Hyper liquid Hyper liquid is a position-control workflow from entry to close
Hyper liquid is a perpetual position-control workflow built around USDC margin, one-way long or short exposure, and explicit orders for entry, adjustment, and exit. A trader opens with a market or limit order, reads signed size, entry price, mark price, margin, and unrealized PnL, then reduces or closes exposure with reduce-only instructions. Position-linked take-profit and stop-loss orders automate exits without turning a closing order into a new opposite position.
Posted:
On this page
Key takeaway: It is a perpetual exchange requiring USDC margin, where traders open positions, track unrealized PnL, adjust size, and exit with reduce-only orders.
Open a perpetual with one deliberate order
A Hyper liquid perpetual position begins when a market or limit order fills on the HyperCore order book. The user chooses BTC, ETH, SOL, HYPE, or another listed contract, selects long or short, sets an allowed integer leverage, and enters size in units of the underlying asset.
HyperCore represents each linear perpetual contract as 1 unit of the underlying asset; collateral is USDC, while most oracle prices are denominated in USDT. The order's notional equals price multiplied by size. A perpetual order below 10 USDC is rejected, so sizing begins with this protocol floor before available margin and the asset's leverage ceiling are considered.
Price and size formatting matter before the order reaches the book. An accepted price carries no more than 5 significant figures, and perpetual price precision is limited to 6 minus the asset's size-decimal setting; size itself is rounded to the asset's declared size decimals. MetaMask, Rabby, Coinbase Wallet, or WalletConnect supplies the account signature, while USDC deposited through Arbitrum supplies standard perpetual collateral.
Read the live position as a signed state
Position state is one-way per asset: a positive signed size represents a long, while a negative signed size represents a short. Hyper liquid displays entry price, position value, margin used, liquidation price, return on equity, cumulative funding, and unrealized PnL around that signed size.
Entry price changes only when exposure increases. It becomes the size-weighted average of the previous entry and the new opening fill; a partial close leaves entry price unchanged. Unrealized PnL follows side × (mark price − entry price) × position size, with side set to 1 for a long and −1 for a short. The displayed PnL is frontend accounting, while HyperCore's fundamental state consists of trades and margin balances.
Mark price drives the position's live unrealized number rather than the latest isolated trade. Funding posts once every 1 hour; its fixed interest component starts from 0.01% per 8 hours, converted to 0.00125% per hour, while the premium component follows contract-versus-oracle pricing. Funding changes account balance without changing signed size, so it belongs in a separate reconciliation when exposure stays constant but account value moves.
Choose cross or isolated margin before resizing
Margin mode determines where a position draws support as its PnL changes. Cross margin is the default and shares USDC across cross positions; isolated margin confines allocated collateral to one asset and permits margin additions or removals after entry.
Initial margin equals position size multiplied by mark price and divided by leverage, while leverage accepts any whole number from 1 through the contract's configured maximum. Raising leverage on an existing position does not rewrite its fill history or entry price; it changes margin allocation. The protocol sets maintenance margin at 50% of the initial margin required at maximum leverage, a durable rule even though each asset has its own maximum and margin tiers.
Margin removal has another fixed guardrail: the remainder must cover both the initial-margin requirement and at least 10% of total open notional. On a cross account, negative unrealized PnL elsewhere absorbs shared availability, so closing one position does not map one-for-one to immediately withdrawable USDC. With isolated margin, adding collateral changes the buffer around that position while its size and entry price stay intact.
Adjust size with reduce-only and a precise time in force
For the closely linked subject, see Using Hyper liquid. Position adjustment changes signed size through another order, while reduce-only makes the change one-directional. A sell reduce-only order closes part or all of a long, and a buy reduce-only order closes part or all of a short; the protocol rejects an instruction that would increase exposure.
One-way state matters after partial fills. The same asset holds a single signed position, so an ordinary order that trades past zero reverses direction; reduce-only blocks that reversal. A partial opening fill recalculates entry price only for the filled amount, whereas a partial closing fill preserves the entry price on the remainder. Review resting orders after resizing because GTC instructions continue independently until they fill or are canceled.
| Order option | Position-adjustment behavior | Standard duration or count |
|---|---|---|
| GTC | Rests at its limit price and accepts partial fills | Until filled or canceled |
| IOC | Fills immediately against available orders and cancels the remainder | One matching attempt |
| ALO | Adds liquidity only and is rejected if it would match immediately | One placement; if accepted, until filled or canceled |
Large routine reductions also support TWAP rather than one immediate order. TWAP sends 1 suborder every 30 seconds, limits each slice to 3% maximum slippage, and caps a catch-up slice at 3 times the normal suborder size. Those constants govern execution, yet completion still requires matching liquidity; an unfinished TWAP leaves a smaller open position that must be read from current state rather than inferred from requested size.
