Position size is not a preference you bring to a trade. It is a number that falls out of a calculation, and the only input that calculation needs from you is where your stop goes. Traders who decide "I'll take 500 shares" before they have located a stop are not sizing a position — they are choosing a stop by accident, wherever 500 shares happens to land.
That ordering problem is the single most common mistake in trade risk, and it is entirely fixable with arithmetic you can do in your head. It is also arithmetic worth handing off: tools built for AI for day trading return a size alongside the stop, so the ordering cannot go wrong.
Most traders solve this backwards
Ask a trader how they size a position and a common answer is some version of "I usually do a few hundred shares" or "I put a couple thousand dollars into a setup." That answer describes a habit, not a calculation, and it means the stop — if one gets placed at all — is whatever is left over once the size has already been chosen.
Run the same decision in the correct order and the stop comes first, because the stop is the only part of the trade that is actually about the chart. It is the price at which the reason you entered has stopped being true: a swing low that would have to break, a level that would have to fail, a range that would have to be violated. Once that price exists, the distance from your entry to that price is fixed, and your risk budget in dollars divided by that distance is your size. There is no second decision to make.
This is easy to agree with in the abstract and hard to actually do, because "a few hundred shares" is a comfortable habit and a stop-first calculation is not. The habit version feels like sizing because a number gets chosen and a trade gets placed. It is not sizing in any meaningful sense — it is picking an amount of capital to expose and then discovering, after the fact, how much room that amount happens to leave for the trade to be wrong. Sometimes that room lines up with a real level by coincidence. Most of the time it does not, and the trader finds out during the trade rather than before it.
Same $300 risk budget, two orders of decision
| Order of decisions | Where the stop sits | Stop distance | Shares at $300 risk |
|---|---|---|---|
| Stop first | Structural — below the last swing low | $0.60 | 500 (falls out of the distance) |
| Size first | Wherever 1,000 shares' worth of room happens to land | $0.30 — inside normal noise | 1,000 (chosen before the chart) |
The bottom row is not a hypothetical. It is what happens whenever a trader has a target share count in mind before they have looked at where the structure actually is. The stop in that row is not protecting anything — it is a distance that was reverse-engineered from a number of shares, and it will get hit by ordinary noise long before it says anything true about the trade.
The only formula you need
The calculation has two steps and neither one is negotiable.
Step one: account size × risk percentage = dollar risk budget. A $30,000 account risking 1% per trade has a $300 budget for this trade, full stop, regardless of how good the setup looks.
Step two: dollar risk budget ÷ stop distance per share = position size. If your stop sits $0.60 below your entry, $300 ÷ $0.60 = 500 shares. If it sits $1.20 below, the same $300 only buys you 250 shares of exposure. The stop distance is doing all the work in that second step, which is exactly why it has to be decided first.
Two things trip people up here. The first is confusing dollar risk with position value. Five hundred shares at $18.20 is a $9,100 position, but the risk on it is still $300 — the difference between position value and dollar risk is exactly the stop distance, and conflating the two is how a trader ends up thinking a large position is automatically a risky one, or a small one is automatically safe. The second is applying the formula to the stop distance you wish you had instead of the one the chart actually gives you. The formula does not care which stop distance you plug in; it will hand back a confident, precise-looking share count for a made-up stop just as readily as for a real one.
Nothing about this formula asks what you were hoping to buy. If the honest stop distance for a setup only lets you take 40 shares, 40 shares is the trade. Widening the stop until a rounder number of shares fits is not a sizing decision — it is choosing to risk more than your budget allows and calling it something else.
Where the stop actually comes from
A structural stop sits at the price that would prove the idea wrong: the far side of the level you are trading, the low of the pullback that formed the setup, the point where the pattern you identified stops existing. It is not a round number, and it is not a dollar figure chosen because it feels tolerable. Those two are the most common substitutes, and both of them put the stop inside the range of normal price movement, where it gets triggered by noise rather than by the trade actually failing.
Finding that level is a reading-the-chart problem before it is a maths problem, which is where most of the actual skill in this sits. An AI chart analysis can point out the swing low or the level your stop should reference, but the stop distance it hands back is only as good as the level underneath it — a level a foot inside the real structure produces a stop that is technically correct and practically useless.
It also helps to notice what a good structural stop is not. It is not the price at which you would feel uncomfortable. It is not a number that makes the arithmetic land on a round lot. And it is not something you move once the trade is open because price is getting close to it — a stop you adjust mid-trade to give it "more room" is a stop that was never really testing the idea in the first place, it was testing your patience for watching a number go the wrong way. If the level was right when you entered, it is still right an hour later. If it was wrong, moving it does not fix that; it just delays finding out.
Risk percentage decides what a losing streak costs
The stop distance sets your size for one trade. The risk percentage you chose sets what happens across a run of losing trades, and every strategy eventually produces one — a bad week is not a sign the system is broken, it is a sign the system is working the way probability said it would.
Notice what does not appear in that table: how skilled the trader is, or how good the stops were. Eight losses is eight losses whatever caused them. The only variable a trader controls going into that stretch is what each one was allowed to cost, which is the risk percentage half of the formula, decided before any of the eight trades happened. Keeping a trading journal that records the risk percentage on every entry is the only way to notice, after the fact, that a bad month was made worse by sizing creep rather than by a run of bad reads.
Building the habit
The formula is two steps of arithmetic. What actually needs practice is doing the steps in order, every time, including on the trades that feel obvious.
Before you enter, in this order
-
Find the structural level first — The swing low, the broken level, the point that proves the idea wrong — before you think about how many shares you want.
-
Measure the distance from entry to that level — This number is fixed by the chart. It is not something you adjust to taste.
-
Divide your dollar risk budget by that distance — The result is your size. If it is an awkward number, it is still the number.
-
Check the size fits your capital — If the required size costs more than you can deploy, the trade is a pass — not a reason to shrink the stop distance instead.
-
Widen the stop to make a rounder share count fit — This is the calculation running backwards, and it is how a 1% risk budget quietly becomes a 3% one.
None of this makes a bad setup good, and it does not remove the risk of trading your own money against people who do this for a living. What it does is make sure the risk you are taking on any single trade is the number you actually chose, rather than whatever number a stop placed for convenience happened to produce. That distinction is invisible on a single trade and decisive over a hundred of them.
