How Much To Risk Per Trade Without Blowing Up

You’ve probably typed "how much to risk per trade" into a search bar after a bad week.

Balance down. Confidence down. You start wondering whether 5% per trade was maybe, possibly, slightly ambitious.

The internet tells you everything from 0.25% to 10%.

Some of it from a bloke whose backtest started last Tuesday.

Everyone asks the wrong question first

The question everyone starts with is: "How fast can I grow this account?"

The question that actually matters is: "How much drawdown can I survive without quitting or blowing up?"

Risk per trade is not a growth lever first. It’s a survival setting.

Get that backwards and the maths will fix it for you. Violently.

How much to risk per trade: start from drawdown, not dreams

Forget percentages for a moment and think in pain.

Say your account is £10,000 and you swear you’ll tolerate a 30% drawdown.

That’s £3,000 of pain. On a chart it looks like a neat valley. In real time it feels like a personality test you’re failing.

The job of how much to risk per trade is to make the probable losing streaks fit inside that £3,000.

Not the worst theoretical losing streak the universe can generate.

The one your system is likely to throw at you in the next few years.

Why the win rate lies to you

Take a simple hypothetical system.

It wins 50% of trades. Average winner equals average loser. Expectancy is zero, but we’ll fix that in a minute.

Even with no edge, you can still learn the key bit: streaks.

The probability of N losses in a row with a 50% win rate is roughly (0.5)N.

So:

• 4 losses in a row ≈ 6.25% chance at any given point
• 6 losses ≈ 1.56%
• 8 losses ≈ 0.39%

"That’s small, I’ll be fine," you think.

Until you look at how many trades you actually take.

If you run a system for a few hundred trades, the maximum losing streak you’re likely to see is usually much worse than you thought the first time you eyeballed the stats.

There’s a whole piece on that here: /blog/how-long-a-losing-streak-should-you-actually-expect.

Short version: the streak eventually turns up. Often on a Monday.

The 1%, 2%, 5% per trade myth

You’ve heard the rules.

• "Pros risk 1–2% per trade"
• "Aggressive is 5%"
• "Go big or stay small forever" from someone selling a course

None of these are laws of nature.

Let’s add a real edge and see what actually happens.

Hypothetical system:

• Win rate: 45%
• Average winner: 2R
• Average loser: -1R

So expectancy per trade = 0.45×2 – 0.55×1 = 0.35R.

Good system on paper.

Now we decide how much one R is.

*Order-of-magnitude numbers for a few hundred trades with a 45% win rate. The precise figure varies, the point does not.

Same system. Same edge. Same trades.

The only thing we changed was how much you chose to hurt when it went cold.

At 5% risk per trade, one routine nasty streak takes you from £10,000 to roughly £4,000–£5,000 after some minor bounces.

At that point, "sticking to the plan" requires monk-level discipline or amnesia.

Risk per trade is a brake, not an accelerator

This is the bit people hate hearing.

Increasing risk per trade beyond a certain point does not increase your long-term growth. It throttles it.

The shape of the equity curve becomes so volatile that your odds of hitting an account-killing drawdown rise faster than your theoretical gain.

This is essentially the same problem we talked about with profit factor and expectations here: /blog/what-is-profit-factor-and-when-is-it-actually-good.

A system can be mathematically good and psychologically untradeable at the size you’re running.

Risk per trade is how you turn a good idea into something you can actually sit through.

Fixed fractional: the boring answer that usually wins

The cleanest position sizing method is fixed fractional.

You risk a fixed percentage of current equity on each trade, based on your stop distance.

Account grows, size creeps up. Account shrinks, size shrinks.

Example.

You decide 1% risk per trade on a £10,000 account. Stop is 50 pips. You set your lot size so that 50 pips = £100.

If you drop to £9,000, the number becomes £90. Same 1%, smaller nominal hit, smaller emotional damage.

It does three important things:

• Keeps you technically away from ruin (percentage of a smaller base is smaller).
• Automatically scales up when things are going well, without chasing.
• Stays simple enough that you can actually implement it with or without automation.

And no, it’s not exciting. That’s the feature.

How to pick your number like an adult

Let’s walk through a sane process to decide how much to risk per trade.

Four steps. No Lamborghinis.

1. Decide your maximum drawdown in pounds, not vibes

Pick a level where you can still trade rationally.

For many people that number is smaller than they say on forums.

Be honest: if your £10,000 account went to £7,000, would you calmly follow the next signal, or would you go system shopping on YouTube?

Write down a number between 10% and 40% that you could sit through.

Call it Dmax.

2. Estimate a realistic losing streak

You need some stats for this.

If you haven’t read it yet, the piece on /blog/how-many-trades-do-you-need-to-test-a-strategy-2 is worth a look.

Assuming you have a few hundred trades tested, look at:

• Win rate (p)
• Number of trades per year

There are formulas for expected maximum losing streak, but a rough and useful approach is:

• 50–60% win rate: expect 6–10 losses in a row at some point
• 40–50% win rate: expect 8–14 losses in a row
• <40% win rate: expect 12+ losses in a row

Call your chosen planning streak L.

3. Back out risk per trade from your drawdown limit

Now link the two together.

