The first time I got liquidated on a coin-margined position, the price had barely moved against me. I remember staring at the fill and doing the arithmetic twice because it did not add up. My stop was where I thought it was, the leverage was modest, and yet the position was gone. What I had missed is the thing almost nobody models when they flip from USDT-margined to coin-margined contracts, which is that the two settle in different currencies, and that single fact bends the whole risk curve.
If you have only ever traded linear USDT-margined perps, coin-margined ones feel familiar and behave completely differently once the market moves. So it is worth walking through exactly where they diverge before you size a single trade in one.
The mechanical difference in one paragraph
A USDT-margined contract is linear. Your collateral is a stablecoin, your profit and loss is denominated in that stablecoin, and one dollar of adverse move is one dollar of loss no matter where the price is. A coin-margined contract is inverse. You post the underlying asset itself as collateral, say BTC, and the contract pays out in that same asset. The payoff is built around the inverse of price rather than price directly, which means your PnL curve is not a straight line. It is convex on the way up and convex against you on the way down, and the collateral you are standing on is the very thing whose value is moving.
That last part is the trap. On a linear contract your margin is inert. A dollar is a dollar whether BTC is at 20k or 60k. On an inverse contract, when you are long and price falls, you lose on the position and the coins backing that position are worth less at the same time. Two forces pull in the same direction. When you are short and price rises, same problem in reverse, your short bleeds and your collateral does not appreciate to cushion it because it is denominated in the asset you are short. The losses compound in a way the linear contract simply does not reproduce.
Why the losses feel bigger than your leverage suggests
Here is the part traders skip. When you punch in 5x leverage on a coin-margined long, you are not carrying the same real risk as 5x on the linear version of the same trade. The inverse payoff plus the falling collateral value means your effective exposure grows as the market goes against you, right when you least want it to. Your liquidation price sits closer than the naive calculation implies, and it moves toward you faster as losses accumulate.
The rough mental model I use is this. On a linear long, a 10 percent drop in price is roughly a 10 percent loss on notional. On an inverse long, that same 10 percent drop costs you more than 10 percent in coin terms, and because those coins are also worth less in dollars, the dollar hit is larger still. The exact numbers depend on entry and the contract, but the direction is always the same. Convexity is not your friend when you are on the wrong side of it.
The flip side is real too. When an inverse position goes your way, the convexity works for you, and gains in coin terms can outrun what the linear contract would have paid. That asymmetry is exactly why some traders like these contracts. It is also why undisciplined position sizing on them ends accounts. You feel like a genius on the good days and you get carried out on the bad ones, and the bad ones arrive convex.
Who coin-margined contracts actually suit
These contracts were not designed to give retail traders a spicier way to gamble, even though that is often how they get used. They make the most sense for people who already hold the underlying and think in terms of it rather than in dollars.
- Miners who earn coin and want to lock in a price without selling their inventory. A coin-margined short lets them hedge in the same asset they are paid in, so the collateral and the hedge speak the same language.
- Long-term holders who measure their wealth in BTC or ETH rather than fiat, and who want to hedge downside or earn funding without converting to stablecoins and creating a taxable or logistical event.
- Anyone whose base unit of account is genuinely the coin. If a good outcome for you means ending the year with more BTC, not more dollars, an inverse contract is arguably the more honest instrument.
If none of that describes you, and you think in dollars, and you would be annoyed to end up with more coins but fewer dollars, then a coin-margined contract is adding a currency mismatch on top of your directional bet for no reason. The linear contract keeps your PnL in the unit you actually care about, which is usually the right call.
How to adjust when you switch contract types
The practical work is resizing and rebuffering. Do not carry your linear habits across untouched. A checklist I run before opening an inverse position:
- Cut the leverage you would have used on the linear version. If 5x felt right there, treat 3x as your starting point here and only add back if you have genuinely modeled the inverse payoff. The convexity is doing hidden work.
- Compute your liquidation price from the contract mechanics, not by eyeballing it as a fixed percentage. It sits nearer than the linear intuition suggests, and the exchange calculator is your friend here. Do not trust a round number in your head.
- Widen the liquidation buffer beyond what you would keep on a linear trade. Because the collateral value drops with the position on a long, a wick that would have been survivable on the linear contract can close you out on the inverse one. Give it more room.
- Decide your unit of account before you enter, not after. Are you trying to protect dollars or coins? The answer changes whether the trade even makes sense and how you judge the result.
- Watch funding as a separate line item. Funding on coin-margined perps is paid and received in coin, so a long stretch of adverse funding quietly erodes your coin balance on top of any PnL. On a linear contract that drain is in stablecoin and easier to notice.
The failure mode I see most often is someone who has done well on linear perps, moves to coin-margined to chase the convex upside, keeps the same leverage and the same mental liquidation buffer, and gets stopped out on a move that would have been a non-event on the contract they were used to. Nothing about their trading got worse. They just changed the shape of the instrument and did not change anything else.
So the short version is that these are hedging tools first and speculation tools a distant second. If you hold the coin and want to hedge in kind, they are elegant. If you think in dollars and want directional exposure, the linear contract is usually the cleaner instrument and the one that will not surprise you at the worst moment. Either way, if you are switching between them, resize and rebuffer as if it were a brand new instrument, because in every way that matters to your account, it is.