Deep Dive of Hyperliquid

Structural tailwinds in derivatives

In mature financial markets, derivatives trading volume is typically five to seven times larger than spot. In principle, crypto, with its higher leverage tolerance, should exhibit an even larger multiple. Yet under the data definition we are using, derivatives trading volume is only about 2.3× spot today.

We believe the primary reason is that crypto’s access barriers and regulatory constraints still limit institutional participation, which suppresses derivatives demand relative to what the market’s leverage appetite would imply. As market-making infrastructure improves and compliance pathways broaden, the volume gap between derivatives and spot is more likely to widen. This is also why exchanges, from a business-model perspective, often prioritize perpetuals (perps) where fee revenue scales more directly with notional exposure, and accordingly split product lines into Perps vs Spot.

On-chain, the broader “derivatives TVL” category is in a clear uptrend, and DEX market share is gradually rising, suggesting a longer-term migration path away from purely off-chain venues. These two trends together have driven many teams to build on-chain derivatives DEXs, from early pioneers like dYdX and GMX to the focal case here: Hyperliquid.

Hyperliquid perps volume, source: Defillama

This article does not revisit the well-trodden technical debate between order books and AMMs. Instead, it analyzes Hyperliquid’s success through an operational lens: how timing, marketing, product, and price reinforced each other.

Timing: the window opened by dYdX’s strategic detour

To discuss Hyperliquid as the new leader, it is useful to examine why the former leader dYdX lost momentum, despite once commanding roughly 90% of the segment’s volume.

In 2022, the industry narrative was dominated by competition between Ethereum Layer 2 approaches and “app-chain” architectures popularized by Cosmos. App chains were often discussed almost exclusively through the Cosmos lens, even as L2-based app-chain stacks like Arbitrum Orbit, Optimism OP Stack, and Starknet/StarkEx existed. At the time, some venture investors publicly pivoted hard toward Cosmos ecosystems. Against that backdrop, dYdX announced that v4 would move away from StarkEx and launch as an independent chain built with the Cosmos SDK.

The stated rationale centered on performance and UX: Ethereum’s roughly 12-second block time, plus StarkEx’s perceived performance limitations and high costs, were seen as constraints. After the migration direction was set, dYdX also adopted features such as zero fees and off-chain matching, similar to what Hyperliquid later offered.

In our view, the core issue was not the technical ambition, but the ecosystem trade-off. After moving into the Cosmos app-chain world, dYdX faced an environment with materially weaker user density, DeFi liquidity, and infrastructure. The user journey from Ethereum to Cosmos introduced additional cross-chain complexity. Market makers could not easily deploy “single-point” liquidity across venues on a Cosmos sovereign chain, liquidity depth suffered, and the chain became more of an isolated island.

At the same time, dYdX removed trading-mining incentives, which accelerated user outflows. Hyperliquid launched in November 2022, turned on trading incentives, and used the HLP mechanism to thicken the order book. With Hyperunit bringing spot access and with a smoother cross-chain UX and strong performance, Hyperliquid was positioned to capture the flow just as dYdX’s growth stalled.

dydx perps volume, source: Defillama

Operation: marketing, product, and price

Rather than starting from outcome metrics alone, it is instructive to look at Hyperliquid’s early go-to-market and product shape. We frame the execution playbook around three coordinated levers: marketing, product, and price.

Marketing: successful cold start without VC backing

A key question is how Hyperliquid achieved cold start without VC backing. A large part of the answer is a sustained, multi-year points program that turned prospective airdrop value into a continuous trading incentive.

From our mapping of perps volume against campaign timing, Hyperliquid’s breakout began around Season 1, with peak daily volume reaching about $4.3B. After Season 1 ended, the market shifted into a meme-coin dominated phase. Hyperliquid responded by launching an “official” meme distribution via PURR, keeping its growth rhythm aligned with the prevailing narrative. Campaigns had essentially no idle periods, and listing and incentive decisions tracked market regimes. The platform listed a range of meme assets (e.g., WIF, BOME, POPCAT, BRETT, MEME) and issued its native HIP-1 token PURR as a trading incentive anchor.

