Home › FX Trading School › Pro
Advanced Position Sizing: Fixed-Fractional and Beyond
Advanced position sizing is where trading stops being about finding good entries and starts being about surviving long enough for those entries to pay off. If you've completed the earlier risk modules in the PipTax FX Trading School — particularly the lesson on basic risk-per-trade and stop-loss placement — this module builds directly on that foundation and takes it into territory used by professional desks: fixed-fractional models, volatility scaling, and correlation-aware limits.
Why Fixed Position Sizes Fail Over Time
A flat lot size (say, always trading 1 standard lot regardless of account balance) feels simple, but it quietly punishes you exactly when you can least afford it. After a losing streak, that same lot size represents a larger percentage of your shrunken equity — you're risking more, relatively, right when confidence and capital are lowest. After a winning streak, the opposite happens: you're under-risking relative to your new equity, leaving compounding on the table.
Fixed-fractional sizing fixes this by recalculating position size before every trade based on a constant percentage of current equity:
- Losses shrink position size automatically — a 20-trade losing streak at 1% risk per trade leaves you with roughly 82% of starting equity, not a wiped account
- Wins grow position size automatically — equity compounds without manual adjustment
- The maths is simple: Position size = (Equity × Risk%) ÷ (Stop distance in pips × pip value)
The catch: this only works if you actually recalculate every time, which most retail traders skip when busy or emotional.
The Fixed-Fractional Formula in Practice
Here's the calculation broken into steps, using round numbers for clarity:
1. Equity: £10,000 2. Risk per trade: 1% = £100 3. Stop distance: 25 pips 4. Pip value (varies by pair and lot size — check your platform's contract specification) 5. Position size = £100 ÷ (25 pips × pip value per lot)
If pip value per standard lot is roughly £10 (for GBP-denominated accounts on many majors), that's £100 ÷ £250 = 0.4 lots.
Do this calculation for every trade, not just when you remember to. Most platforms — including MetaTrader on Pepperstone's servers and IG's own platform — let you build this into a spreadsheet or a simple position-size calculator that pulls live pip value. What they won't do automatically is factor in your actual spread and commission cost, which is why sizing purely off the stated stop distance often understates true risk. Run your instruments through PipTax's /audit.html cost tool so your "stop distance" reflects real round-trip cost, not just the chart distance.
Volatility-Based Sizing: Adjusting for Market Conditions
Fixed-fractional sizing answers "how much equity to risk" — but it doesn't answer "how wide should my stop be." That's where volatility-based sizing comes in, typically using the Average True Range (ATR):
- High ATR environment (news events, illiquid sessions): wider stops needed, so smaller position size for the same % risk
- Low ATR environment (quiet ranges): tighter stops possible, so larger position size for the same % risk
- Common approach: stop = 1.5× to 2× the 14-period ATR, position size calculated from that distance
This matters because a fixed pip stop (say, always 20 pips) makes no sense across both a quiet Asian session and a high-impact US data release. The stop gets hit by noise in the first case and is far too tight to survive normal movement in the second. Volatility-based sizing keeps your risk-in-money constant even as the risk-in-pips shifts with conditions.
Correlation: The Hidden Multiplier of Real Risk
Per-trade sizing discipline can still leave you badly over-exposed if you don't account for correlation between open positions. Three separate 1%-risk trades can add up to far more than 3% real risk if they're all long the same underlying theme — for example, long GBP/USD, long EUR/USD and short USD/JPY simultaneously.
Practical steps:
- Group correlated pairs before sizing — most majors cluster around USD strength/weakness, commodity currencies, or risk sentiment
- Cap total risk per correlated group, not just per individual trade — a common rule is 1.5-2x single-trade risk as the ceiling for any one group
- Recheck correlations periodically — they shift with macro regime, and a pair that moved independently six months ago may not now
This is genuinely one of the most under-taught parts of retail risk management, and it's why traders following "1% per trade" rules can still blow up: the rule was never broken on paper, but the real portfolio risk was 3-4x what anyone accounted for.
Costs, Swaps and the True Size of Your Risk
Position sizing calculated on stop distance alone ignores three real cost drags that erode your actual "R":
| Cost type | Impact on sizing | |---|---| | Spread | Widens effective stop distance, especially on entry | | Commission | Fixed per-lot cost, disproportionately affects tight stops | | Swap/rollover | Compounds on positions held overnight, especially in carry-heavy pairs |
None of these are fixed across brokers or account types — Pepperstone and IG, for instance, offer different account structures (raw spread plus commission vs all-in spread) that change how these costs bite. Rather than guessing, run your actual instruments and stop distances through /cost-impact.html to see how costs change your effective risk before you commit capital, and compare account types on /brokers/index.html.
