Liquidation & Position Size Calculator

Maintenance margin is the minimum equity an exchange requires you to keep behind an open position, quoted as a percentage of its notional value. It is not zero because closing a large position takes time and moves the price, so the venue needs a buffer to unwind you without the account going negative and the loss landing on everyone else. First-tier rates are usually between 0.4% and 1%, and every exchange raises the rate in tiers as the position grows, which is why a very large trade is liquidated sooner than a small one at the same leverage. The rate matters more than it looks: it sets a hard ceiling on usable leverage at 1 divided by the rate, so a 0.5% maintenance rate makes 200x mathematically impossible to hold.

Almost always because the liquidation price sat closer to entry than the stop did, which happens the moment the leverage is high enough. A stop is a decision about the chart; a liquidation is a consequence of arithmetic, and neither knows the other exists. At 20x with a 0.5% maintenance rate the position dies on a 4.4% move, so a 6% stop is decoration. The second cause is the mark price: exchanges liquidate against an index or mark price rather than the last trade on their own book, so a wick that never touched your stop on the chart can still have taken the mark price through your liquidation level. This page checks the first cause for you. Only a wider buffer protects you from the second.

They are safe in different directions, and the honest answer is that it depends on which failure you are protecting against. Isolated margin caps the loss at the amount posted to that one position, so a liquidation costs exactly that and nothing else in the account is touched. Cross margin lets the whole free balance stand behind the trade, which pushes the liquidation price much further away and makes it far less likely to happen at all, but when it does happen the loss is not capped at one position's margin. Isolated is the better default because it makes the worst case a number you chose in advance. Cross is defensible on a single, carefully sized position, and it is genuinely dangerous with several open at once, because they share one balance and each one's loss brings the others closer to being closed.

Yes, and this is the one adjustment that reliably helps. Topping up an isolated position increases the margin term while the notional stays exactly the same, so the liquidation price moves further away in direct proportion. Doubling the margin on a 20x position turns it into a 10x position and roughly doubles the room. What surprises people is the opposite move: closing part of the position does almost nothing to the liquidation price, because it shrinks the margin and the notional together and the ratio between them is what the formula depends on. Closing half the position halves the loss you are carrying but leaves the liquidation price roughly where it was, so if you are trying to survive rather than to reduce exposure, adding margin is the lever and cutting size is not.

It is 100 times the profit per unit of margin on any move you survive, and the catch is in that last clause. At 100x with a 0.5% maintenance rate the position is closed on a 0.5% move against you, which on most pairs is one candle and sometimes just the spread. The profit on a 3% winner is the same 3% of the notional whether you posted 1% of it or 50% of it as margin, so leverage changes the denominator, not the trade. What it really buys is a much higher chance of being removed from the position before the move you were right about happens. That is why the ladder on this page shows the surviving move rather than a profit multiple: at 10x you have 9.5% of room, at 100x you have 0.5%, and the trade idea did not get ten times better.

Most position-sizing rules land between 0.5% and 2% of the account on any single trade, and the reason is arithmetic rather than tradition. At 2% per trade, ten losers in a row take about 18% of the account and it is recoverable. At 10% per trade, ten in a row take 65% and getting back to even requires nearly tripling what is left. Losing streaks of that length are ordinary in any strategy that wins half the time. The percentage also has to survive being wrong about the plan, not just about the direction: fees, slippage on the stop, and a gap through your level all make the realised loss larger than the budgeted one, which is why this page includes the fees on both sides in the size it returns.

Several small reasons, each worth a few basis points. Exchanges tier the maintenance margin rate by position size, so the rate on your ticket may not be the one you typed here. Some compute the maintenance requirement on the entry notional rather than on the mark notional, which is where the popular shortcut formula comes from and which shifts the answer slightly. Perpetual funding is charged or paid every few hours and changes the equity behind the position, so a liquidation price that was accurate this morning is not accurate tonight. Cross accounts with more than one open position share a balance, so every other trade you hold moves this one's liquidation price. And unrealised profit on other positions may or may not count as margin depending on the venue. Treat the figure here as the correct arithmetic on the inputs you gave it, and your exchange's own number as authoritative for your actual account.

A short, on both counts, and it is not close. The liquidation is slightly tighter: the maintenance requirement is charged on the notional at the current price, so as a short goes against you the position gets larger and demands more margin at exactly the moment you have less. At 10x with a 0.5% maintenance rate a long survives a 9.55% fall while a short survives a 9.45% rise. The bigger difference is what lies beyond the liquidation. A long's worst case is bounded because an asset can only fall to zero, so a 1x long can never be liquidated at all. A short has no ceiling: the price can rise by any multiple, which is why every exchange treats short exposure as the riskier side and why a short at any leverage above 1x has a liquidation price sitting above it.

The bankruptcy price is where the equity behind the position reaches zero, which sits a little further out than the liquidation price. The gap between the two is the maintenance margin, and it exists so the exchange has something to work with while it closes you out. In the normal case the liquidation engine sells the position somewhere between the two prices, keeps the remainder as a liquidation fee, and nobody else is affected. In a fast market it may not manage that, and the shortfall is covered by the venue's insurance fund. When the insurance fund cannot cover it either, most exchanges fall back on auto-deleveraging, which force-closes profitable traders on the other side of the book at the bankruptcy price. None of that is modelled here, because it depends on the exchange's own liquidity and rules, but it is the reason your loss on a violent move can exceed the number on this page.

No. There is no price feed, no API key, no exchange connection and no account. Every figure comes from a number you typed, the arithmetic runs in your browser, and nothing you enter leaves the page. The maintenance rate presets are labelled typical values to save you a trip to the risk-limit table, not quotes from any venue, and you should overwrite them with the rate your exchange applies at the size you are trading. The same goes for the example entry price and fee: they exist so the page has something to show, and every one of them is meant to be replaced.