Hyperliquid: $100 Million Revenue per Employee
Hyperliquid has become the leading decentralized perpetual-futures exchange within the last year, surpassing rivals in both performance and market share. Since our December 2024 review, key metrics have multiplied:
Daily trading volume rose from low single-digit billions to more than USD 10 billion on average, peaking above USD 30 billion.
Open interest climbed from a few billion to roughly USD 15 billion.
HYPE market capitalization expanded from about USD 4 billion fully diluted value at launch to over USD 16 billion circulating (≈ USD 48 billion FDV) by 28 August 2025.
Infrastructure now settles trades in under one-second finality (median ≈ 0.2 s) while handling roughly 200,000 orders per second.
Buybacks have removed roughly 8 to 9 percent of total token supply, underpinning a fifteen-fold price increase since launch.
Market share in decentralized perpetuals trading has leapt to roughly 75 to 80 percent, far ahead of dYdX, GMX, and other competitors.
Hyperliquid now sets the benchmark for on-chain trading speed and liquidity. New challengers, such as order-book DEXs on Sei or hybrid models, are emerging, and the first major token unlock in November 2025 is attracting scrutiny, yet Hyperliquid still holds a commanding lead. In addition, Hyperliquid ranks first worldwide in revenue efficiency, generating USD 102.4 million per employee with a team of just 12. For comparison: Tether, USD 93 million; OnlyFans, USD 37.6 million; Nvidia, USD 3.6 million; Apple, USD 2.4 million; Meta, USD 2.2 million.
Core Principles
In contrast to VC-driven projects with locked-up allocations and insider-friendly token economics, Hyperliquid reserved 70% of $HYPE’s total supply for the community. No private investment rounds diluted user interests, ensuring distribution remains broad and egalitarian. By eschewing external capital, Hyperliquid aligns incentives with its user base and encourages organic network effects.
The Hyperliquid Labs team, a core contributor to the growth of Hyperliquid, is led by co-founders Jeff and iliensinc, both Harvard classmates. The team includes members from prestigious institutions such as Caltech and MIT, with professional experience spanning leading firms like Airtable, Citadel, Hudson River Trading, and Nuro.
Initially focused on proprietary crypto market-making in 2020, the team shifted to DeFi in 2022, driven by frustration with the inefficiencies of existing platforms, including poor market design, subpar technology, and clunky user interfaces. Leveraging their expertise, the team aims to bridge the gap between decentralized and centralized trading experiences by building a seamless, efficient, and user-friendly product.
Updates since our last paper
Hyperliquid’s rapid ascent began with the 29 Nov 2024 “genesis” airdrop that distributed 31 % of the one-billion HYPE supply, valued at roughly 1.5 billion USD. The token opened near 4.8 USD, trading volumes hit 165 million USD in the first hour, and by mid-December the exchange was already clearing a few billion dollars per day with deposits near the one-billion mark. Alongside the launch, HyperEVM, an EVM-compatible layer sharing state with the core order-book chain, went live and laid the groundwork for future DeFi applications.
Early 2025 brought structural growth. In January the on-chain Assistance Fund started using trading fees to buy back HYPE, validators expanded from five to sixteen, and Asian marketing drives highlighted the platform’s sub-second execution. HyperEVM formally activated on 18 Feb 2025, giving Solidity contracts the same fast finality as trading. March added validator voting on delistings, total exchange volume passed the one-trillion-dollar mark, and whales demonstrated that nine-figure positions could be managed on-chain.
During April and May, monthly turnover topped 200 billion USD, exceeding Robinhood’s equity flow, while Builder Codes let external front-ends tap Hyperliquid liquidity for a fee share. In June the testnet of HIP-3 opened permissionless market creation, and the validator set reached its planned twenty-one independent nodes. New listings followed, from wrapped BTC and ETH to meme coins and the SPX perpetual index.
July proved decisive: a record 330 billion USD in monthly volume, open interest above 15 billion, and Hyperliquid’s share of decentralized perp trading above seventy-five percent. Media coverage highlighted revenue of roughly 1.17 billion USD annualized produced by only eleven core contributors. In August, BitMEX founder Arthur Hayes predicted outsized upside for HYPE, daily volume touched 12 billion USD, and the token printed a new high near 50 USD. Institutional signals arrived as BitGo and Phantom added custody and trading support, and by 28 Aug 2025 the Assistance Fund had accumulated almost thirty million HYPE while staking participation reached forty-three percent of supply. The team now focuses on bringing HIP-3 to mainnet, further open-sourcing, and preparing for the first large team-treasury unlock scheduled for November.
