Position Sizing and Risk per Trade: The Math Retail Investors Skip in 2026

Ask a retail trader how many shares they bought and you’ll usually get a number that traces back to a feeling: the position “felt right,” or it was a round dollar amount, or it was “all in because this one’s a lock.” Ask a desk trader the same question and you’ll get a formula. That gap — sizing from conviction versus sizing from risk — is the single piece of math that separates an account that survives a cold streak from one that doesn’t.

This isn’t about picking better stocks. You can be right 55% of the time and still blow up if your position sizing is wrong, and you can be right 45% of the time and grind out a return if it’s right. The sizing decision is upstream of the entry decision. Here’s the part most people skip.

Size from risk, not from conviction

The core idea: decide how much money you’re willing to lose before you decide how many shares to buy. That dollar amount — your risk per trade — should be a fixed, small fraction of your account, not a function of how excited you are about the idea.

The formula has two inputs and one output:

• Account risk = the dollars you’ll lose if the trade hits your stop. Commonly 0.5% to 2% of account equity.
• Per-share risk = the distance from your entry price to your stop-loss.
• Position size (shares) = Account risk ÷ Per-share risk.

Work a concrete example. You have a $25,000 account and you cap risk at 1% per trade, so your account risk is $250. You want to buy a stock at $50 and you’ve decided that if it falls to $48, your thesis is broken — that’s your stop. Per-share risk is $50 − $48 = $2. Position size is $250 ÷ $2 = 125 shares, or a $6,250 position.

Notice what just happened. The position size fell out of the risk and the stop. You didn’t pick $6,250 — the math did. Tighten the stop to $49 and per-share risk drops to $1, so the same $250 of risk now buys 250 shares ($12,500). Loosen the stop to $45 and you can only hold 50 shares. The stop distance, not your enthusiasm, controls the size.

Fixed-fractional, fixed-dollar, and the Kelly trap

There are three common ways to set the risk-per-trade number, and they are not equally good.

Fixed-fractional is the workhorse. Because you re-base off current equity, a losing streak automatically shrinks your dollar risk — at 1%, a $25,000 account risks $250, but after it falls to $20,000 it risks $200. The method gets defensive exactly when you need it to, without you having to remember.

The Kelly criterion is where smart retail traders hurt themselves. Kelly tells you the bet fraction that maximizes long-run growth: for a simple bet, f = edge ÷ odds. The problem is that full Kelly assumes you know your win rate and payoff precisely. You don’t — you’re estimating both from a small, noisy sample. Overestimate your edge by a little and full Kelly tells you to bet a lot too much, and the drawdowns become brutal. The standard fix is fractional Kelly: bet a half or a quarter of what the formula says. Half-Kelly gives up only about a quarter of the theoretical growth rate while cutting the volatility of returns roughly in half. For most retail accounts, a flat 1% fixed-fractional rule lands in a similar place with far less to get wrong.

The drawdown math that decides survival

Here’s why 1% versus 2% isn’t a small detail. Losses don’t add — they compound, and recovery is asymmetric.

Run a string of ten losses in a row (which a 50%-win strategy will produce more often than you’d guess). Risking 1% each, your account multiplies by 0.99 ten times: 0.99^10 ≈ 0.904, a drawdown of about 9.6%. Risking 2%, it’s 0.98^10 ≈ 0.817 — an 18% drawdown. Double the risk, roughly double the hole.

Now the asymmetry. A 10% drawdown needs an 11% gain to get back to even. A 20% drawdown needs 25%. A 50% drawdown needs a 100% gain — you have to double what’s left just to return to where you started. This is why capping risk per trade is the whole game: it keeps your worst realistic losing streak inside a hole you can actually climb out of. At 1% per trade, even a punishing run leaves you down single digits. At 5% per trade, ten losses cut your account roughly in half, and now you need to double it.

The practical discipline that ties this together is the R-multiple. Define 1R as the dollars you risk on a trade — your $250. Then stop tracking trades in dollars and start tracking them in R. A winner that made $500 is +2R; a loser that hit its stop is −1R. Because every trade risks the same fraction, R-multiples are comparable across positions of wildly different dollar sizes, and your whole track record collapses into one honest number: your average R per trade, your expectancy. If it’s positive, position sizing is just the throttle on a working engine. If it’s negative, no sizing scheme saves you — and R-multiples are how you find that out before the account does.

The reason to log this somewhere structured rather than in your head is that memory is kind to winners and quietly deletes the losers. A database doesn’t. After 50 trades, sort by R and you’ll see the truth: whether your average is positive, how fat your worst loss really got, and whether you’ve been quietly creeping your risk up on the trades that “felt like locks.”

Building the habit

None of this is hard arithmetic — it’s a discipline problem. Pick one risk percentage (1% is a defensible default), compute your share count from the stop every single time, and log the result in R. The math protects you only if you run it before every trade, including the one you’re certain about. Especially that one.