If you risk R% per trade, and you hit L straight losses, the "straight line" drawdown is roughly L × R.

Reality is noisier, but again: we’re planning, not pretending.

So you want: L × R ≤ Dmax.

Rearrange for R: R ≤ Dmax / L.

Example: Dmax = 25%, L = 10 → R ≤ 2.5%.

That gives you the hard stop.

Then you knock it down a bit for safety and psychology.

If the maths says 2.5% is the ceiling, 1%–1.5% is where a more conservative, long-term trader probably lands.

4. Adjust for correlations and open positions

The above assumes one position at a time.

In reality, many systems run multiple trades, often on correlated markets.

If you run three FX pairs that all move with USD, or multiple gold trades, your "total portfolio risk per trade" is higher than you think.

Rule of thumb.

• Set a per-trade risk (e.g. 0.5%–1%).
• Set a maximum portfolio risk at one time (e.g. 2%–3%).

Once your open risk hits the portfolio cap, you stop opening new positions, even if the system gives another signal.

Is that perfectly optimal?

No. It is survivable.

Why small accounts and gold make this harder

Now the annoying part.

Position sizing is easy to write down and harder to do cleanly in live markets, especially for small accounts.

Minimum lot sizes and instrument volatility complicate the "nice smooth percentage" you drew on a spreadsheet.

Gold (XAUUSD) is a regular culprit.

Minimum 0.01 lots on many brokers. For a small balance, that unit size can represent a large chunk of equity, especially with a 500–1,000 pip stop.

You think you’re running 1% risk per trade; you’re actually swinging 3–5% every time the gold system fires.

Two options if you want consistent risk management on a small account:

• Avoid instruments where the smallest contract size is still too big for your desired risk per trade.
• Deliberately lower your target risk per trade (say 0.25%–0.5%) so that rounding up to the minimum lot doesn’t blow your plan apart.

Either way, the constraint is real. The market doesn’t care that your spreadsheet promised 1.3478 micro-lots.

Automation: what it actually fixes, and what it doesn’t

Automation is excellent at doing what you told it to do.

That’s the blessing and the curse with risk per trade.

Good news first.

• It calculates position size correctly, every time, based on your rule.
• It doesn’t revenge-trade by quietly tripling size after a loss.
• It doesn’t "forget" to cut size after a nasty drawdown.

The spreadsheet logic actually makes it into live orders.

But.

Automation does not change the distribution of outcomes.

If you set 3% risk per trade into a system with a 12-loss worst-case streak, the machine will calmly walk you into a 30–40% drawdown while you watch.

It also doesn’t fix bad assumptions.

If your backtest overlooked slippage, or you only tested in neat conditions, the true risk per trade can be higher than you think once real fills arrive.

There’s a separate piece on that here: /blog/what-is-slippage-trading-and-how-it-eats-your-edge.

So automate the execution.

But do not outsource the responsibility for telling it what to risk per trade.

Why "optimal" sizing usually isn’t

At this point someone usually brings up Kelly sizing or some "optimal f" formula.

The maths is elegant. The equity curves rarely are.

Kelly-based risk per trade is aimed at maximising long-term geometric growth assuming:

• Your edge estimate is perfectly accurate.
• The distribution of returns doesn’t change.
• You are indifferent to giant drawdowns on the way.

All three assumptions are heroic.

In real trading:

• Your estimate of expectancy is noisy.
• Market regimes change; systems decay (/blog/what-is-system-decay-and-when-to-retire-a-strategy).
• You probably don’t enjoy 50% drawdowns, however "optimal" they are on a log scale.

A more sensible approach is "fractional Kelly" in spirit, without needing a PhD to operate it: take whatever clever formula spits out, look at the drawdown it implies, and then cut the risk several times until you no longer feel slightly ill.

So what’s a sane number?

Nobody sensible is going to give you one magic percentage that fits every system, every market, every human.

But there are ranges that tend to keep people alive.

• 0.25–0.5% per trade: cautious, good for volatile instruments (gold, indices), correlated portfolios, or newer systems.
• 0.5–1% per trade: common sweet spot for people running multiple trades with a tested edge and realistic expectations.
• 1–2% per trade: only if you have very strong, walk-forward tested confidence in the system, moderate trade frequency, and you accept larger swings.

Above 2% per trade, the burden of proof on your system quality and your own discipline grows quickly.

Most traders are not as robust as the backtest print-out on their desk.

Your actual job: choose something you can survive

So the next time you’re wondering how much to risk per trade, don’t start with the upside.

Start with the worst week you’re realistically going to see, in money terms, with your expected losing streak and your chosen percentage.

If that number makes you wince, your future self will probably bin the system before the edge plays out.

Survival first.

Growth is what happens if you don’t blow up.

Plain risk statement: Trading leveraged products carries a high level of risk to your capital; losses can exceed deposits and you should only trade with money you can afford to lose.

Start your free 14-day ArcisTrade demo →

Watch systematic strategies run in real time, with position sizing, drawdowns and losing streaks visible, not guessed.

No card, no promises of overnight profits, just live data to test your risk comfort against.

P.S. Pay more attention to how the equity curve falls than how it rises. That’s where your true risk per trade reveals itself.