Traders spend most of their attention on the question "what should I buy?" and almost none on "how much?" — yet the second question decides who survives. Two traders can take the identical entries all year, and end the year one solvent and one blown up, purely on sizing. This guide covers the arithmetic that governs that outcome, none of which requires anything beyond division.
The asymmetry that rules everything
Losses and gains are not symmetric. Lose 10% and a 11% gain restores you. Lose 25% and you need 33%. Lose 50% and you need 100% — a double, just to get back to even. Lose 80% and you need 400%. This curve is why the first job of sizing is not maximizing gains but capping the depth of drawdowns: damage compounds against you faster than recovery compounds for you. Every rule below is downstream of this one piece of math.
Risk per trade: the fixed-fractional method
The standard professional framework is embarrassingly simple: risk a fixed small fraction of your account on any single trade — commonly 1%, sometimes 0.5% or 2% depending on style. "Risk" means the amount lost if the stop is hit, not the position's total value.
The sizing formula falls straight out:
- Position size = (account × risk %) ÷ (entry price − stop price).
Concrete example: a $20,000 account risking 1% ($200) on a stock bought at $50 with a stop at $46. Per-share risk is $4, so the position is $200 ÷ $4 = 50 shares — a $2,500 position. Notice what happened: the stop distance determined the size. A tighter stop at $48 would allow 100 shares; a wider stop at $42 only 25. Size is the output of the risk decision, never the input. Traders who pick a "round" position size first and bolt a stop on afterward have the process exactly backwards.
Why 1%-ish numbers? Because losing streaks are normal, not exceptional. A strategy that wins 50% of the time will produce a streak of seven straight losses roughly once every hundred trades — routine over a trading year. At 1% risk, that streak costs about 7% — annoying. At 10% risk, the same ordinary streak costs over half the account, and the asymmetry curve above takes over. Position sizing is not about any single trade; it is about making the inevitable bad run survivable.
R-multiples: a common language for outcomes
Once every trade risks a defined amount, that amount becomes a natural unit — call it R. A trade that risked $200 and made $600 is a +3R trade; a stopped-out loss is −1R. This bookkeeping has two virtues. It makes performance comparable across position sizes and account growth. And it reframes the goal honestly: a trader with 40% winners is profitable if winners average +2R and losers stay at −1R — expectancy per trade is what compounds, not win rate. Chasing win rate while letting losers run past −1R is the retail signature; the R framework makes that failure visible in the ledger.
The rules that protect the rules
- The stop is part of the trade, not a suggestion. A stop moved "just this once" converts a −1R plan into an unbounded loss. The entire framework rests on losses staying the size they were designed to be.
- Cap correlated exposure. Five 1% positions in five semiconductor stocks are not five independent risks — on a sector selloff they are one 5% trade wearing disguises. Cap total risk per theme, not just per ticker.
- Size down around known event risk. Stops do not protect against gaps: a stock can open far below your stop on earnings morning (see the earnings guide), and a currency can gap across a weekend. Holding through binary events with full size means accepting that −1R can become −3R on a gap.
- Respect volatility. A stop 2% away in a sleepy utility and a stop 2% away in a crypto pair are not the same tripwire — one is a real level, the other is noise that will trigger by lunchtime. Stops belong beyond meaningful levels (see support and resistance), and size adjusts to the distance, per the formula.
The honest conclusion
No sizing scheme creates edge. If a strategy loses money on average, fractional sizing only schedules the funeral politely. What sizing does is guarantee that if you have an edge, you stay in the game long enough for it to pay — and that no single idea, however convincing, can end you. That is the actual dividing line between trading and gambling: not the instruments, but whether ruin is a possible outcome of one decision. Keep every trade's risk defined, denominated in R, and small enough that being wrong is boring — then spend your creativity on the entries, where it belongs.