Perpetual Swap Mechanics
How perp futures actually work — funding rates that keep them pegged to spot, mark vs index price, insurance funds that backstop losses, and the liquidation cascades that wreck overleveraged accounts in seconds.
The basics
What Makes a Perp Different
Traditional futures contracts expire on a specific date — after which both sides settle and the contract ceases to exist. Perpetual swaps (perps) have no expiry. They were popularised for crypto by BitMEX in 2016 (the idea goes back to Shiller's 1993 proposal) and have become the dominant crypto derivative: in the first half of 2023, perpetuals made up 75% of the average daily volume of listed Bitcoin futures. Because perps never expire, they need a different mechanism to stay tied to the underlying asset's spot price. That mechanism is the funding rate.
The Funding Rate
Every 8 hours on many major centralised contracts (some contracts settle every 4 hours, and Hyperliquid pays hourly), traders holding perps pay or receive a funding rate based on whether the perp is trading above or below the spot index. If the perp is trading at a premium (longs are paying more than spot), longs pay shorts a funding fee proportional to position size. If perp is at a discount, shorts pay longs. This incentivises arbitrageurs to push the perp back toward spot. Funding rates are typically 0.01% per 8h in calm markets (~11% APY) and can swing to 0.5%+ per 8h (~500% APY) on volatile smaller contracts during squeeze events, within exchange-set caps that vary by contract.
Leverage and Margin
Perps offer leverage — you can control a $10,000 position with $1,000 of collateral at 10x leverage, or $100 at 100x. The collateral is your 'margin'; the position is the notional size. If the price moves against you, your margin gets depleted; when your margin falls below the 'maintenance margin' threshold, the exchange liquidates your position to prevent losses exceeding your collateral. Higher leverage means smaller price moves liquidate you.
- Perps never expire; funding rate keeps them pegged to spot
- Funding paid periodically (every 8 hours on major centralised contracts): longs pay shorts when perp trades above spot, vice versa
- Leverage: control a big position with small collateral
- Liquidated when margin falls below the maintenance threshold
Key Takeaways
- Perps are the dominant crypto-derivative product by volume
- Funding rate is the mechanism that keeps perp price near spot
- Leverage amplifies both gains and losses; high leverage = quick liquidation
- Liquidations happen automatically and predictably when conditions are met
Going deeper
Mark Price vs Index Price vs Last Trade
Exchanges use three distinct prices for perps. Last trade price: the price of the most recent perp trade — can be manipulated short-term. Index price: a volume-weighted average of spot prices from several reference exchanges (Binance, Coinbase, Kraken, OKX, etc.) — much harder to manipulate. Mark price: the price used for liquidation calculations — usually the index price plus a smoothing factor to prevent flash-crash liquidations from temporary venue-specific blips. Understanding which price your exchange uses for liquidations is critical; some venues have liquidated users on last-trade-price during low-liquidity flash crashes, producing controversy.
Insurance Funds
When a position gets liquidated, the exchange sells the position at market. If the market price is below the bankruptcy price (where the user's margin is zero), the exchange takes a loss — the closeout price doesn't cover the position. The insurance fund absorbs these losses. It's funded in venue-specific ways (under dYdX v4's default settings, for example, a liquidation penalty of up to 1.5% is transferred to the insurance fund) and depleted when liquidations happen below the bankruptcy price. Major exchanges' insurance funds are in the hundreds of millions to billions of dollars (as of 9 November 2026, Binance's public insurance-fund endpoint showed about $1.8 billion in stablecoins across its contract groups). When the insurance fund is exhausted, the exchange resorts to 'auto-deleveraging' — clawing back winning positions to cover the loss.
Cross vs Isolated Margin
Cross margin: all your collateral backs all your positions. A profitable position offsets a losing one's margin draw. Less likely to liquidate individual positions, but a single bad position can drain your whole account. Isolated margin: each position has its own margin allocation. Position A can liquidate without touching Position B. Common pattern: cross margin for hedged positions, isolated for high-leverage degens. The choice cascades into liquidation behaviour, fee economics, and risk management.
Funding-Rate Carry as Yield
Buying spot and shorting an equal-size perp captures the funding rate as yield, hedged against price movement. When funding is +0.05%/8h (~55% APY), this 'cash and carry' trade is a major source of basis-trade flow on ETH and BTC. Sophisticated funds run this systematically; basis trading accounted for part of 2024-2025's ETF demand (see the etf-institutional and crypto-news-literacy ETF lesson).
Key Takeaways
- Three prices matter: last-trade (manipulable), index (resistant), mark (used for liquidation)
- Insurance funds absorb closeout-below-bankruptcy losses; auto-deleveraging is the fallback
- Cross margin pools collateral; isolated margin compartmentalises risk
- Funding-rate carry is the foundation of basis trades and a major institutional flow
Under the hood
Liquidation Cascades
When the price drops sharply, leveraged longs hit their liquidation prices. The exchange sells those positions, pushing the price down further, hitting the next layer of liquidations, and so on. This is the cascade. Famous examples: May 19, 2021 (BTC dropped ~30% in hours), November 2022 FTX week, August 5, 2024 (Japan carry trade unwind + crypto cascade), October 10, 2025 (more than $19B of leveraged positions liquidated in 24 hours, the largest on record per Galaxy Research). Cascades are mechanical, not sentimental. Once they start they accelerate themselves until forced selling exhausts the leveraged positions in the danger zone.
