You cannot control whether your next trade wins. You cannot control where price goes, what a central bank announces, or how other people react. The one variable that is entirely yours is how much you put at risk. That makes position sizing the most important decision in trading — and, oddly, the least discussed.
Position sizing is the process of choosing how many units — shares, coins, contracts, lots — to trade, so that a losing trade costs a pre-decided, survivable amount. It is not about maximising the win; it is about capping the loss.
Two traders can take the identical entry, stop and target, and end up with completely different outcomes over a year, purely because one risked a consistent fraction of their account and the other sized by gut feel. Same signals, different survival.
Entry techniques get all the attention because they feel like the skill. But consider what actually destroys trading accounts. It is almost never one bad signal — it is a normal losing streak taken at an abnormal size.
Losing streaks are not a malfunction; they are a statistical certainty. Even a genuinely profitable system loses often. SRA Quant's own published methodology is explicit that its historical baseline won only 42.2% of the time — the edge came from winners being roughly twice the size of losers, not from being right frequently. A system that loses nearly six trades in ten will hit runs of five, eight, sometimes ten consecutive losses across a few hundred trades. Sizing is what decides whether that streak is an annoyance or an ending.
Here is the arithmetic most people never run — what ten straight losses does at different risk levels, and the gain then required just to get back to even:
| Risk per trade | Account after 10 straight losses | Gain needed to recover |
|---|---|---|
| 1% | −9.6% (90.4% remains) | +10.6% |
| 2% | −18.3% (81.7% remains) | +22.4% |
| 5% | −40.1% (59.9% remains) | +67.0% |
| 10% | −65.1% (34.9% remains) | +186.8% |
Losses and recoveries are asymmetric: lose 50% and you need +100% to break even. Small risk per trade keeps you in the game long enough for any edge to express itself.
The most widely used approach is fixed-fractional sizing: risk the same small percentage of your current account on every trade — commonly 0.5% to 2%. The formula is simple:
position size = (account × risk %) ÷ distance to stop
Worked example in forex terms. Say the account is $10,000 and the risk budget is 1%, so a losing trade should cost about $100. The setup calls for a stop-loss 50 pips away (a pip is the standard smallest price step in a currency pair). Then the position must be sized so each pip is worth $100 ÷ 50 = $2. On a pair where a mini lot pays about $1 per pip, that means two mini lots — not "whatever feels right today".
Same logic in crypto terms: account $10,000, risk 1% ($100), and the setup's invalidation level sits 4% below the entry. The position value is $100 ÷ 0.04 = $2,500. If the stop is hit, you lose 4% of $2,500 — the planned $100 — regardless of how volatile the coin is.
Notice what the formula requires: a defined stop before entry. If you don't know where the trade is wrong, you cannot size it. This is why serious frameworks attach an invalidation level to every setup — SRA Quant, for instance, derives it from recent volatility (1.5× the Average True Range) so the risk distance is known before any alert goes out.
Same unit count on every trade. Buying "1 lot, always" means your real risk swings wildly with each setup's stop distance. A 20-pip stop and a 100-pip stop at the same lot size differ in risk by 5×.
Sizing by conviction. "This one is obvious" is exactly the trade people oversize — and confidence is not correlated with outcome nearly as strongly as it feels. If the evidence is genuinely stronger, that should be visible in objective confluence, not adrenaline.
Doubling after losses. Martingale-style recovery sizing turns a routine streak into ruin. The table above assumes constant risk; increase it while losing and the numbers get much worse, fast.
Ignoring correlation. Three open positions in three highly correlated cryptocurrencies is closer to one position at 3× size. Risk budgets apply to exposure, not tickers.
Position sizing is one leg of a three-legged stool. The second is the risk-reward ratio — how much a win pays relative to what a loss costs — which determines whether a modest win rate can still produce positive expectancy. The third is selectivity: taking fewer, higher-quality setups instead of sizing down to justify taking everything. Get all three right and no single trade, or single week, can hurt you much. Get sizing wrong and the other two cannot save you.
One honest caveat to finish: correct sizing does not create an edge. It only ensures that if an edge exists, you survive long enough to realise it — and that if it doesn't, you find out cheaply. Deciding how much to risk is always yours; no analysis tool can or should do it for you.
SRA Quant applies these principles automatically. Every setup it scores carries a volatility-based invalidation level and a minimum risk-reward requirement before it can ever become an alert — so the risk is defined before the trade idea exists.
SRA Quant provides market analysis and educational content only. Nothing on this page or the platform constitutes financial advice, and past or backtested performance does not guarantee future outcomes. Trading involves substantial risk of loss.