Architecture and Performance
Hyperliquid runs on a purpose-built Layer 1 that merges a high-speed order-book chain, HyperCore, with an embedded EVM runtime called HyperEVM. HyperCore handles all trades, funding payments and liquidations on-chain with a tendermint-style consensus that rotates twenty-one top-staked validators. Blocks close in roughly one second, trades confirm in about two-tenths of a second, and the matching engine has been load-tested above two-hundred-thousand orders per second. Users pay no gas; fees are volume-based and fund both validator rewards and an on-chain buy-back treasury.
Throughput headroom remains large because each market’s order book is sharded in memory across validators, so even the busiest day of about thirty-billion dollars volume used only a fraction of CPU capacity. Horizontal scaling with additional matching engines is possible but not yet needed. Finality is immediate after two-thirds validator signatures, giving central-exchange-like execution without the risk of chain reorganisations. Deterministic price-time priority and a single global order book minimise MEV; no exploit has been observed.
Risk management is on-chain and cross-margined. Positions are checked every block against oracle prices, with liquidations open to anyone willing to take the other side. A small protocol-run liquidator pool backstops the process. Oracles combine internal trade-weighted prices with external feeds such as Chainlink or Pyth, and permissionless market deployers must post a one-million HYPE bond that can be slashed for faulty data.
HyperEVM sits inside each HyperCore block, sharing the same validators and state, so smart contracts talk to the order book without cross-chain latency. Gas is paid in HYPE, block gas limits are conservative and throughput is akin to a fast sidechain, easily handling typical DeFi workloads. Assets reach the network through HyBridge, which has remained exploit-free after multiple audits. A delegated “session key” lets traders place orders without constant wallet signatures, enabling high-frequency strategies.
Network resilience has been solid, with no outages and tolerance for up to one-third faulty validators. Since launch the validator set has widened from five foundation nodes to twenty-one independent operators, and about forty-three percent of HYPE is staked. Governance is still validator-centric, code for the matching engine is partly closed, and full open-sourcing is promised once further battle-testing is complete.
HYPE Tokenomics
Supply & current float
Hard cap: 1 B tokens, no inflation.
Airdrop at launch (Nov 2024): 31 % ≈ 310 M.
Circulating as of 28 Aug 2025: 334 M (33.4 %). Slight rise from the airdrop reflects small liquidity-mining releases or late claims.
Locked: 666 M (66.6 %) split roughly into 388.8 M future rewards and 300 M team / treasury.
Unlock schedule
Team / treasury block (≈ 300 M) is expected to start vesting after a one-year cliff on 29 Nov 2025, then release over 2–3 years.
Future-rewards pool drips out via incentive programs at the foundation’s discretion.
Staking & trader utility
Unlike many L1 tokens, HYPE did not initially have a staking yield for holders because the first validators were foundation-run. In Jan 2025, as validators expanded, HYPE staking rewards became a factor: validators earn a portion of trading fees as rewards, which they share with their delegators. Hyperliquid directs a fixed % of protocol fees to stakers/validators (the exact share is not officially stated, but likely a small portion since most fees go to buybacks). Validators compete by offering a high commission to delegators. As of Aug 2025, staking yield is not publicly disclosed but inferred to be relatively low (since $1B+ fees mostly go to buybacks, not payouts). The bigger incentive to stake is governance power and the prospect of HYPE appreciation.
Fees, buybacks, burns
Trading fees are paid mainly in USDC.
HIP-1 converts any spot-market fees taken in tokens to HYPE and burns them – around 360 k HYPE burned so far.
Most protocol fees flow to the on-chain Assistance Fund (AF). The AF spends USDC to market-buy HYPE, then holds it. Balance on 27 Aug 2025: 29.8 M HYPE, effectively reducing the liquid float to ~304 M.
Outlook & risks
No ongoing emissions beyond the locked allocations, so circulating supply grows slowly until the Nov 2025 unlock.
Upcoming team unlocks could add up to ~25 M tokens per month in 2026. Hyperliquid’s buyback pace (≈ $100 M per month) will need to match any selling to avoid price pressure.
The AF’s growing treasury plus high staking ratio help keep market supply tight, but transparency on the exact vesting schedule would improve visibility.
Liquidity as a Service: The “AWS of Liquidity” Concept
AWS transformed computing by abstracting away bare-metal servers, allowing developers to focus on business logic rather than infrastructure. Hyperliquid applies a similar logic to liquidity. Developers who build on Hyperliquid need not reconstruct order book logic, liquidity incentives, or complex backend systems. Instead, they access a ready-made liquidity pool and order matching infrastructure at the protocol level.
