Position Sizing and the Divide-by-20 Rule

Position sizing is the only control in trading that works regardless of whether you are right. Everything else — entries, exits, stops, filters — depends on the market cooperating in some way. Sizing does not.

The rule

Available trading capital divided by twenty, as the maximum committed to any single position.

For options, that ceiling applies to the premium at risk on a long position. For futures, it applies to the dollar risk per contract — tick value multiplied by stop distance — not to the margin posted, which is a permission figure rather than a risk figure.

The number twenty is a convention, not a law. It reflects a working assumption of holding up to twenty units of exposure at once, and it is deliberately round.

Why crude beats sophisticated here

Sophisticated sizing methods exist and most of them share a weakness: they take inputs.

A method that scales position size by recent volatility depends on a volatility estimate. One that scales by conviction depends on a conviction score. One derived from an edge estimate depends on the edge estimate being correct, and when it is not, the method sizes up on exactly the strategies that do not deserve it.

Every input is something that can be wrong, stale, or fitted to a historical sample. A fixed divisor has none. It cannot be overfitted because there is nothing to optimise, and it cannot degrade quietly because there is no estimate feeding it.

That is the argument. Not that crude sizing is more accurate — it is that accuracy is not the property that matters in a control whose job is to survive being wrong.

Why the recovery maths demands it

The arithmetic underneath, and it is the reason sizing dominates every other risk control.

Losses and gains are asymmetric, because a loss shrinks the base the recovery must be earned from. A 10% decline needs about 11% to recover. A 20% decline needs 25%. A 50% decline needs 100%. A 75% decline needs 300%.

The curve steepens rather than rising evenly. Past a point, recovery stops being a matter of continuing the strategy and starts requiring returns the strategy has never produced.

Sizing determines whether a bad sequence lands in the recoverable region or outside it. Nothing else does.

Where the divisor should be larger

Twenty assumes twenty units of exposure. Three situations break that assumption, and the honest response is to raise the divisor rather than pretend they do not apply.

Averaging in. If one signal can produce two entries, each signal consumes two units rather than one. A strategy that averages into every position needs roughly double the divisor.

Higher concurrency. If you routinely hold more than twenty positions at once, twenty is not the right number. Size to worst-case simultaneous exposure, not typical.

Correlation. Five positions in instruments that move together is closer to one large position than five small ones. A strategy taking correlated exposure needs more capital to run at genuinely bounded risk than the position count suggests.

The general form: divide by the maximum number of simultaneous exposure units you could realistically hold. A rule sized to typical conditions fails in unusual ones, which is when it is needed.

The constraint people try to work around

Contracts are indivisible. You cannot buy a third of an option contract.

So if a strategy's typical position costs more than your ceiling, you have two choices and one of them is bad. Exceeding the ceiling means running multiples of the intended risk on every trade — a trader taking positions at a quarter of their account instead of a twentieth is running five times the intended risk, repeatedly. Skipping trades means running a filtered version of the strategy, filtered by cost rather than by anything related to expected value.

The honest conclusion in that situation is that the strategy is not appropriate for the account, not that the ceiling should be raised. Micro futures contracts at one-tenth notional are the one genuine escape, and they have no options equivalent.

What it does not do

It does not make a strategy profitable. Correct sizing on a losing strategy produces a slower loss. It buys time, not an edge.

It does not eliminate large losses. A gap through a stop, a halt, or an assignment can cost more than the position's expected maximum. Sizing bounds the damage; it does not prevent it.

It does not optimise returns. A smaller position produces smaller gains proportionally. This is a deliberate trade, and anyone who wants maximum returns should be clear that they are choosing a different objective.

Enforcing it in software

A sizing rule that lives in a spreadsheet is a intention. One enforced in code is a control.

Enforce in the component that places orders, not in whatever generates signals. A limit implemented per strategy cannot see the aggregate, and a second signal source bypasses it entirely. One enforcement point, applied to every order regardless of origin.

Treat incoming quantities as untrusted. A payload arriving over a webhook is user input. An endpoint that submits whatever quantity it is handed is one malformed alert from a very large order. Compute size from your own capital and bound it server-side.

Compute from risk, not from permission. Margin capacity tells you what you can open. It is not a risk assessment, and sizing from it is the fastest way to destroy a futures account.

Persist the state. A system that restarts and recomputes limits from current balance has silently reset its own measurement.

On a self-hosted deployment those limits live in your own environment, where no vendor default and no incoming signal can override them. Keeping the calculation off the submission path — the worker thread pattern — means the check cannot delay the order it is gating.

The honest limits

Twenty is arbitrary. It is a reasonable default, not a derived optimum, and someone trading a single uncorrelated strategy might justify a smaller divisor while someone running many correlated positions needs a larger one.

A fixed rule ignores information a more sophisticated method would use, and that is the cost of having nothing to misconfigure.

And no sizing rule saves a strategy with negative expectancy. It determines how long you survive being wrong, which matters enormously and is not the same as being right.

Frequently asked questions

What is the divide-by-20 rule? Capping any single position at available trading capital divided by twenty — $500 per position on a $10,000 account.

Why twenty specifically? It is a convention reflecting up to twenty simultaneous units of exposure. The right divisor depends on your concurrency and correlation.

Should I size from margin requirement? No. Margin tells you what you are permitted to open, not what you should. Size from dollar risk per position.

What if one contract costs more than my ceiling? The strategy is not appropriate for the account at that size. Micro futures at one-tenth notional are the one genuine workaround, with no options equivalent.

Does position sizing improve returns? No. It reduces them proportionally while bounding loss. That is the trade being made deliberately.

Disclaimer: This article is educational content about trading mechanics and software. It is not investment advice, financial advice, tax advice, legal advice, or a recommendation to buy or sell any security, nor a recommendation of any strategy, order type, or position size. Any instruments, figures, or examples are used solely to illustrate mechanics. Options and futures trading involve substantial risk of loss and are not suitable for all investors. Please read Characteristics and Risks of Standardized Options before trading options. Automated trading carries additional risks including software defects, connectivity failures, broker API changes, and outages that may prevent orders from being placed, modified, or cancelled. Stop orders do not guarantee an execution price and stop-limit orders may not execute at all. Past performance does not indicate future results, and no configuration, order type, position-sizing rule, or risk setting can guarantee a profit or prevent a loss.

StaxInvesting LLC sells self-hosted trading software. It is not a broker-dealer, investment adviser, or financial institution, and it does not manage accounts, hold member funds, place trades on behalf of members, or access member brokerage accounts. Members run the software in their own cloud environment, connect their own brokerage accounts under their own credentials, and are solely responsible for their configuration, their credential security, and every trade executed in their account. Broker order handling and available order types vary; verify against your broker's current documentation. Consult a qualified financial adviser and tax professional regarding your individual circumstances.