Risk · Lesson 8 of 8 · 18 min read
Position sizing and risk
Every other lesson was about the market. This one is about the only input you fully control. Size is not a feeling — it is what you are willing to lose divided by the distance to the exit that says you were wrong, and every other number on the ticket falls out of it.
- Assumes you have read
- What a trade really costs,Perpetuals and funding
- After this you can
- Derive a position size from a written-down exit and a fixed fraction of your balance, rather than from how confident you feel.
The short version
- Size, exposure and risk are three different numbers. On the worked trade they are 0.083 BTC, 4,980.00 USDT and 99.60 USDT — the last is 1.00% of the account while the middle one is 49.80%.
- The method is a division: risk budget ÷ distance to your exit = position size. Leverage is not in that equation and cannot be. It decides the collateral, not the size.
- The quantity always rounds down to the market’s lot step. A size rounded up risks more than the budget it came from, which is the one rounding direction that turns a rule into a lie.
- Above 41x on this trade, the engine’s liquidation sits nearer than your stop and fires first. That crossing is arithmetic, not risk appetite, and you can compute it before you place anything.
- A 1% rule really costs 1.05% when the stop is hit, because both taker fees land on top of the loss. 20 losses in a row still leaves 81.8% of the account.
Ask a beginner how big their position is and they say a number of dollars. Ask them how much they are risking and they say the same number, or they say “it depends”, or they say nothing — because they have not decided. That gap is where accounts go. Not in the entry, not in the analysis, and almost never in the direction of the trade.
This lesson closes it with arithmetic. Every figure below is computed at render time by the same initialMargin, liquidationPrice and roundTo the futures ticket previews with and the engine settles with, so the numbers on this page and the numbers on that screen cannot disagree.
Three numbers, not one
Before any method, three words have to come apart. They get used interchangeably and they are not close to the same thing.
- Position size is the quantity you hold — 0.083 BTC. It is what a price move is multiplied by.
- Exposure, or notional, is what that quantity is worth — 4,980.00 USDT, or 49.80% of the account. It is what fees and funding are charged on.
- Risk is what you lose if the trade is wrong. And “wrong” has to mean something concrete: a price at which you decided in advance that the idea failed, and at which you exit. Without that price, risk has no value — not zero, no value.
A position without a written-down exit does not have a risk number. It has a maximum loss, and the maximum loss is everything.
The whole method is one division
Fix what you are willing to lose on one idea, as a fraction of the balance. Commonly cited ranges sit between half a percent and two percent; this lesson uses 1%, which is not a recommendation so much as a number to run the arithmetic with. Then:
- Account balance10,000.00 USDT
- Risk budget — 1% of it100.00 USDTThe most this idea is allowed to cost. Chosen first, before anything about the trade.
- Distance from entry to your exit1,200.00 USDT60,000.0 down to 58,800.0 — 2.00%. This comes from the chart, not from the budget.
That is the entire method. Two of the three inputs are decisions you make before you look at a ticket, and the size is the consequence. Reverse the order — pick a size, then wonder where the stop goes — and the risk becomes whatever the chart happens to allow, which is a different number on every trade and never the one you would have chosen.
The remainder the lot step leaves behind
0.08333333 BTC is not a size any venue will accept. BTCUSDT’s futures lot step is 0.001 BTC, so the order is 0.083 and the real risk at the stop is 99.60 USDT — 0.40 under budget. That slack is the correct direction to be wrong in, and on a cheaper asset with a coarser step it can be a lot more than a rounding error. Check the number rather than assuming it landed where you aimed.
The trade, worked end to end
Same trade, all the way through, at 5x leverage. This is what the ticket would show you before you pressed anything, computed the same way.
- Position size0.083 BTCFrom the division above. Fixed before leverage was chosen.
- Exposure at the entry4,980.00 USDT49.80% of the account is now moving with the price.
- Initial margin at 5x996.00 USDT9.96% of the account, posted as collateral and locked while the position is open.
