Understanding Mark Prices in Perpetual Trading

A perpetual futures contract has no expiration date, which means there is no natural convergence point where the contract price must equal the underlying asset's spot price. To keep the perpetual price tethered to reality, exchanges use two key mechanisms: the funding rate and the mark price.

The mark price is used to calculate unrealized profit and loss (PnL) and, critically, to determine whether a position should be liquidated. It is distinct from the last traded price. While the last traded price reflects the most recent transaction on the order book, the mark price is designed to be a more stable and manipulation-resistant representation of fair value.

If the mark price deviates significantly from the true fair value, traders can be unfairly liquidated even when the broader market has not moved against them. Conversely, an inaccurate mark price can allow positions to remain open when they should have been liquidated, creating systemic risk for the exchange's insurance fund and other traders. Mark price accuracy is not a minor technical detail — it is a core component of a fair trading platform.

What Are Hyperps?

Hyperps (hyper-perpetuals) are perpetual contracts on Hyperliquid covering assets that may not have deep, established spot markets. These include pre-launch tokens, newly listed assets, and other instruments where traditional price discovery is less mature.

Because these assets often lack the deep liquidity pools and well-established oracle feeds that major cryptocurrencies enjoy, pricing them accurately presents unique challenges. The mark price for a hyperp must be responsive enough to track genuine price movements, stable enough to resist manipulation, and fair enough to protect traders from artificial liquidations.

The XPL Volatility Event

XPL, traded as a hyperp on Hyperliquid, experienced an approximately 2.5x increase in mark price over a relatively short period. This dramatic move raised immediate questions about whether the mark price was accurately reflecting the asset's true market value.

The issue was rooted in how hyperp mark prices were calculated at the time. For assets without deep external spot markets, the formula relied heavily on the Hyperliquid order book itself. While this worked under normal conditions, it became problematic during extreme volatility. In a thin market, aggressive buying could push the order book midpoint significantly higher, and the mark price would follow — even when external price references suggested the move was exaggerated.

Traders with short positions faced mark prices that may have overestimated fair value, while long traders saw unrealized PnL that might not have been realizable in the broader market. The mark price was not fulfilling its core purpose.

Community Feedback and the Engineering Response

The community response was swift and data-driven. Traders shared comparisons between the XPL hyperp mark price and external pre-launch perpetual prices on other venues. The evidence was clear: there was a meaningful divergence.

Rather than treating this as a one-off anomaly, the Hyperliquid team acknowledged the issue publicly and committed to improving the calculation. The engineering team identified the key vulnerability: the mark price formula for hyperps was not adequately incorporating external reference prices where they were available. Even when external pre-launch perpetual markets existed, the hyperp mark price was calculated primarily from internal order book data.

The Improved Mark Price Formula

The updated approach applies the normal mark price formula to hyperps, incorporating external pre-launch perpetual prices as a reference. When external markets exist for a hyperp asset, those prices are factored into the calculation alongside Hyperliquid order book data using a weighted combination.

External pre-launch perpetual prices are sampled from reference venues and combined into a composite index. This index is blended with the Hyperliquid order book midpoint using a formula that accounts for the relative liquidity and reliability of each price source. The result is a mark price that is more robust against manipulation on any single venue.

Importantly, the new formula does not simply peg the hyperp mark price to external venues. Hyperliquid's own order book data still contributes, which is appropriate since the platform often has significant liquidity and genuine price discovery. The improvement is in the balance — the mark price is no longer overly dependent on internal data when external references are available.

Why Accurate Marking Matters Beyond Individual Trades

Mark price accuracy has systemic implications. When positions are liquidated at inaccurate prices, the insurance fund absorbs preventable losses. Over time, a pattern of inaccurate liquidations can deplete the insurance fund, creating risk for all users.

Mark price accuracy also affects the funding rate mechanism. The funding rate is calculated based on the difference between the perpetual price and the mark price. If the mark price is distorted, funding rate signals become unreliable and the mechanism that keeps perpetual prices aligned with fair value stops functioning correctly.

For market makers, mark price reliability is a prerequisite for providing tight spreads. If a market maker cannot trust the mark price, they must widen spreads to account for additional risk, directly increasing trading costs for all participants.

Price Data on HyperX

HyperX displays real-time mark prices for all Hyperliquid markets including hyperps. Our TradingView charts show position entry lines and liquidation levels based on current mark prices.

The Broader Pattern

The XPL event and its resolution illustrate a characteristic pattern in Hyperliquid's development: issues are identified through real-world usage, community feedback is incorporated into the analysis, and improvements are implemented transparently. Hyperps represent an ambitious product category — offering perpetual trading on assets without mature market infrastructure elsewhere. Ambition of this kind inevitably encounters edge cases that cannot be fully anticipated.

What matters is the response. The mark price system for hyperps is now more robust, more resistant to order book manipulation, and better anchored to the broader market's consensus on fair value. For traders, the practical result is that positions are less likely to be affected by mark price distortions, and the trading experience on hyperps more closely mirrors the reliability expected from established perpetual markets.