Product: liquidity first

Marketing often runs ahead of product completion; product iterates in the gaps between campaigns. Hyperliquid’s core product goal is to fuse off-chain efficiency with on-chain transparency.

Relative to CEXs, an on-chain order book DEX has higher transparency and stronger composability potential. But for any exchange, the first principle is unchanged: liquidity depth is everything. Hyperliquid’s early design used the HLP pool to provide baseline depth by acting as a counterparty to on-chain retail flow.

Liquidity is a positive feedback loop: better depth attracts more volume, and more volume improves depth. In cold start, many platforms must bootstrap market making internally. In CEX contexts, internal market making is often criticized as opaque or “trading against users.” DEX transparency changes the optics: self-built liquidity more naturally takes a GMX-style LP vault form. Even if the strategy logic runs partially off-chain, governance and vault accounting are visible, and retail users can participate in HLP to earn hedged returns, making the model easier to accept.

This approach is not risk-free. The JELLYJELLY episode highlighted tail-risk exposure, and under the current mechanism there are still latent vulnerabilities. Based on our observation, LP PnL annualizes around ~10%, and that yield must be evaluated against the true risk surface.

The deeper moat for a DEX is not only thick liquidity, but the ability to let that liquidity spill into a composable DeFi ecosystem. Hyperliquid is still weaker here. Founder Jeff has acknowledged that liquidity accessibility via HyperCore is not very efficient. Although HyperEVM has attracted 200+ projects, directly tapping HyperCore order-book liquidity requires a bridging component called CoreWriter. CoreWriter is non-atomic: contract calls and order-book execution cannot be finalized in the same transaction and block. Contracts therefore cannot confirm within-block whether an order state has changed successfully, which lengthens the interaction path and degrades UX.

With liquidity established, the next focus is matching performance, especially for market makers. Key order-book metrics include throughput (TPS/OPS), spreads, depth, and slippage. A simplified comparison is below:

Hyperliquid outperforms dYdX on these metrics, but still trails top centralized venues like Binance. At current on-chain user scale, performance is adequate; however, if on-chain trading becomes mainstream and user count spikes, the infrastructure may hit limits. In practice, outages have already occurred.

Price synergy: token price as marketing, tokenomics as flywheel

In crypto, “marketing cost” has a broader meaning than in traditional industries. Token price appreciation itself can function as a powerful distribution and attention mechanism. Hyperliquid’s price coordination effects are particularly important.

Price synergy requires the ability to control enough supply to influence price action. We observed some anomalous patterns: during the closed alpha campaign, certain related parties may have “farmed” points late in the program and captured outsized airdrop allocation. Because points allocation and final distribution were not fully transparent, there is some possibility of discretionary allocation that could have contributed to later supply concentration—creating favorable conditions for coordinated price dynamics.

On the other hand, the eventual airdrop design sought to distribute tokens to more “real” users. Contribution was measured by absolute P/L, discouraging pure volume farming and many forms of Sybil behavior. This design encouraged whales to hold more organically rather than immediately sell.

A further advantage was the lack of meaningful early VC inventory. In a scenario where market participants are bullish, both VCs and retail must buy on the secondary market, and there is less risk of early VC dumping. A commonly cited analog is StepN: when public sentiment is positive and early VC supply is limited, VCs and retail often purchase directly from the market without the overhang of early VC dumping.

Tokenomics became a second narrative pillar, especially the “fees → buyback → burn” flywheel. Hyperliquid enabled fees early. Roughly 93% of fees are used to buy back and burn HYPE, with the remaining ~7% routed to the HLP treasury. This created visible positive reinforcement in the early phase. As of February 10, 2026, cumulative buyback and burn exceeded 40.5M HYPE, valued at roughly $1.3B at an average buyback price near $32. For value-oriented holders, the observable linkage between fees and buybacks increases perceived predictability and can support valuation.

In short, once cold start succeeded, the lack of early VC inventory became an advantage. Secondary-market FOMO from VCs and retail, project-side inventory concentration, potential points opacity, and an explicit buyback-burn mechanism reduced circulating float and strengthened buy pressure. The resulting price rise reinforced attention and trading activity, creating another feedback loop.