Builder Codes are a key innovation. They enable third-party developers, regional exchanges, and specialized front-ends to integrate directly with Hyperliquid’s order book. Each builder can define its own fee structure and user interface without custodial control over user funds. This model dramatically lowers the barrier to entry, allowing a thousand exchanges to bloom—each tailored to specific user demographics, geographies, or trading strategies, but all leveraging Hyperliquid’s deep liquidity.
EVM compatibility ensures that lending protocols, stablecoins, yield aggregators, and insurance products can interoperate with Hyperliquid’s liquidity engine. This composability creates a “financial Lego” environment, where advanced strategies like automated hedging, cross-margining, and yield optimization emerge naturally. Over time, this could spawn a new class of DeFi primitives that rely on liquid, real-time markets at their core.
Growth and Distribution
Hyperliquid’s expansion rests on a few well-timed programs:
Early points campaign (Jun–Nov 2023): hefty volume-based points later became the HYPE airdrop, attracting ~50 000 active addresses.
Referral and fee discounts (from Dec 2024): 4 % lower fees for new traders and 10 % of taker fees for referrers motivated influencers, Telegram groups, and Discord servers to push traffic.
Builder codes (mid-2025): teams that embed Hyperliquid in custom wallets, analytics dashboards, or bots keep 20–50 % of generated fees. Integrations in Phantom and other frontends quietly funnel users onto the exchange.
Alongside these incentives, community-led trading contests, research reports, and endorsements from names like Arthur Hayes helped win over quant funds and prop shops. Regional hubs in China, Korea, and Turkey amplified the story, especially where on-chain venues face fewer regulatory headaches than offshore CEXs.
User Base and Behavior
640 k unique trading addresses as of Aug 2025 (≈200 k real users) versus 200 k a year earlier.
Monthly active traders are estimated in the tens of thousands, with whales accounting for most volume; Pareto effects dominate.
Retail interest shows up in 400 k+ spot USDC holders on HyperCore, but many hold only stables or farm yields rather than trade.
Volumes are growing faster than the user count, implying strong retention and heavier use by existing pros. The main challenge now is turning the large pool of passive retail wallets into active traders without diluting the platform’s professional edge.
Hyperliquid scaled quickly by rewarding early risk-takers, outsourcing acquisition to referrers and builders, and staying visible through expert commentary rather than flashy marketing. To unlock the next leg of growth, it will need simpler experiences or campaigns that persuade its sizable passive audience to place real trades.
Competitive Landscape
The decentralized perpetuals exchange space in 2025 is crowded, but Hyperliquid stands clearly at the top in volume and market share. Still, it’s instructive to compare it against key competitors in both DeFi and CeFi.
Hyperliquid leads on on-chain metrics (volumes ~5× larger than closest on-chain rival Drift, ~4× dYdX’s). Its liquidity (spread/depth) is best among DEXs, facilitated by being on a custom chain with robust MM support. Only Binance and OKX have more raw liquidity, but Hyperliquid is narrowing that gap for top assets. In terms of finality, Hyperliquid and Solana (Drift) are fastest on-chain; others like Arbitrum-based have slight delays.
Decentralization: GMX and Vertex rely on Arbitrum (which has central elements), dYdX has an independent chain (with ~16 validators – somewhat similar to HL’s early stage). Hyperliquid’s 21 validators put it in line with some mid-sized POS chains (e.g. Tron has 27 SRs). It’s not as decentralized as Ethereum or Solana yet, but more than any competitor except maybe Drift (which inherits Solana’s large set).
Token Sinks/Value: Hyperliquid uniquely has a direct revenue-to-buyback mechanism benefiting HYPE. dYdX’s token is only for governance and had high emissions (they ended rewards, but no fee burn – fees go to a community treasury not currently redistributed). GMX has a strong token model (real yield in ETH to GMX stakers), which has kept GMX popular despite lower volumes. But GMX’s growth has stagnated relative to HL because GLP yields have dropped (GLP APR ~10 % now vs earlier 20 %+, as trader losses slowed). HL’s HLP yields ~>100 % historically (not directly given to HYPE holders, but one could simulate by also being in HLP).
New Entrants (like GTE): GTE is a new on-chain trading venue on MegaETH that bundles a launchpad, AMM, central limit order book, and a price aggregator, positioning it directly against Hyperliquid’s perps-first model. It markets CEX-level performance with fully non-custodial execution, citing 100k orders per second at ~1 ms latency and venue-scanning to route orders to best prices. In June 2025 it raised a $15 million Series A led by Paradigm, with backing from top market makers and funds, and is currently live on testnet. Strategically, GTE competes by owning the full token lifecycle from launch to spot to leverage, while Hyperliquid focuses on high-performance perpetuals with an established liquidity moat.
