TLDR: A crypto perpetual can be settled in two ways, and the difference decides how it behaves in a crash. A linear (USDT-margined) perpetual posts a stablecoin as collateral, so your margin holds its dollar value while the trade runs. An inverse (coin-margined) perpetual posts the coin itself, say BTC, as collateral, and its profit and loss follow a reciprocal, convex curve. That combination means a coin-margined long gets hit twice in a sell-off: the position loses value and the collateral loses value in the same move, which pushes the account toward liquidation faster than an equivalent linear position. The single read is the collateral denomination: coin-margined longs carry reflexive, self-reinforcing downside that stablecoin-margined longs do not. This is education only, not a buy or sell call.
Most traders pick a perpetual by ticker and leverage and never look at what backs the margin. On the big venues you can trade the same asset two ways: a BTCUSDT contract margined in USDT, and a BTCUSD contract margined in BTC. They track the same price, but they are not the same instrument once volatility arrives, because the collateral behaves differently. Here is what the two settlement styles are, how the inverse payoff curves, why coin-margined longs amplify their own liquidations, why the market migrated to linear, and how to tell when coin-margined risk is actually in play.
What is the difference between inverse and linear perpetuals?
The difference is what you post as margin and what your profit and loss are paid in. A linear, or USDT-margined, perpetual (for example BTCUSDT) uses a stablecoin as collateral and settles gains and losses in that stablecoin, so one USDT balance can margin any market and the collateral holds a steady dollar value. An inverse, or coin-margined, perpetual (for example BTCUSD) uses the base coin as collateral and settles in that coin, so a BTC-margined position is funded, marked, and liquidated in BTC. Same underlying price, different collateral: that single choice is what makes the two contracts diverge under stress.
How does an inverse contract's payoff differ from a linear one?
A linear contract has straight-line profit and loss, while an inverse contract is convex. For a linear long, profit in USDT is just size times the price change, a straight line. For an inverse position the profit is paid in the coin and follows a reciprocal formula:
Profit (in BTC) = Contract size in USD times ( 1 / entry price minus 1 / exit price )
Because the payoff depends on 1 divided by price, the curve bends. A coin-margined long earns progressively fewer coins as price rises and loses progressively more coins as price falls, and as price approaches zero the coin-denominated loss grows without bound. A worked example makes it concrete. Take a 10,000 USD notional long, entry 64,000, and a 20 percent drop to 51,200:
| Linear (USDT-margined) long | Inverse (coin-margined) long | |
|---|---|---|
| Collateral posted | 10,000 USDT (stable) | 0.15625 BTC (worth 10,000 at entry) |
| Position P and L on a 20% drop | about -2,000 USDT | about -0.0391 BTC (about -2,000 at exit) |
| Collateral value after the drop | still 10,000 USDT | 0.15625 BTC now worth about 8,000 |
| Total dollar hit to the account | the position loss only | the position loss plus about 2,000 of collateral depreciation |
The linear long loses on the position but its collateral is untouched in dollar terms. The coin-margined long takes the position loss and watches its BTC collateral lose value in the very same move.
Why do coin-margined contracts amplify liquidation risk?
Because the collateral and the loss move together, a coin-margined long faces a reflexive double hit that a stablecoin-margined long avoids. As one exchange puts it, when your margin is denominated in BTC a price crash hits you twice: your position loses value and your collateral loses value, which can accelerate liquidation, whereas a linear position keeps its collateral at roughly the same dollar value through the drawdown. The convex payoff makes it worse, because the coin-denominated loss on a long accelerates as price falls rather than growing in a straight line. So the margin ratio on a coin-margined long erodes faster than on an otherwise identical linear long, liquidations trigger sooner, and the forced selling can push price down into the next cluster of coin-margined longs. That self-reinforcing loop is why coin-margined-heavy markets have historically produced sharper long-liquidation cascades. The asymmetry matters too: it is the coin-margined long in a sell-off that carries the dangerous convexity, while a coin-margined short's maximum coin loss is bounded, so the risk is directional, not symmetric.
Why did the market move to USDT-margined perpetuals?
