Oracle
The Hyperliquid Oracle System provides reliable price data for funding rates, margining, liquidations, and triggering Take Profit/Stop Loss orders through robust, manipulation-resistant mechanisms.
Spot Oracle Price 📈
Validator-computed: Published by validators every 3 seconds for high-frequency market updates
Weighted median of major CEX prices: Binance (3), OKX (2), Bybit (2), Kraken (1), Kucoin (1), Gate IO (1), MEXC (1), Hyperliquid (1)
Final clearinghouse price: Weighted median of validator-submitted prices (validators weighted by stake)
Usage: Primary input for funding rate calculations and mark price component
Special cases:
Hyperliquid-native assets (e.g., HYPE): External sources excluded until sufficient external liquidity
External primary liquidity (e.g., BTC): Hyperliquid spot prices excluded from oracle
Mark Price 📊
Unbiased, robust estimate of fair perpetual contract price using median of three components:
Oracle price + 150-second EMA adjustment • EMA of difference between Hyperliquid mid-price and oracle price • Smooths short-term fluctuations and integrates market deviations
Hyperliquid order book median • Median of best bid, best ask, and last trade price
External perpetual median • Median of mid-prices from Binance (3), OKX (2), Bybit (2), Gate IO (1), MEXC (1)
Special handling: If only 2 of 3 components exist, a 30-second EMA of Hyperliquid bid/ask/last trade is added
Update frequency: ~3 seconds (synced with oracle updates) Usage: Liquidations, margin calculations, TP/SL triggers, unrealized PnL
Example
Spot Oracle Price: Calculated as $10,000.
150-Second EMA: Indicates Hyperliquid’s mid-price is on average +$20 above the oracle price (i.e. $10,020).
Hyperliquid Median (Bid/Ask/Last Trade):
Suppose the current best bid is $10,005, the best ask is $10,015, and the last trade occurred at $10,010.
The median of these three numbers would be $10,010.
External Perpetual Median:
Imagine the mid-prices from these exchanges are $9,995, $10,000, and $10,010.
The median of these values is $10,000.
The mark price is computed by combining the spot oracle price with the above inputs. In our example, the final mark price might then be in the vicinity of $10,010 to $10,020 (the exact mathematical formula is not publicly disclosed), reflecting both the consensus market price and the slight premium observed on Hyperliquid.
Comparison Table: Spot Oracle Price vs. Mark Price ⚖️
Source of Data
Aggregated CEX prices and Hyperliquid price
Combination of oracle price, Hyperliquid order book data, and external perpetual markets
Update Frequency
Every 3 seconds
Every 3 seconds
Calculation Method
Weighted median (using set weights per exchange)
Composite calculation: spot oracle price + 150-second EMA adjustment + medians from Hyperliquid and external sources
Primary Use
Funding rate calculations
Liquidations, margin calculations, TP/SL triggers, unrealized PnL computation
Additional Adjustments
N/A
Special inclusion of a 30-second EMA if one main input is missing
Margining & Contract Types 💵
Standard contracts:
USDC collateral, USDT-denominated prices - maximizes liquidity and accessibility
Quanto structure: No USDC/USDT conversion applied (USDT P&L directly in USDC)
USDC-denominated contracts:
PURR-USD, HYPE-USD - use USDC pricing due to primary Hyperliquid spot liquidity
Uniswap Perpetuals:
Isolated margin only - no cross margining or manual margin removal
Uniswap V2/V3 AMM prices converted to USDT using robust CEX oracle prices
Margin adjustment: Only through position closure (partial/full)
Anti-Manipulation Features 🛡️
Oracle independence: Spot oracle completely separate from Hyperliquid market data
Weighted medians: Resistant to single-source manipulation
Multi-component mark price: Combines internal and external data sources
EMA smoothing: Prevents short-term price manipulation from affecting critical calculations
Validator stake-weighting: Ensures oracle integrity through economic incentives
This robust oracle architecture ensures fair, transparent, and manipulation-resistant price discovery for all Hyperliquid trading operations.
Resources
Last updated