Centralized vs Decentralized: A couple of comparisons to highlight:
Volume: Hyperliquid’s ~$10 B daily (Aug avg) vs Binance’s ~$80 B. HL ~12 % of Binance futures volume. That is remarkable, given Binance’s multi-year head start and wider asset offerings. HL’s trend is upward, so that share could increase. If crypto markets enter a strong bull, Binance might do $200 B days, but HL could see $30 B days regularly (as Coindesk said “rivaling largest CEXs” on some days).
Users: Hyperliquid’s user count (few hundred k) is dwarfed by Binance’s ~150 M registered. But DeFi is more about quality of volume than quantity of users at times – a handful of quant firms on HL can produce volumes equal to millions of retail on Binance. But for a sustainable ecosystem, growing the user base remains key (especially for decentralization – thousands of small traders means less volume concentration risk).
Feature sets: Many CEX features (stop-loss, take-profit, advanced order types) are already in Hyperliquid. It also offers one-click trading. In contrast, dYdX and others have had limited order types historically (dYdX v4 improving it, but HL was ahead in offering e.g. bracket orders from day one). GMX still lacks advanced orders (besides basic stop losses as separate triggers). This sophisticated feature parity with CEX is another reason quant traders felt comfortable migrating to HL.
Competitive Moats and Risks:
Hyperliquid’s moat: liquidity + performance + now brand recognition in DeFi. Also, being its own chain means if a competitor wanted to replicate it, they’d either need to start a similar chain (hard to bootstrap liquidity now that HL is established) or try to do it on an L2 (but then performance suffers).
However, competition from CeFi: Binance isn’t standing still. It has launched some DeFi-like initiatives, but nothing serious in perps. If Binance felt threatened, it could, for example, list HYPE and maybe co-opt HL’s liquidity by offering cross-listing (just speculation). Or more likely, Binance could reduce fees even more or run promotions to keep traders on their platform.
Competition from other DeFi primitives: Another competitor category is things like protocol-controlled perps (e.g. Perennial offers a different design with long-lived pools instead of orderbooks, aiming for low slippage). So far, none have challenged HL in volume. But new ideas like “RFQ-style perps” or “pure on-chain HFT via L3s” could emerge.
dYdX vs Hyperliquid: dYdX is the closest conceptually (custom chain orderbook perps). It’s behind in volume and arguably tech (dYdX chain is cosmos-based with 2 s blocks, slower finality, plus their tokenomics – heavy inflation to stakers – is considered weaker). However, dYdX has a strong brand and regulatory goodwill (they geofenced US from day one, which might pay off in being seen more compliant). If dYdX finds a second wind (perhaps by pivoting token value accrual or adding features like spot trading or running on multiple providers), it could regain share. But for now, dYdX’s market share plummeted from ~60 % in early 2024 to <20 % while HL took >75 % – a dramatic shift.
CeFi crackdown/regulatory arbitrage: If CEXs get hamstrung by regulations (like requiring KYC or leverage limits), traders might move more to DEXs like HL for freedom. That’s partially already happening due to US and EU derivative restrictions. This macro tailwind benefits HL. Conversely, if regulators target DeFi perps specifically, HL as the biggest would be a prime target (see next section).
In summary, Hyperliquid is currently out-competing its DeFi rivals on all key metrics and is chipping away at centralized incumbents. The competitors listed (dYdX, GMX, etc.) each have unique strengths (e.g. GMX’s community and token yield, dYdX’s prior user base), but none match Hyperliquid’s combination of speed, liquidity, and aggressive token buybacks. The competitive landscape likely cements into Hyperliquid as the leader, dYdX and a few others sharing a distant second tier, unless a new paradigm (like truly on-chain orderbook on Ethereum L2 if it became feasible) emerges. Hyperliquid’s job is to maintain this lead – which means continue technical optimization, keep market makers happy, and navigate regulatory waters safely so users have confidence to stick with it.
Key Metrics
Sources
Cover Artwork
The Consequences of War
Peter Paul Rubens, c. 1638
Risk Disclaimer:
insights4.vc and its newsletter provide research and information for educational purposes only and should not be taken as any form of professional advice. We do not advocate for any investment actions, including buying, selling, or holding digital assets.
The content reflects only the writer's views and not financial advice. Please conduct your own due diligence before engaging with cryptocurrencies, DeFi, NFTs, Web 3 or related technologies, as they carry high risks and values can fluctuate significantly.
Note: This research paper is not sponsored by any of the mentioned companies.
Hype it up