A perpetual futures contract is an agreement to track the price of a crypto asset with no expiry date, kept anchored to the spot market by a periodic funding payment between longs and shorts. That single design choice, the funding mechanism, is what separates perpetuals from dated futures and what makes them the dominant leveraged instrument on Binance, OKX, Bybit, Bitget, KuCoin and Gate. Understanding how funding, liquidation, leverage and order types interlock is the difference between using the contract as a precise tool and discovering its mechanics the hard way. This article walks through each mechanism in turn, then shows how they combine in practice.
Funding rates: the tether between perpetual and spot
A perpetual contract has no settlement date, so nothing forces its price to converge on spot at expiry. Exchanges solve this with a funding rate, usually paid every eight hours, sometimes more frequently on volatile pairs. When the perpetual trades above spot, the rate is positive and longs pay shorts. When it trades below, the rate is negative and shorts pay longs. The payment is calculated on the notional value of the position, not on margin, so a trader using high leverage pays funding on the full contract size.
The rate itself is typically composed of two parts: an interest rate component, often a small fixed value, and a premium component that reflects how far the perpetual has drifted from spot. Exchanges clamp the total between a floor and a cap to prevent extreme payments. The practical consequence is that funding acts as a slow tax on crowded positioning. When everyone is long and the rate stays positive for days, holding that position costs money every eight hours, and the cost compounds against the trade's edge.
Traders who analyse funding often look at it as a sentiment gauge rather than a cost line. Persistently high positive funding across major pairs suggests leveraged longs are crowded, which historically precedes long squeezes. Deeply negative funding suggests the opposite. Reading this alongside price structure is a common approach, and it is one of the inputs a multi-model stack can surface without the trader having to poll each exchange manually. The AI trend analysis layer, for instance, treats positioning extremes as context for whether a trend has room to continue.
Leverage and margin: the two numbers that define risk
Leverage is expressed as a ratio, such as 10x or 50x, and it describes how much notional exposure a trader controls per unit of margin. Margin is the collateral locked to support that exposure. The two are inverses: at 10x leverage, the initial margin required is one tenth of the position's notional value. Exchanges distinguish between initial margin, the amount needed to open, and maintenance margin, the minimum required to keep the position alive. The gap between them is the buffer that absorbs adverse price movement before liquidation.
Most exchanges use a tiered margin system. Small positions enjoy the highest available leverage and the lowest maintenance margin requirement. As position size grows, the maximum leverage falls and the maintenance margin rate rises. This exists because a very large position that gets liquidated can move the market on the way out, and the exchange wants more collateral behind it. The tiers are published per contract and are worth reading before sizing a trade, because the effective leverage available at one notional size may be half of what it is at another.
| Concept | What it represents | How it changes |
|---|---|---|
| Initial margin | Collateral to open a position | Set by chosen leverage |
| Maintenance margin | Collateral to keep it open | Rises with position size tier |
| Funding rate | Periodic payment between longs and shorts | Floats with premium to spot |
| Liquidation price | Level where the engine closes the position | Moves with margin added or removed |
Isolated margin and cross margin are the two account modes. In isolated mode, only the margin assigned to a specific position backs it, so a liquidation there cannot touch the rest of the account. In cross mode, the entire available balance supports every open position, which lowers the chance of a single liquidation but exposes the whole account to a cascading loss. The choice is a risk architecture decision, and it interacts directly with the position sizing framework covered in crypto position sizing.
Liquidation: how the engine actually closes you out
Liquidation is the forced closure of a position when its margin falls below the maintenance requirement. The liquidation price is not a fixed distance from entry; it depends on leverage, the maintenance margin rate for the position's tier, and any margin added or removed after opening. Higher leverage pulls the liquidation price closer to entry. Adding margin to an isolated position pushes it further away. This is why two traders with the same entry and the same direction can have very different liquidation prices.
When the mark price reaches the liquidation threshold, the exchange's risk engine takes over. On most venues the position is closed at the bankruptcy price, and any remaining margin is returned. If the engine cannot close the position at a price that covers the loss, the insurance fund absorbs the shortfall. Insurance funds exist precisely because liquidations during fast moves can execute worse than the theoretical bankruptcy price. Some exchanges also use an auto-deleveraging mechanism, which closes opposing profitable positions to balance the book when the insurance fund is insufficient.
