Position sizing that survives a stop-out

The three inputs, in order

Stop distance, risk percentage, per-lot value. That is the whole list, and the order matters, because each one constrains the next.

The stop comes from the chart first, measured in pips, points or cents. It sits below the swing, above the wick, outside the noise band, never on a round figure picked for comfort.

The risk percentage is a policy number, chosen once and kept. It does not move with confidence or with the last result. The per-lot value is a market fact, looked up per instrument and nothing more.

Only once all three exist is the size touched, and from there it is pure division. On $100,000 at 1% risk the money at risk is $1,000. A 20 pip stop on a USD-quoted major costs $200 per lot, so the size is $1,000 divided by $200, five lots.

The per-lot value per instrument

The same arithmetic runs on every instrument; only the value changes. A JPY pair is quoted through USD/JPY, so the per-pip value is 1,000 divided by that rate: at 150, roughly $6.67 per lot, and the size rises.

Gold moves in cents and one lot is 100 ounces, so a dollar of movement is $100 per lot and gold sizes come out smallest. An index moves in the broker's stated points, often $1 each, with wider slips.

Exotics and crosses carry wider spreads. Fold the spread into the effective stop, or the size is quietly optimistic on every trade.

Rounding down, minimums and steps

Brokers trade in lot steps, usually 0.01, sometimes larger on indices. Divide by the step, round down, accept the leftover.

There is also a minimum. On a small account 1% can compute below it, which forces a tighter stop or a lower risk percentage. Never cap the size by a comfortable lot figure; the minimum is a floor from the broker, not a target.

After rounding, check the money once: size times stop times per-lot value must land on the planned risk. If it does not, fix the size.

Where the stop goes wrong

The classic error is sizing first. The trader likes the setup, decides on 2 lots, then hunts a stop that makes 2 lots feel safe. The stop ends up inside the noise, so normal movement takes it while the thesis is alive.

Another version is padding the stop for volatility and then keeping the size, which risks more than the plan.

A third is an honestly wide stop that produces a tiny size. That is arithmetic, not a failure: the setup is expensive, so it earns fewer lots. Take the small size or skip it.

Recalculating after every close

Equity moves with every close, so size is recomputed after every close. A spreadsheet row is enough and takes seconds.

Losing five in a row while sizing off the old balance risks more than 1% of what is left, exactly when it is least affordable. Winning five quietly drifts risk downward instead.

On an evaluation the limits move too. A daily drawdown measured from the day start means an early loss today shrinks what is left for later trades today. Prop firms sometimes fix the percentage as a rule; then the size follows that fixed number.

A worked session, one trade at a time

A 12 pip stop on EUR/USD: $1,000 risk divided by $120 per lot is 8.33 lots, and the real risk lands at $999.60, a hair under target.

The same session on gold with a 350 cent stop gives 2.85 lots and $997.50 risk. An index with a 60 point stop at $1 per point gives 16.66 lots and the same $999.60.

Three very different sizes, identical risk. That equality is the entire objective. Write all three rows with the money check beside each.

Common fixes and how to test them

If sizes consistently come out tiny, the stop is too wide for the instrument; tighten the stop to structure rather than raising the percentage. If sizes come out absurdly large, the per-lot value is probably misread against the broker's contract size.

Keep a one-line reference table of per-lot values so the figure is never re-derived from memory, and test the routine on paper for twenty trades, checking that every full stop costs the intended money.

When the routine is stable it takes under thirty seconds, which is what lets it survive a busy session. Review the percentage once a month against drawdown depth, not against confidence.