- Taker fee, 0.05% of exposure2.49 USDT
- Cost to open (margin + fee)998.49 USDTThe ticket shows this exact figure on the Cost (margin + fee) row.
- Liquidation price the engine would give this48,241.219.60% below the entry — 9.8 times further away than your stop. Shown as Est. liq. above the Open long button.
The last row is the one worth sitting with. A stop being hit is not an accident — it is the rule doing exactly what it was written to do, at exactly the cost it was written to cost. The only surprising number in the whole ledger should be none of them.
What leverage actually changed
Run the identical position at every leverage the ticket offers. The quantity does not move, because nothing in the division that produced it depends on leverage. Only the collateral behind it moves, and the liquidation price follows the collateral.
Swipe horizontally to compare all columns.
| Leverage | Margin | Liquidation | Below entry | Which exit comes first |
|---|---|---|---|---|
| 2x | 2,490.00 | 30,150.8 | 49.75% | Your stop fills |
| 5x | 996.00 | 48,241.2 | 19.60% | Your stop fills |
| 10x | 498.00 | 54,271.4 | 9.55% | Your stop fills |
| 20x | 249.00 | 57,286.4 | 4.52% | Your stop fills |
| 40x | 124.50 | 58,794.0 | 2.01% | Your stop fills |
| 41x | 121.46 | 58,830.7 | 1.95% | The engine liquidates |
| 50x | 99.60 | 59,095.5 | 1.51% | The engine liquidates |
| 75x | 66.40 | 59,497.5 | 0.84% | The engine liquidates |
Every row holds 0.083 BTC. A 2x position and a 75x position of the same size lose the same amount per dollar the market moves — the 75x one has simply put up 66.40 of collateral instead of 2,490.00 to hold it.
Leverage does not add risk. It removes the buffer between your position and the engine’s exit, which is a different and much more sudden problem.
The leverage that liquidates you first
Read down the last column and there is a line in it. Below a certain leverage, the market reaches your stop before the engine reaches its threshold: you lose the 1% you planned and the position closes at a price you chose. Above it, the order reverses.
For this trade the crossing is at 41x. At 40x the liquidation sits at 58,794.0, below the stop at 58,800.0 — your exit wins by six dollars. One notch of the slider later it sits at 58,830.7, above it, and the engine wins.
What actually happens when the engine wins is worth spelling out, because it is worse than losing the same money a different way:
- The position closes at the mark, not at your level — a price the market chose during the move, with the ordinary 0.05% taker fee charged on it.
- You lose the whole posted margin, less whatever equity survived to maintenance. On the 41x row that margin is 121.46 USDT, so the loss is capped low — but so is your control over when it happens.
- Every open order you had on that contract is cancelled, including the stop that was about to protect you and any take-profit attached to the entry.
This is the whole practical argument against high leverage, and it is not a moral one. At 41x and above, your stop is decoration: it will be cancelled by a liquidation before it can fire. The previous lesson covers what that settlement actually pays out.
The 1% rule costs more than 1%
The risk budget covers the price move. It does not cover the fees, and the fees are charged on exposure rather than on the loss.
- Loss at the stop99.60 USDT1.00% of the account — the number you signed up for.
- Entry fee, 0.05% of 4,980.002.49 USDT
- Exit fee, 0.05% of 4,880.402.44 USDT
Two consequences follow, and neither is obvious. A tighter stop means a larger position for the same risk budget, which means larger fees on the same 1% — so the tightest stop is not automatically the cheapest plan. And on spot, where the taker rate is 0.10% rather than 0.05%, the same arithmetic doubles. What one round trip costs and why is the whole of the costs lesson.
Why the fraction has to be small
Everything above works at any fraction. The reason to keep it low is that losses are not symmetric with gains, and the asymmetry gets worse the deeper you go.
Swipe horizontally to compare all columns.
| Drawdown | Gain needed to get level |
|---|---|
| 5% | 5% |
| 10% | 11% |
| 20% | 25% |
| 25% | 33% |
| 40% | 67% |
| 50% | 100% |
| 75% | 300% |
Lose a tenth and you need about an eleventh back — close enough to fair. Lose half and you need a double. Lose three quarters and you need to quadruple what is left, which nobody does by trading better. The only reliable way to avoid a deep drawdown is to make every individual loss shallow, and the only way to do that is to size for it.