Building a Hard Equity Floor Alongside Fractional Sizing
Fixed-fractional and volatility sizing reduce the *speed* of drawdown, but they don't prevent a slow bleed from an underperforming strategy or a bad trading period. Pair your sizing model with a hard stop-trading trigger:
- Monthly drawdown limit — e.g. stop trading new positions if the month is down 5-6%, review before resuming
- Consecutive-loss trigger — pause after a defined number of losses in a row to check for a regime change, not just bad luck
- Equity floor from peak — a trailing maximum drawdown level that forces a full strategy review, not just a pause
These floors exist because percentage-based sizing alone still allows a strategy with a genuinely broken edge to grind an account down slowly and steadily — the % risk per trade stays "correct" while the strategy itself stops working.
Testing Your Sizing Model Before Going Live
Advanced position sizing is not something to trust on theory alone. Before running any new model — fixed-fractional, volatility-based, or correlation-capped — test it against your own last 50-100 trades (not a vendor backtest, not a demo run by someone else):
- Recalculate what position size the model would have given for each historical trade
- Check the resulting drawdown curve, not just the ending equity
- Compare against your current sizing approach to see the actual difference in worst-case drawdown
Trading is inherently risky and no sizing model prevents losses — it only shapes how losses and gains compound over time. The goal of advanced position sizing isn't to win more often; it's to make sure that when you're wrong, being wrong doesn't end the game.
Key takeaways
- Fixed-fractional sizing (risking a constant % of equity per trade) compounds losses down and gains up automatically — but the % you choose matters more than any entry signal
- Volatility-based sizing (using ATR or standard deviation) keeps risk consistent across pairs and regimes, unlike fixed pip-stop sizing
- Correlation between open positions can quietly multiply your real risk well beyond your per-trade %, especially across GBP, EUR and commodity-linked pairs
- Spreads, commissions and swaps eat into the 'R' you think you're risking — always size using worst-case cost assumptions, not headline numbers
- Use a hard equity floor (e.g. a monthly stop-trading trigger) alongside fractional sizing, since % risk alone doesn't prevent death-by-a-thousand-cuts drawdowns
- Test any sizing model on your last 50-100 trades before trusting it live — theory and your actual trade distribution often disagree
Frequently asked questions
- What percentage should I risk per trade with fixed-fractional sizing?
- There's no universal number, but most professional and prop-desk frameworks sit between 0.25% and 1% of equity per trade. Higher percentages (2%+) survive fine in backtests with short losing streaks but rarely survive real drawdowns, where 8-10 losses in a row happen more often than traders expect. Start conservative and only raise the % after you have a verified track record on your own execution, not a demo or vendor backtest.
- Is fixed-fractional better than fixed lot-size position sizing?
- For most retail traders, yes. Fixed lot sizing means a losing streak costs proportionally more as your account shrinks, and winning streaks don't compound. Fixed-fractional automatically scales position size to current equity, which smooths the risk of ruin — but it requires you to recalculate lot size before every trade rather than using the same size out of habit.
- How does volatility-based position sizing actually work?
- Instead of a fixed pip stop, you size the position so that a stop placed at a multiple of the pair's Average True Range (ATR) still only risks your chosen % of equity. Wider ATR (more volatile conditions) means a smaller position for the same % risk; tighter ATR means a larger one. This keeps your dollar risk constant even as market conditions change.
- Do spreads and commissions really affect position sizing calculations?
- Yes, and it's often overlooked. If your stop is 20 pips and the spread plus commission round-trip is worth 2-3 pips, your real risk is 10-15% higher than the headline stop distance suggests. This matters more on wider-spread pairs and during news events. Check live costs on PipTax's cost tool before finalising size on a new instrument or broker.
- How do I account for correlation when sizing multiple open positions?
- Group your open trades by rough correlation (e.g. EUR/USD and GBP/USD often move together) and cap total risk per correlated group, not just per trade. A common approach is to treat a correlated group as a single position for risk purposes, capping combined risk at roughly 1.5-2x your normal single-trade risk rather than letting it stack to 3x or more.