Centralization in security and governance

Post-TGE, Hyperliquid also encountered several public incidents:

In the JELLY episode, Hyperliquid shut down the market and delisted the perp, then closed/settled the market at a specific price point and announced that users (except “flagged addresses”) would be made whole by the Hyper Foundation, with compensation executed based on on-chain data.

After the large ETH liquidation loss, Hyperliquid raised margin requirements to reduce systemic impact from large positions, effective shortly after the incident. Updating margin frameworks is normal and often prudent, but it highlights that the protocol’s risk constraints are not fully ossified. They can be revised by a coordinating authority on short notice, which is again a form of centralized rule-setting (even if the change is publicly announced and arguably beneficial).

In the API/frontend outage: a “DEX” trading experience can still depend on centralized service layers (API gateways, front ends, routing infrastructure). When that layer failed, Hyperliquid decided to reimburse affected users and implemented a process that included tiered criteria and KYC requirements above a threshold.

These incidents reveal a broader theme: centralization in incident response and rule-setting. As Arthur Hayes has argued, Hyperliquid’s decentralization can be performative. We partially agree. Hyperliquid decentralizes certain rules and execution details, but key hubs such as points distribution, token allocation, and HLP operations remain comparatively centralized with limited disclosure. This governance style can stop attacks quickly, but it increases the “trust budget” required and invites skepticism about decentralization claims.

From a pragmatic standpoint, this trade-off is not necessarily unreasonable today. Given weak on-chain infrastructure, frequent adversarial events, and an early-stage product, limited centralization can support stability.

What does the next era of DEX look like?

From first principles, product is the 1 and marketing is the 0. For a perps DEX, matching performance and liquidity depth are the irreducible core. Marketing and tokenomics determine how far a platform can go once it clears that baseline.

Even Hyperliquid remains behind Binance-class CEXs in raw performance, and there is room to improve. The team’s approach is pragmatic: with on-chain users still far fewer than off-chain users, current OPS is “good enough,” and more attention is placed on operational cadence.

Looking forward, the natural DEX advantage is to extend thick liquidity into a broader DeFi stack via transparency and composability. Hyperliquid’s EVM ecosystem has reportedly surpassed 200 projects, and TVL disclosures claim it is above $4B.

The core limitation is still composability efficiency. All chains and all dual-layer systems face liquidity fragmentation and the rollback risk introduced by non-atomic flows. Ethereum liquidations, for example, often span multiple blocks and require waiting for soft confirmation, making real-time settlement difficult.

Hyperliquid uses a dual execution-layer design within the same consensus domain: contracts originate on the EVM side, while core matching is processed by HyperCore in subsequent blocks, bridged by CoreWriter. The bottleneck is less about throughput and more about an atomicity gap. It is difficult to close “order → hedge → settle” within one transaction and one block. Contracts cannot confirm order outcomes immediately, which degrades UX.

Soft settlement or pre-confirmation can mitigate latency, and this is a common approach in Solana-adjacent practices and elsewhere. However, it often converges toward the same family of designs as intents, solvers, and shared sequencers: better UX at the cost of potentially evolving into low-latency races, side-channel bribery, and quasi-centralization. If a key node stalls, composability can fail along the critical path. This is an industry-wide issue, not unique to Hyperliquid.

Conclusion

The success of Hyperliquid is attributed by right timing, successful operations and appropriate tokenomics. While on-chain user scale remains limited, it delivered a gasless and smooth UX, relatively high throughput, and thick liquidity that served both HFT-style traders and retail. Marketing cadence, token incentives, and pragmatic operations amplified this advantage precisely as dYdX lost momentum.

Near term, Hyperliquid’s weaknesses are not fatal. The centralization trade-offs in governance are balanced by operational stability, and its competitive edge is likely sustainable in the medium term. The principal risk is that OPS is still not CEX-level. If a major on-chain market cycle drives exponential user growth, infrastructure limits could resurface. In that window, the market will favour platforms with higher performance, deeper liquidity, and more atomic composability.