Hyperliquid’s Vertical Evolution

When a dApp Outgrows Its Blockchain: Hyperliquid and Arbitrum

In crypto theory, the hierarchy is usually drawn intuitively: blockchains first, applications on top. dApps compete within ecosystems and inherit the limits of their base layer. The reality is more complex, and the Hyperliquid–Arbitrum relationship illustrates this shift clearly.

Hyperliquid began as an Arbitrum-settled perpetuals platform, one of many DeFi applications benefiting from cheap execution and deep liquidity. Now, Hyperliquid commands derivatives volume that rivals major centralized exchanges, with open interest consistently measured in the billions. And, it developed and runs on its own execution chain.

In the process, Hyperliquid did not meaningfully reinforce Arbitrum’s application layer, nor did it attract other perpetuals platforms to build alongside it, and the two are no longer closely associated.

This creates an uncomfortable feedback loop for L1s and L2s. Chains subsidize applications, provide liquidity and visibility, and absorb early-stage risk. The most successful applications then internalize execution, reduce dependency on shared infrastructure, and leave chains with settlement duties and diminishing strategic leverage.

In that sense, the application layer's success becomes extractive for blockchains. The better a chain is at nurturing winners, the more likely those winners are to outgrow it.

After separation, Arbitrum still served as Hyperliquid’s primary on- and off-ramp, with the majority of traders’ funds transferred to and custodied on Arbitrum. From its roughly $8–10 billion in total value locked, over $4 billion in USDC (70% of all USDC on Arbitrum) was tied directly to Hyperliquid-related flows.

However, with the launch of native USDC on Hyperliquid in late 2025, this relationship has also changed. The capital that previously relied on Arbitrum for settlement is now migrating directly to Hyperliquid’s chain. Consequently, Arbitrum is losing even its status as the "vault" for these billions in liquidity, transitioning from a necessary settlement layer to an optional bridge.

Further, Hyperliquid introduced in 2025 its own stablecoin, USDH, designed for margin efficiency and internal incentives. Issuing a venue-native stablecoin followed a familiar playbook from traditional exchanges: first control execution, then collateral flows, and eventually the unit of account itself.

dApp Driven Blockchain Development

Early blockchain development was driven by protocol performance and decentralization as first principles. Increasingly, it is driven by function. Execution, settlement, governance, and distribution are being optimized separately, and applications are reorganizing the stack around their own needs rather than those of the underlying chain. Hyperliquid’s trajectory fits this pattern, but it is far from an isolated case.

Hyperliquid is steadily building a vertically integrated venue, internalizing execution while externalizing settlement only where it remains advantageous. The logic is not ideological; it is operational. As soon as a shared blockchain becomes a bottleneck rather than a multiplier, it is bypassed.

The same dynamic played out earlier with Uniswap. Uniswap outgrew Ethereum blockspace years ago, expanded across multiple chains, and ultimately moved toward its own execution environment. Unichain is optimized less as a general-purpose blockchain and more as a proprietary execution layer tailored to Uniswap’s needs, closer in spirit to a dedicated trading backend than a neutral public chain.

Other examples reinforce the pattern. dYdX left Ethereum entirely to launch its own appchain, prioritizing control over execution and market structure. Lido became systemically important to Ethereum without binding itself to Ethereum’s execution layer, operating instead as a protocol-level liquidity coordinator. In each case, the application’s economic gravity ultimately exceeded that of the chain hosting it.

What emerges is a reversal of the original crypto hierarchy. The center of innovation has shifted upward in the stack, from protocols competing for developers to applications competing for global market share. Blockchains increasingly play a supporting role, while applications define the user experience, liquidity, and economic gravity.

This shift also invites a more fundamental question: what was the blockchain promise in the first place? Much of the complexity around decentralized data processing was engineered to support a base digital asset and a neutral execution layer. Platforms like Hyperliquid and Uniswap are clearly innovating in market structure and trading efficiency, but describing these systems as decentralized is difficult at their scale, especially when execution, custody, and control consolidate around application-specific infrastructure.

The House of All Finance

Hyperliquid’s recent evolution suggests a reframing of the problem rather than a rejection of it. Its development roadmap increasingly resembles the early construction of a blockchain, instead, it is not driven by ideology, but by accumulated user demand. The sequence of Hyperliquid Improvement Proposals (HIPs) passed over the past year reads less like incremental DeFi upgrades and more like the foundations of a new base layer.

HIP-1 introduced a native token standard alongside a spot order book, establishing a common asset layer comparable in spirit to Ethereum’s ERC standards.

HIP-2 followed with liquidity injection mechanisms for HIP-1 tokens, emphasizing the trading use case at the protocol-infrastructure level.

HIP-3 allowed developers to deploy perpetual markets directly. Participation requires a substantial economic commitment — staking 500,000 HYPE, roughly $12.5 million — and successful market creators can earn up to 50% of generated fees. This mirrors the function of smart contracts more than classic DeFi permissionlessness: execution is open, but only to those willing to bear real economic responsibility.

Hyperliquid is assembling the components of a general-purpose financial platform shaped by what traders actually demand: performance, liquidity, and composability at scale. The strongest use case for blockchain — trading — is driving design decisions here, including deliberate trade-offs around decentralization.

Whether Hyperliquid can ultimately build a true “zero-level” decentralized financial platform — a global base layer for digital assets where value cannot be double-spent — remains an open question.