Attach exits that follow the position you hold
Position-linked take-profit and stop-loss orders use mark price as their trigger and reduce exposure when activated. Opening them from the position form sizes each exit to the entire position by default, which aligns the control with an intended full close.
Market TP/SL orders apply a 10% slippage tolerance after the mark-price trigger fires. A limit TP/SL separates 2 prices: the trigger decides when the order reaches the book, and the limit sets the worst acceptable execution boundary. That boundary prevents fills beyond it, but a fast move can leave the limit order resting, making price control and immediate completion distinct outcomes.
Exits attached to a parent entry order behave differently. The 2 child orders - 1 TP and 1 SL when both are configured - stay untriggered until the parent fills; canceling a partially filled parent also cancels its children, requiring fresh protection for the filled portion. A specifically entered TP/SL size remains fixed and does not resize with later position changes, so compare every open trigger with live signed size after each manual adjustment. TradingView chart handles expose the same triggers visually, as explained in Hyper liquid fees.
Close, then verify zero exposure and no orphaned orders
Position closure is complete only when signed size reaches 0 and all unwanted open orders for that asset are gone. Hyper liquid records the closing fill against the existing entry price, realizes PnL into account balance, and removes the open position after no size remains.
A market close prioritizes immediate matching, while a reduce-only limit close controls the execution boundary and remains pending if the book never reaches it. After any close request, read current position size, remaining order size, fill history, and account value; button submission alone does not prove completion. Cancel leftover GTC, TP, SL, or scale orders whose purpose ended with the position because an open instruction retains its own lifecycle.
Automated accounts query clearinghouseState for signed size, margin, entry price, cumulative funding, and unrealized PnL, then use orderStatus or userFills to confirm the transition. The official Python SDK exposes the same HyperCore state for monitoring. Keep one post-close rule: zero signed size proves exposure is flat, while zero relevant open orders proves the exit plan has no pending actions.
Hyper liquid FAQs
Can sub-accounts isolate position exposure from the master account?
Yes, each sub-account has its own address-level clearinghouse state, positions, margin summary, and open orders, so a BTC position there does not merge with the master account's BTC position. Position management remains per user even where another feature aggregates activity differently. Select the intended sub-account before placing, modifying, or closing an order, then verify that same account's signed size.
When should an automated trader schedule cancellation of open orders?
An automated trader should schedule cancellation when open orders must not survive a stopped process or monitoring gap. The scheduleCancel action requires a time at least 5 seconds in the future, permits no more than 10 triggers per day, and resets the count at 00:00 UTC. Re-arm it while the strategy is healthy; once triggered, verify positions separately because the action cancels orders rather than closing exposure.
Which position controls apply to HIP-3 perpetual markets?
HIP-3 perpetual markets inherit HyperCore order books and margining, so market, limit, reduce-only, and supported trigger orders follow the same position-lifecycle model. The market deployer defines the oracle, contract specifications, and leverage limits, which makes sizing boundaries market-specific. Read the selected DEX's metadata and clearinghouse state before applying an automation rule copied from a validator-operated BTC or ETH market.
Does changing leverage rewrite the displayed entry price?
No, entry price reflects opening fills and changes through a size-weighted average when exposure increases; closing trades leave it unchanged. Changing leverage alters initial-margin allocation rather than the trade history used for that display. After the adjustment, compare leverage value, margin used, position value, and signed size because those fields show whether the account accepted the new setting without implying another fill.
Which records explain a completed position after it leaves the open-position table?
User fills, historical orders, funding history, and account balance changes reconstruct a completed perpetual position after it leaves the open-position table. Match entries by asset, side, size, timestamp, and order identifier, then separate opening fills from closing fills. Closed PnL and entry price are frontend-derived convenience fields, while the durable accounting trail consists of trades, funding transfers, and margin balances recorded by HyperCore.