Position size from risk and stop

Many beginners size a trade by money: "I'll put in 3,000 and divide by the price." That fixes the position, but the risk then changes from trade to trade, because stops sit at different distances.

Nora's 3,000 in Northwind Foods at 48.00 is about 62 shares; with the stop 1.50 lower, the stop costs about 62.5 × 1.50 ≈ 94 dollars. Her 3,000 in Spark Coin with a stop 15% away costs 3,000 × 0.15 = 450 dollars if the stop fills. Same amount in, nearly five times the risk.

So turn the question around. Start from the risk you allow (say 1% of the account, 100 dollars on 10,000) and the risk per unit (the stop distance plus costs), and let the number of units come out of them. Then every trade costs about the same share of the account when it fails, whether it's a calm fund or a wild coin.

Position size comes from two numbers you already have.

Risk in dollars = account × risk share. On 10,000 at 1%, that's 100.

Risk per unit = entry − stop + costs per unit (the spread and commission for buying and selling, per share).

Units = risk in dollars ÷ risk per unit, rounded down to whole units.

Marc's Northwind idea: entry 48.00, stop 46.50 where the idea fails, costs about 0.05 a share for the round trip. Risk per share: 1.50 + 0.05 = 1.55. Units: 100 ÷ 1.55 = 64.5, so 64 shares. Check: if the stop fills, 64 × 1.55 = 99.20, just inside the 100 limit.

Last, look at the price tag: 64 × 48.00 = 3,072 out of a 10,000 account, which the account can pay for. Notice what you never chose directly: the number of shares. It came out of the risk rule and the stop.

The formula rarely gives a whole number. Round it down, never to the nearest unit.

For Marc's Northwind trade, 100 ÷ 1.55 = 64.5. Rounding to the nearest gives 65, and 65 × 1.55 = 100.75: over the limit. Rounding down gives 64, and 64 × 1.55 = 99.20: inside it. Even 100 ÷ 1.45 = 68.97 goes down to 68: 69 × 1.45 = 100.05 is already over.

Seventy-five cents sounds like nothing, but the point isn't the cents. A rule you break a little "because it's closer" stops being a rule; next month it's broken by 75 dollars. Rounding down keeps the planned loss at or under your limit every time.

For instruments that trade in fractions or small lots, the same idea holds: take the largest size whose loss at the stop stays within the limit.

Leaving costs out of the formula makes the size too big. Nora sizes the Northwind trade without the 0.05 of costs: 100 ÷ 1.50 = 66.7, so 66 shares. When the stop fills, she pays the spread and commission anyway: 66 × (1.50 + 0.05) = 66 × 1.55 = 102.30, over her 100 limit.

Two dollars over looks harmless. But it happens on every losing trade, and it hides a habit: treating costs as something that "doesn't count". Module 1 showed how quickly costs add up.

Costs per unit also vary a lot. On a calm share they may be a few cents, a small slice of the stop distance. On a coin with a wide spread, or a share traded after hours, when spreads widen, they can be a real part of it. So the rule is simple: costs go into the risk per unit every time, not only when they look big.

The formula gives a maximum, not a target. A second ceiling sits beside it: what the account can actually pay for.

Marc's Harbor 20 idea: entry 112.00, stop 111.40, costs 0.04. Risk per share is 0.60 + 0.04 = 0.64, so the formula allows 100 ÷ 0.64 = 156.25, that is 156 shares. But 156 × 112.00 = 17,472, and the account holds 10,000. Pebble's practice account doesn't lend money.

So take the smaller ceiling: 10,000 ÷ 112.00 = 89.3, so 89 shares, worth 9,968. Close stops on slow instruments often hit this cap.

The wrong fixes: borrowing with leverage to reach 156 (that puts more than the whole account into one trade, and a gap would hit all of it), or moving the stop to make the formula agree. And even 89 shares means almost the whole account in one fund, a choice worth weighing separately; Marc would probably take fewer.

Whenever a ceiling changes the size, work out the risk again, because it's no longer the number you started from.

Marc capped Harbor 20 at 89 shares. Each share still risks 0.64 (0.60 to the stop plus 0.04 of costs), so if the stop fills he loses 89 × 0.64 = 56.96. On a 10,000 account that's about 0.57%, below his 1% rule.

That's the right direction. The two ceilings can only make the risk smaller than the rule, never bigger, and taking less than 1% is never a problem. What would be a problem is "topping up" the risk some other way, like moving the stop further away so the 89 shares risk the full 100. The stop stays where the idea fails.

Writing down the final risk also feeds your journal: later you'll compare planned losses with real ones, including the times a gap made them bigger.

With the same dollar risk, the stop distance decides the size: you divide the same 100 by a bigger or smaller number.

A slow fund like Harbor 20 needs a close stop, 0.64 a share with costs, so the formula allows many shares. A jumpy instrument like Spark Coin needs a wide stop, so the size shrinks by itself. That's the formula protecting you: a wild instrument automatically gets a small position, without anyone asking how you feel about it.

The same fact hides a temptation. Pull the stop closer and the formula hands you a bigger size, which feels like more profit for the same risk. But a stop pulled inside the normal daily wiggle gets hit by noise, as Nora's 40-cent stop on Bluefin did. Big size plus frequent noise exits is a quick way to collect many small losses.

So the order never changes: the stop goes where the idea is wrong, and the size follows.

The routine before every order is the same four steps.

1. Risk in dollars: account × risk share.
2. Risk per unit: the distance from entry to stop plus costs. For a short position, which profits when the price falls, the stop sits above the entry, so it's stop − entry + costs.
3. Units: risk ÷ risk per unit, rounded down.
4. Cap: no more than the account can pay for; then recheck the risk.

A short at 30.00 with the stop at 31.20 and 0.05 of costs: risk per share 1.25; 100 ÷ 1.25 = 80 shares; 80 × 30.00 = 2,400, inside the cap.

The step people skip most is the cap: the formula looks so precise that they forget the price tag.

How it fits together: risk share and stop give the size, rounded down and capped, with costs always in. Next lesson: a case where you size three trades for Nora's account.