.

Compounding isn't mystical. It's just "same percentage on a bigger base" repeated enough times without blowing up. If you haven't read about expectancy, start there: What Is Trading Expectancy (And Why Your Win Rate Is Lying To You)?

Let’s put some numbers on it

Assume a hypothetical strategy with this profile:

• Win rate: 50%
• Average win: +2R
• Average loss: -1R
• No position sizing changes other than rules we're testing

So expectancy per trade is +0.5R. The edge exists. The only variable is how you size.

Scenario 1: fixed lot returns, no compounding

Say you trade £100 risk per loss. Fixed. Always.

Over 200 trades, expected profit in R is 200 × 0.5R = 100R.

In money that's 100 × £100 = £10,000. So if you started with £10,000, you'd expect to end around £20,000 if your system behaves perfectly and the fixed risk stays sensible relative to your equity the whole way.

Now adjust for reality.

If drawdown hits early and brutal — say -40% at the low — your £100 fixed risk is now 1.7% of equity instead of 1% at the start.

Variance gets magnified. The path dependence bites. The same edge now has a higher chance of catastrophic ruin because your risk per trade quietly inflated while you were busy being annoyed at the last loss.

Same expected total gain in absolute pounds if you survive – but with a higher chance you don't reach the end of the sample.

Scenario 2: fixed fractional, actual compounding

Now risk 1% of equity per trade instead.

Start with £10,000. First loss? -£100. Second trade wins +2R? Now equity is £10,100. Next 1% risk is £101. It's microscopic at first. Looks like nothing.

But after a long sequence, the gap between "fixed £100" and "1% of whatever I have now" is large. Especially if you survive the early drawdowns with smaller pound risk when you're low.

That is compounding in practice: the bet size follows the equity curve, not the memory of a nice round number lot size you picked last spring.

But fixed fractional makes my drawdowns look worse…

This is where people get spooked.

Look at an equity curve with 1% fixed fractional and the percentage drawdown looks the same whether you started with £5,000 or £50,000. 20% is 20%.

Look at fixed-lot, and on a long run the same initial drawdown can look smaller in percentage terms, because as equity grows, that fixed £100 or 0.1 lot becomes a smaller and smaller slice of the pie.

That's not a safety feature.

It's just another way of saying: the system is under-sized now compared with what it could responsibly run at.

You haven't reduced risk; you've reduced the rate at which you apply the edge once it proves itself. You kept the wage the same while your skill went up.

If you've done the work on risk of ruin and sizing — see the Kelly piece here: What Is The Kelly Criterion (And Why Full Kelly Ruins You)? — you're usually better off keeping risk per trade constant as a percentage.

Otherwise you're just donating edge back to the broker because you're scared of higher pound numbers.

How the two methods change risk of ruin

Risk of ruin is "what's the chance I blow this account or hit a level where I'm effectively done".

For a system with positive expectancy, fixed fractional sizing has a very particular property: if you never raise the fraction, the theoretical probability of literal ruin (equity hitting zero) can be driven arbitrarily low.

Because as equity falls, pound risk falls; the system asymptotically limps instead of faceplanting.

Fixed lot doesn't do this.

If the lot size is large relative to equity, a long enough bad run can easily take you to margin call levels. The percentage risk per trade rises as equity falls, then at some point you hit a spiral: fewer bullets left, but each one is large relative to the account.

That's why sizing is in the same family as stop placement, not an afterthought. It's a risk-control lever, not a cosmetic one.

If you want more on streaks and why they feel worse than the maths suggests, read How Many Losing Trades In A Row Is Normal?

Short version: both methods face losing streaks. Fixed fractional shrinks the bet when they arrive. Fixed lot quietly raises the percentage bet size as the streak deepens — unless you manually intervene.

The automation trap: fixed lot because it's easier to code

Now bring automation into it.

Plenty of retail bots simply ask: "What lot size do you want?" Then they blast that same size forever, come rain or margin call.

Fixed-fractional sizing means either calculating the position size dynamically from balance/equity and the stop distance, or letting the platform handle % balance risk for you.

One of those is a little more effort in code.