What a losing streak actually does
Fixed-fraction sizing has a property worth understanding: because each position is sized off the surviving balance, positions shrink automatically as the balance does. It is the mathematical opposite of doubling down.
Swipe horizontally to compare all columns.
| Risk per idea | After 20 losses in a row | Gain needed back |
|---|---|---|
| 1% | 81.8% | 22% |
| 2% | 66.8% | 50% |
| 5% | 35.8% | 179% |
| 10% | 12.2% | 723% |
20 losses in a row is a run almost nobody has. At 1% a trade it leaves an account that recovers with 22% of gains — a normal quarter. At 10% a trade the same run leaves 12.2% and needs an eightfold recovery. Nothing about the trades changed between those rows. Only the fraction did.
What the ticket does and does not do
This product will help you place the position and hold the exit. It will not decide the size for you, and it is worth being exact about where the boundary sits.
The percentage buttons size exposure, not risk
The 25 / 50 / 75 / 100% row under the quantity field sizes an order against your available balance. On futures that means it solves for the largest position that budget can collateralise: budget ÷ (price × (1/leverage + taker rate)), rounded down at the lot step. It is a real answer to a real question — how much can I afford — and it is not the question sizing asks.
- The 25% button at 5x0.207 BTC2,500.00 of budget, buying 12,420.00 of exposure on 2,484.00 of margin.
- Risk against the same stop248.40 USDT2.48% of the account.
What it will hold for you
- The quantity always rounds down. Sizing at 100% can never ask to spend more than you have, and a size derived from a budget can never exceed it.
- Futures entries can carry attached exits. Tick TP / SL on the ticket and each fill places two reduce-only orders sized to that fill. The stop leg is a stop-limit whose limit sits 0.5% through the trigger, and when one leg closes the position the sweep cancels the other within about five seconds.
- A gap can still get through both prices. The ticket says so itself: if the market jumps past the trigger and the limit together, the stop leg can be left resting rather than filled. A stop is an instruction, not a guarantee, on every venue there is.
- Leverage is fixed at open. While a position or a resting entry exists on that contract the slider locks and names which of the two locked it. You cannot re-lever a position you are already in — which is a constraint that mostly protects you.
- Spot is different. A spot buy can attach one exit — a take-profit or a stop-loss, not both, because the ledger has no OCO yet — and only on a market buy. The ticket shows the control only where it works rather than accepting a pair it cannot place.
Five ways this goes wrong
Mistaken belief: “I only used 10x, so the risk is small.”
What actually happens: Leverage says nothing about risk on its own. Risk is size times stop distance, and size came from a division leverage was not in. What leverage decided was how close the engine’s exit sits to yours — on this trade, everything from 41x up puts it in front.
Mistaken belief: “I sized at 25% of my balance, which is conservative.”
What actually happens: That is an exposure decision with no risk number attached. On this trade the 25% button at 5x risks 248.40 USDT against the same stop — 2.48% of the account, 2.5 times the intended amount.
Mistaken belief: “My liquidation price is far away, so I have room.”
What actually happens: A liquidation price is not a plan, it is a failure mode with a fee attached. It also moves: funding charged against margin walks it toward the market every crossing. The number that bounds your loss is the exit you wrote down, and only if it is nearer than the engine’s.
Mistaken belief: “I will widen the stop rather than take the loss.”
What actually happens: Widening a stop after the position is open changes the risk you already accepted into a larger one, at the worst possible moment to be deciding. The size was derived from that distance; moving the distance without resizing breaks the only equation holding the plan together.
Mistaken belief: “A 50% drawdown just needs a 50% gain.”
What actually happens: It needs 100%. The money that would have earned the second half is the money that is gone. That asymmetry is the entire reason the risk fraction is a small number instead of a comfortable one.