Position sizing: the risk calculation that matters more than conviction

Position size is the bridge between an idea and an account. Two traders can hold the same view and reach very different outcomes because one risks a fixed, survivable amount while the other lets size expand with confidence. A good process makes risk predictable before a trade is opened, and it does so with arithmetic that takes about ten seconds.

Start with a fixed account-risk budget

Choose an amount, or a percentage of equity, that represents the maximum loss if the invalidation is reached. This is a personal risk parameter and there is no universal number. The test is behavioural: it should be small enough that a sequence of ordinary losses does not change how you act. If four consecutive losses would make you want to size up to recover, the budget is too large — and four consecutive losses is not unusual, as the arithmetic in losing streaks and drawdown maths shows.

Calculate from the distance to the stop

The relationship is one line: notional exposure multiplied by the percentage distance to invalidation equals the risk budget. Rearranged, exposure equals risk budget divided by stop distance.

Every row risks the same $200. That is the entire point: the four trades feel completely different and are identical in the only dimension that compounds. The wider the invalidation, the smaller the position — not because the idea is worse, but because the arithmetic says so.

Build size from loss, not from buying power

Buying power tells you the largest position a venue will permit; it says nothing about the largest loss an account can absorb. Start with the cash loss that is acceptable if the thesis is invalidated, then derive exposure from the distance to that point. This reverses the emotional sequence. The position is no longer selected because it feels small or large — it is selected because the defined failure is survivable.

Leverage enters only at the last step, and only as a funding question. In the second row of the table, $8,000 of notional on a $25,000 account needs no leverage at all. The same $8,000 at 10x requires $800 of margin — the price risk is unchanged, but the position now sits about 10% from a maintenance-margin level, which may be closer than the invalidation. When leverage puts the liquidation price inside the stop, the venue is managing the trade, not the plan.

Include the costs that are always there

• Fees on entry and exit, which on a round trip can be a meaningful fraction of a tight stop.
• Expected slippage — stops fill worse in fast markets, and worst during the moves that trigger them.
• Funding, if the instrument charges it, multiplied by the intended holding period.
• The spread, particularly outside the deepest liquidity hours.

On a 1% stop, a round-trip cost of 0.12% plus 0.1% of slippage is over a fifth of the planned risk. It does not make the trade unworkable; it makes the real risk $244 rather than $200, and a process that ignores it will consistently run larger than it believes.

Treat correlated trades as one risk book

Several positions can look diversified while sharing a single driver. A book that is long high-beta equities, long crypto and long a growth-sensitive currency is one risk-on position expressed three ways, as the correlation between crypto and equities makes plain. Three trades at 0.8% each is a 2.4% event, not three 0.8% events, whenever the shared driver moves.

A workable rule is a cap on aggregate risk per driver: no more than a set percentage of equity exposed to the same macro factor at once, reviewed before adding rather than after. The review matters most around scheduled macro events and in thin liquidity, when correlations rise exactly when you would want them not to.

Position-size checklist

1. Use only capital that can bear loss without changing essential obligations.
2. Fix the risk budget for the idea before looking at the chart again.
3. Measure entry-to-invalidation as a percentage.
4. Divide to get notional; check the implied leverage and the liquidation distance.
5. Add fees, spread and plausible slippage to the true risk figure.
6. Cap combined risk across positions driven by the same factor.

The purpose of sizing is not to maximise any single opportunity. It is to make it possible to take the next good opportunity with the same clarity after this one is wrong. Consistency is the thing that makes a process measurable over a meaningful sample — everything else is a story about one trade. The risk disclosure covers the same territory in formal terms, and it is worth reading once properly rather than scrolling past.