So a lot of bots default to fixed lots because it was quicker for the developer, not because it was better for your risk.

Layer in volatile products — gold (XAUUSD), indices, high-beta FX pairs — and the problem magnifies. On XAUUSD, for example, a 0.01 lot minimum means a small account can be taking massively chunky risk in percentage terms without the trader quite realising it.

"It’s only 0.01" on a £500 account with a $15 stop can easily be north of 2–3% of equity. That’s before slippage does its usual impression of a helpful stranger.

Automation executes your rule. It doesn't fix a bad one.

When fixed lot can be justified (and when it's just laziness)

There are legitimate uses for fixed lots.

• Very small accounts hitting broker minimum lot sizes.
• Manual intraday scalping where calculating proper % risk impairs execution.
• A deliberate decision to under-size a system far below its optimal risk because it's only one small sleeve in a larger portfolio.

In those cases, you're consciously accepting that risk will drift with equity.

The problem isn't fixed lot itself; it's blind fixed lot.

Picking 0.1 lots on day one of a gold system and never revisiting it while the account swings from £3,000 to £900 is not a plan. It's a vibe.

If you insist on fixed lot, at least map what percentage of equity that lot represents at different balance levels and decide in advance when you'll step it down or up.

Write the rule once, instead of negotiating with yourself mid-drawdown.

Why fixed fractional feels worse (and why that’s good)

There’s a psychological tax with fixed fractional.

When the account's down 20%, 1% per trade is a smaller pound amount than it was at the peak. It feels like you're "not making it back fast enough".

Some traders respond by bumping risk to 2–3% "just until I recover". This usually ends exactly how you'd expect.

Fixed lot is sneakier.

Because the pound amount doesn't change, you feel stable. £100 risk is £100 risk. The brain likes that. But the actual risk — as a percentage of your shrinking pot — is going up, not sideways.

This is why position sizing should be set with a calculator and your eyes off the equity curve, not on it. The maths needs to win that argument, not the part of your brain that wants to be square with the market by Friday.

If you haven't yet thought about how overfitting plays with sizing — i.e. sizing up on something that only ever worked in-sample — this is where people detonate accounts. See What Is Overfitting In Trading (And How To Spot It)?

Portfolio reality: different systems, different fractions

In practice, you rarely have one system.

You might have a trend follower on FX, a mean-reversion model on indices, and something carefully designed for gold's habits.

Each has different volatility, trade frequency, and expected drawdown. A flat "0.1 lots each" makes even less sense in that setting than on a single system.

Fixed fractional lets you express risk in the only unit that matters: percentage of your total equity.

You can assign, say, 0.5% per trade to a high-frequency mean reverter, and 1% to a slower trend system that trades once a week. You can cap portfolio-level exposure by saying: "I never have more than 3% of equity at risk at once across all open trades."

Those rules are only coherent when position size is defined as a function of equity. Otherwise, one system is quietly running leverage 4x the other just because its stop is tighter or the instrument has a different pip value.

And you're left with that classic line: "I don't know why that one blew up; they were all at 0.05 lots."

So which sizing actually compounds?

If the question is literally: fixed fractional vs fixed lot — which method compounds? — the answer is straightforward.

Fixed fractional by definition compounds, because bet size grows with equity on the same percentage risk.

Fixed lot does not compound in the same way. It gives you linear-ish pound returns over time if you survive, while the effective percentage risk per trade drifts.

There are edge cases where fixed lot is all you can do sensibly, usually because of broker constraints or unusually tiny accounts.

But if you’re running a systematic approach, aiming for long-term growth, and you actually care about risk of ruin, the default answer is boring and clear: fixed fractional.

Understand the edge. Understand the risk. Automate the execution. Don’t hard-code yesterday’s lot size and call it a plan.

And if any sizing scheme promises to recover losses by increasing lots into drawdown, you've probably just met martingale in a fake moustache.

Trading involves real risk. You can lose part or all of your capital, and no position sizing method removes that risk.

Start your free 14-day ArcisTrade demo →

Watch automated systems with fixed, transparent sizing across FX, indices, and gold. Winners and losers side by side, no martingale, no grid — just rules you can actually inspect before you trust them.