On-Chain Perps: dYdX, GMX, Hyperliquid
Decentralised perps have grown from a niche to a substantial market. dYdX V4: standalone Cosmos chain, orderbook model, similar to centralised exchanges. GMX: pool-based model, traders trade against liquidity pools (GM and GLV pools in the current version) at oracle prices, no orderbook — simpler but capital-inefficient. Hyperliquid: vertically integrated L1, orderbook model, among the largest on-chain perp venues by volume (about $8.7 billion of venue-reported 24-hour notional volume on 9 November 2026). Synthetix: now a central-limit-order-book perps exchange margined and settled in USDT on Ethereum Mainnet. Each makes different tradeoffs in latency, capital efficiency, and decentralisation.
Funding-Rate Manipulation
When a perp's open interest is heavily one-sided, the funding rate becomes a target. A coordinated group can push the perp's price up to charge longs an extortionate funding rate, then take the receiving side. Exchanges cap the funding rate to prevent the worst extraction, and the caps vary by venue and contract (on 9 November 2026, for example, Binance's BTCUSDT and ETHUSDT perpetuals were capped at ±0.3% per 8 hours, while many smaller Binance contracts allowed up to ±2% per funding interval), but rates can still become extreme during squeezes. Watching funding rates is a quick read on positioning — extreme positive funding = crowded longs, extreme negative = crowded shorts.
Perp DEX Risk Beyond the Trader
Liquidity providers on perp DEXes (GMX's GM and GLV pools, Hyperliquid's HLP) take the other side of trader PnL — they profit when traders lose, lose when traders win. Historically, retail traders lose money in aggregate on perps (the casino vig of liquidations + funding makes it structurally hard to win), so LP returns can be substantial. But during sustained one-sided market moves (a multi-month bull where everyone is long), LPs can take serious drawdowns. Evaluating perp DEX LP risk requires understanding the protocol's hedging mechanisms and historical trader-PnL data.
High Leverage Is Statistical Suicide
Most retail traders who use high leverage on perpetual futures lose money over time. The math is harsh: at 100x leverage, a 1% adverse price move wipes out 100% of your margin. Markets routinely move 5-10% in volatile sessions. The combination of leverage + funding cost + liquidation cascades makes long-term perp trading structurally unfavourable for retail. Use perps for hedging existing exposure; speculative leverage is a wealth-transfer mechanism from retail to professionals.
Key Takeaways
- Liquidation cascades are mechanical and self-amplifying once triggered
- On-chain perps (dYdX, GMX, Hyperliquid, Synthetix) span orderbook and pool models
- Funding rates can be manipulated when open interest is one-sided; exchanges cap them, and the caps vary by contract
- Perp DEX LPs profit on average from trader losses but can take serious drawdowns in sustained one-sided markets
- Most retail perp traders using high leverage lose money; high leverage is structurally adverse over time
Related Content
Related Coins
Related Topics
More Topics
Browse all 189 topicsVerifiable Random Functions
How VRFs produce randomness that's both unpredictable before commitment and cryptographically verifiable after — enabling fair lotteries, leader election, and on-chain randomness without trusted parties.
Hash Functions Compared
SHA-256, Keccak-256, Blake3, and Poseidon — which one each chain uses, why ZK systems needed a new family of 'arithmetic-friendly' hashes, and what tradeoffs each makes.
Stealth Addresses and Confidential Transactions
Privacy primitives that hide who's receiving what — from Monero's foundational stealth addresses to Ethereum's ERC-5564 and the legal context post-Tornado-Cash.
CDP Lifecycle
How collateralized debt positions work end-to-end — minting DAI against ETH, paying stability fees, surviving liquidations, and the structural lessons from MakerDAO's Black Thursday and Liquity's no-fee model.
Intent-Based Architectures
Why CoW Swap, UniswapX, Across, and 1inch Fusion moved from 'sign a transaction' to 'sign an intent' — and how solver competition delivers better prices and MEV protection.
MEV-Share Mechanics
How Flashbots' MEV-Share lets users share metadata about their transactions in exchange for a cut of the MEV searchers extract — a structural shift from 'searchers prey on users' to 'searchers and users share value.'
Solver Networks
How solver markets shape DEX prices, the competitive structure across CoW Swap, 1inch Fusion, Hashflow, and other venues, and why solver consolidation is a central debate in DeFi market structure.
Liquidation Cascades
The mechanical structure of cascading liquidations — Black Thursday, May 19 2021, August 2024, October 2025 — what amplifies them, who profits, and why over-leverage creates correlated risk across the entire DeFi stack.
References & further reading
- secondary
- secondary
- primaryHyperliquid Docs - Funding
Hourly funding paid at one eighth of an 8-hour rate; interest component 0.01% per 8 hours.
- secondaryAckerer, Hugonnier, Jermann - Perpetual Futures Pricing
Origin of perpetual futures (Shiller 1993; BitMEX 2016) and the 75% share of listed BTC futures volume in H1 2023.
- secondaryGalaxy Research - One Year After 10/10: The $19B Liquidation Event
Size of the 10 October 2025 liquidations and the Hyperliquid auto-deleveraging event.