Jeff Yan
Jeffrey Yan is the co-founder and chief executive officer of Hyperliquid, the largest decentralized perpetual futures exchange by trading volume, whose custom-built Layer-1 blockchain processes tens of billions of dollars in derivatives volume daily and whose 11-person team generated more than $900 million in profit in 2025 without accepting a dollar of venture capital.[1] Background[edit]Early Life[edit]Yan grew up in Redwood Shores, California—a densely technical corridor between San Francisco and Palo Alto where Oracle's headquarters cast a long shadow over the neighborhood and where the children of engineers constituted his peer group. His parents, both Chinese immigrants, divorced when he was in third grade, and he was raised by his mother, an accountant who worked long hours each tax season. Yan has described his childhood as relatively unstructured by Silicon Valley standards; no one pushed him toward competition until a middle school friend recruited him for a math tournament in eighth grade. That first contest, which rewarded not computation but proof-construction—the art of demonstrating why something must be true—proved transformative.[2] In the summer following that tournament, Yan taught himself competition mathematics without a tutor, working through past exam papers each morning. Within a year he had made the U.S. Mathematical Olympiad training team, placing among the top 50 high school students in the country. His interest shifted toward physics after he was eliminated in an intermediate round of mathematics selection at age 16; to fill what he described as a "void," he found upper-level physics textbooks, discovered the Feynman Lectures, and within a year became one of the five highest-ranked young physicists in the United States. In 2012 he represented the United States at the International Physics Olympiad (IPhO) in Tartu, Estonia, and won a silver medal. The following year, in Copenhagen, Denmark, he returned and won a gold medal, ranking 24th in the world among thousands of competitors. Hudson River Trading[edit]After graduating from Harvard in 2017, Yan joined Hudson River Trading (HRT) as an algorithm developer, assigned to the U.S. equities team. HRT is one of the most successful proprietary trading firms in the world, known for its rigorous quantitative methods and infrastructure investment in electronic trading and market microstructure. The firm recruits selectively from top mathematics, physics, and computer science programs, and the environment gave Yan direct exposure to the mechanics of price discovery, liquidity provision, and high-frequency order flow at scale. Within weeks of joining, Yan's manager noticed an unusual dynamic: rather than arriving with questions, Yan arrived with ideas. He moved quickly through the technical ramp-up that typically occupied new hires for months, improving existing systems with an efficiency that his manager later described as rare even by HRT's standards. Yan himself has acknowledged finding the work technically satisfying but ultimately disconnected from the question that preoccupied him—what value does a marginal improvement to an already excellent system add to the world? Eight months after joining, in April 2018, he submitted his resignation. His manager's farewell email, by the accounts of those familiar with HRT's culture, was notably warm.[3] Yan has cited his time at HRT as foundational to the architecture of Hyperliquid. The firm's approach to latency, match engine design, and the discipline of competitive market-making shaped his intuitions about what a performant, fully on-chain order book would require—and how far existing blockchain infrastructure fell short of that standard.[4] Deaux (Prediction Market Project, 2018–2019)[edit]Upon leaving HRT in April 2018, Yan moved to San Francisco with his college friend and co-developer Brian Wong, also a former HRT employee, to build a prediction market platform called Deaux. The project allowed users to wager on the outcomes of real-world events—elections, weather, sporting contests—using a blockchain-based architecture Yan believed to be novel: off-chain matching with on-chain settlement, with user funds held in smart contracts governed entirely by code rather than a corporate custodian. He and Wong developed Deaux at Binance Labs' first incubator class in San Francisco.[5] Deaux attracted approximately 100 users before closing. The project failed for two reasons Yan has since acknowledged clearly: the broader cryptocurrency market entered a severe downturn in 2018, eliminating speculative appetite for experimental applications, and the team had not adequately accounted for the regulatory complexity that would later consume competitors in the same space. Kalshi, founded in 2019 and navigating roughly the same product category, spent three years in regulatory proceedings before launching. Polymarket followed in 2020. Yan returned most of the $450,000 he had raised. Scott Wu, his Harvard classmate and future co-founder of Cognition AI, was among Deaux's five most active users.[6] Chameleon Trading (2020–2022)[edit]After his HRT non-compete expired at the end of 2019, Yan relocated to Puerto Rico, drawn by the territory's advantageous capital gains tax treatment. He had $10,000 in savings and a hypothesis that cryptocurrencies offered structural edges unavailable in traditional markets. Where HRT required connecting to 13 exchanges across three New Jersey data centers and millions of dollars in infrastructure to trade U.S. equities, crypto's open APIs meant a single server on Amazon Web Services could access the same liquidity pools as institutional counterparts. Yan began writing Python scripts to trade automatically on centralized exchanges, monitoring and adjusting algorithms 14 or more hours a day, running his operation on a television commandeered from his living room.[7] Starting from $10,000 in late 2019, Yan grew the operation at a rate of several thousand percent annually for roughly two and a half years, eventually operating what multiple market observers described as one of the largest anonymous trading operations in the crypto industry at the time. The trading handle "Chameleon" was derived from a gaming alias he had used in school. In 2021, recognizing the operation had grown beyond a single-person setup, he formally incorporated Chameleon Trading and brought in iliensinc—a Harvard classmate and former venture capital strategist who had worked across Tokyo, Seoul, and Hong Kong—to manage operations, institutional relationships, and team coordination. Under her stewardship, the firm grew to six employees by 2022 and established itself among the tier of anonymous market-making firms that sit beneath the largest players—Jump Trading, Tower Research Capital, HRT, and Jane Street—but operate with sufficient scale to matter to exchange liquidity.