The mark price deserves separate attention. It is a composite of spot prices across several venues, not the last traded price on the perpetual itself. This design prevents a single wick on a thin order book from triggering a cascade of liquidations. Traders who place stops based on the last price sometimes find their stop untouched while their liquidation price was hit, or vice versa, because the two reference different series. Understanding which price the engine uses is essential when the position is large relative to the book.
Order types: the execution layer of contract mechanics
Order types determine how a position is entered and exited, and they interact with the margin and liquidation mechanics above. A market order fills immediately at whatever liquidity is available, which is simple but can slip in fast conditions. A limit order rests at a specified price and only fills if the market reaches it, giving control over entry but no guarantee of execution. Stop-market and stop-limit orders trigger when a specified price is touched, and they are the standard tools for capping loss on a leveraged position.
Three order types are specific to derivatives and worth knowing. A reduce-only order can only shrink a position, never flip it, which prevents an exit order from accidentally opening a new position in the opposite direction. A post-only order is rejected if it would execute immediately, guaranteeing the trader pays maker fees rather than taker fees. A time-in-force setting such as IOC or FOK controls whether an unfilled order is cancelled or partially filled. Together these give fine control over how a position is built and unwound.
The placement of stop orders relative to the liquidation price is a recurring source of confusion. A stop placed beyond the liquidation price will never trigger, because the engine closes the position first. A stop placed too close to entry on a high-leverage position will be hit by normal noise. The practical approach is to size the position so the stop sits at a level the analysis supports and the liquidation price sits comfortably beyond it, which is a position-sizing problem more than an order-type problem. A distribution-based forecast, such as the one produced by the prediction engine, gives stop and target levels derived from the same probability model rather than from arbitrary distances.
How the mechanics combine in a live position
Consider a trader who opens a long perpetual position at 20x leverage with isolated margin. The initial margin is one twentieth of notional. The maintenance margin requirement for that position size sets the liquidation price some distance below entry. Funding is charged every eight hours on the full notional, so the position bleeds a small amount each period if the rate is positive. A stop-loss order sits between entry and liquidation. If price rises, the trader may add a trailing stop or a take-profit limit order. If price falls, the stop triggers before the liquidation engine does, and the loss is the distance from entry to stop multiplied by position size.
The failure modes are predictable. Funding costs accumulate silently on a position held for days. A stop placed too close gets hit by noise while the trade idea remains valid. A position sized so large that the liquidation price sits inside the noise band gets closed by the engine rather than by the trader's own decision. Each of these is a mechanics problem, and each is solvable by understanding the four components together rather than in isolation.
Platforms that surface these mechanics visually tend to reduce the errors. AlphaMind AI, for example, derives entry, target, stop-loss and position size from a distribution of forward paths using fixed rules, and the multi-model stack provides the regime and volatility context that determines whether a given leverage level is appropriate. The conversational layer, MindX GPT, explains those outputs in plain language without inventing levels of its own. None of this removes the need to understand funding and liquidation; it makes the consequences of each choice visible before the position is opened.
Frequently Asked Questions
What happens if I hold a perpetual futures position through a funding payment?
The funding payment is deducted from or added to your account balance at the scheduled time, based on the notional value of the position and the current rate. If you are long and the rate is positive, you pay. If you are short and the rate is positive, you receive. The payment does not close the position, but repeated payments can erode margin over time and move the liquidation price closer to entry.
Can I change my liquidation price after opening a position?
Yes, by adding margin to an isolated position or by switching to cross margin mode, both of which push the liquidation price further from entry. Reducing the position size or lowering leverage also moves it. The liquidation price is a function of margin, leverage and the maintenance margin tier, so any change to those inputs recalculates it.
Why did my stop-loss not trigger even though price reached it?
Stop orders on perpetual contracts typically trigger on the mark price or the last price, depending on the exchange and the order settings. If your stop was set against one series and the market moved on the other, it may not have triggered. Checking which price reference the exchange uses for stops and for liquidation is the first step in diagnosing the discrepancy.
This article is educational content and does not constitute investment advice. Trading crypto derivatives involves substantial risk of loss and is not suitable for every trader.