Perpetual trading has shifted from bitcoin-margined inverse contracts, which dominated when BitMEX and early Bybit were the main venues, toward USDT-margined linear contracts, and the large majority of BTC perpetual open interest and liquidity now sits in linear products. The reasons are practical. Linear contracts are easier to reason about, and the spread of stablecoins removed the need to keep all your trading capital in a volatile asset, so a trader can post dry powder that holds its value. The two contract types also attract different participants: a coin-margined trader often already holds the coin and is leaning short or hedging the stack they own, while a USDT-margined trader is putting stablecoin dry powder to work and tends to lean long. That structural split shows up in funding. On BitMEX, the funding spread between the bitcoin-margined XBTUSD and the USDT-margined XBTUSDT has averaged roughly 3.93 percent annualized and stayed negative in about 94 percent of rolling 90-day periods, a persistent gap that reflects the different crowds each contract draws.
Where this reading misleads you
The collateral denomination is a risk lens, not a forecast. It tells you how an account will behave in a drawdown, not which way price is going, and it does not replace watching leverage and positioning directly. Three honest caveats. First, the danger is specific: it is the coin-margined long in a falling market that carries the reflexive, convex hit, not every coin-margined position, and a coin-margined short can even be convexity-favored. Second, coin-margined is now a minority of open interest, so it is a concentrated risk pocket rather than the whole market, and you should size its influence to how much of a given venue's book is actually coin-margined. Third, some traders choose coin-margined on purpose, because if you are a long-term holder posting coins you already own to hedge, settling in the coin can be exactly what you want. The read is to know which collateral you are on and what that does to your liquidation behavior, not to avoid one style outright.
A three-step way to tell if coin-margined risk is in play
You do not need the order book to gauge this. A repeatable methodology:
- Check the collateral of the contract you are trading. A BTCUSDT-style ticker is linear and margined in a stablecoin; a BTCUSD-style ticker is inverse and margined in the coin. If you are on the coin-margined side, the reflexive risk applies.
- Read the direction against the denomination. Coin-margined plus long plus a falling market is the convex, double-hit case. Linear, or coin-margined short, does not carry the same collateral-depreciation loop.
- Confirm against where leverage actually sits. Set it next to aggregate open interest and funding across venues. If open interest and crowded funding are building on the long side while a slice of that sits in coin-margined contracts, the cascade fuel is real; if positioning is calm, the denomination is a footnote.
Athenum lines up open interest, funding, and the futures basis across the 14 exchanges it aggregates, so you can see where leverage is concentrated and pair it with the contract denomination you are trading. For the mechanics of how perpetuals stay pinned to spot in the first place, see Athenum. As a reference point, aggregate BTC perpetual open interest across those 14 exchanges was about 17.5 billion dollars with funding near 0.0065 percent per 8 hours and a long/short ratio of 1.42 as of 2026-07-18, positioning that is crowded long but not extreme. Athenum has 34 free calculators at athenum.xyz/tools that need no signup, and a free 7-day Pro+ trial with no card required.
The honest caveat
Knowing whether you are on inverse or linear collateral turns a hidden risk into a visible one, but it is a behavior lens, not a timing tool. Coin-margined longs carry reflexive, convex downside; linear longs do not; coin-margined shorts are a different case again. Use the denomination to understand how your account will act in a drawdown, confirm the market-wide risk against open interest and funding, and never mistake a collateral-structure read for a price forecast.
Education only. Not investment advice. No buy or sell recommendation is made or implied.
Sources
- CoinEx, "The Differences Between Linear Futures Contracts and Inverse Futures Contracts" (linear = USDT-margined; inverse = coin-margined; inverse PnL is non-linear) | coinex.medium.com
- Ethena, "Inverse vs Linear Contracts" (inverse payoff = contract multiplier times (1/entry minus 1/exit); convexity, higher downside exposure) | docs.ethena.fi/solution-overview/underlying-derivatives/inverse-vs-linear-contracts
- Backpack Learn, "BTCUSDT vs BTCUSD Perpetual Futures" (coin-margined collateral volatility; crash hits position and collateral together; stablecoin-margined collateral holds dollar value) | learn.backpack.exchange/articles/btcusdt-vs-btcusd-perpetual-futures
- BitMEX, "Q2 2026 Derivatives Report" (XBTUSD inverse vs XBTUSDT linear funding spread about 3.93% annualized, negative in about 94% of rolling 90-day periods; migration to USDT-margined) | bitmex.com/blog
- Bybit Help Center, "Introduction to Inverse Contract" (inverse contracts margined and settled in the base coin) | bybit.com/en/help-center
- Athenum, aggregate BTC perpetual open interest, funding, and long/short readings across 14 exchanges as of 2026-07-18 | athenum.xyz