[8] The collapse of the FTX exchange in November 2022 was the proximate catalyst for Chameleon's closure. FTX, valued at $32 billion, had been covertly lending user deposits to its affiliated trading firm, Alameda Research. When withdrawals accelerated, the shortfall became public and the exchange collapsed in nine days. The event confirmed for Yan a conviction he had been forming through 2022: that four years of rapid growth in cryptocurrencies had produced exchanges that were structurally indistinguishable from the custodial intermediaries that decentralized finance had been designed to replace. He shut Chameleon Trading without a vote, informing his team that his decision was final. Several members disagreed; several later departed.[9] Hyperliquid (2022–present)[edit]Yan began designing Hyperliquid in late 2022 alongside iliensinc and a small group of engineers, most of them recruited from mathematics and science Olympiad programs. His co-founder, known by the pseudonym iliensinc, had been his classmate at Harvard's Pforzheimer House and is a joint architect of the platform's operational strategy. The team's guiding technical premise was that no existing blockchain was fast enough to support a performant, fully on-chain order book for perpetuals trading. Having reviewed every available infrastructure option, they built a custom Layer-1 blockchain from scratch in approximately three months. Hyperliquid launched publicly at the end of February 2023. The platform's match engine runs a fully on-chain order book powered by a consensus protocol called HyperBFT, written in Rust, capable of processing approximately 100,000 transactions per second with a block time of roughly 200 milliseconds. Unlike most competing decentralized exchanges (DEXs), which use off-chain or hybrid order book architectures, every bid, ask, and fill on Hyperliquid is recorded publicly on the chain—a design choice Yan has described as foundational to the platform's claim of trustlessness. Users custody their own funds via a bridge from Arbitrum, with positions visible on-chain in real time.[10] In May 2023, to seed liquidity before independent market makers arrived, Yan deployed Chameleon Trading's strategies into an on-chain vault called the Hyperliquidity Provider (HLP). Unlike a conventional market-making arrangement—in which an exchange compensates firms in cash, equity, or token allocation for quoting—HLP allowed any user to deposit capital and receive a proportional share of trading revenue, with all positions visible on-chain. The mechanism was explicitly designed to avoid replicating the opaque fund commingling that had destroyed FTX, and its complete transparency was among the features that early users cited as a departure from industry norm. Hyperliquid declined all venture capital. In early 2024, after usage had grown sufficiently to attract fund interest, a prospective investor offered approximately $100 million at a $1 billion valuation. Yan spent a weekend deliberating and declined on Monday morning, subsequently posting four lines to his X account: "There are no investors. There are no paid market makers. The team does not charge commissions. There are no insiders." The platform's operating costs through this period were funded personally by Yan. He has cited bitcoin's value proposition as an analogy: a protocol's neutrality, in his view, depends on the absence of investors who hold preferential claims.[11] On November 29, 2024, Hyperliquid launched its native token, HYPE, via a genesis airdrop distributing 310 million tokens—31 percent of the total one-billion-token supply—to approximately 200,000 eligible addresses, with no allocation reserved for private investors, centralized exchanges, or market makers. The airdrop, valued at approximately $1.6 billion at launch and rising to roughly $4.3 billion at subsequent peak prices, was widely described as the largest by distributed value in the history of cryptocurrency markets. HYPE serves as the gas token for HyperEVM—the platform's Ethereum-compatible execution environment—and as the staking token underpinning the HyperBFT consensus protocol.[12] By 2025 and into 2026, Hyperliquid had grown to command approximately 14 percent of Binance's perpetuals market share—an unprecedented position for a decentralized platform—and was generating roughly $600 million in annualized revenue, essentially all of which was returned to token holders and liquidity providers through buybacks rather than retained by the team. The 11-person organization generated more than $900 million in profit in 2025, a figure that multiple analysts cited as among the highest revenue-per-employee ratios of any company on earth. Hyperliquid Labs, Yan's Singapore-based development entity, operates alongside the Hyper Foundation, which manages the broader ecosystem. In October 2025, Hyperliquid introduced permissionless market creation (HIP-3), enabling independent developers to launch perpetual contracts for any asset with a price source—a step Yan described as moving toward his stated goal of making Hyperliquid a foundational layer for global financial infrastructure.[13][14] Education[edit]Yan attended Harvard College, completing a combined AB/SM program in Mathematics and Computer Science in 2017—earning both a bachelor's and master's degree in four years. He enrolled in 2013, the same year he won the IPhO gold medal. At Harvard, he ranked first among roughly 150 students in Computer Science 124 (Data Structures and Algorithms), a notoriously difficult course primarily taken by upperclassmen, as a freshman. His Harvard cohort included Scott Wu—who would later co-found Cognition AI—and the person known as iliensinc, his future co-founder at Hyperliquid.[15] His practical training extended through a series of technically demanding internships. He interned at Google X, developing tooling for the autonomous vehicle project that later became Waymo. He also interned at Tower Research Capital, a proprietary trading firm, gaining early exposure to quantitative trading infrastructure, and worked part-time at Nuro, another autonomous vehicle company, during his senior year. During his junior year winter break, he and Wu were among the ten participants in Hudson River Trading's inaugural internship cohort—a three-week competitive evaluation in which Yan and Wu took the top two positions in every round. Other participants in that cohort included Alexandr Wang, who later co-founded Scale AI, and Jesse Zhang, who later co-founded Decagon.[16] Yan won a silver medal at the 2012 International Physics Olympiad in Tartu, Estonia, and a gold medal at the 2013 IPhO in Copenhagen, Denmark—representing the United States in both years and finishing 24th worldwide in 2013. He had previously qualified for the U.S. Mathematical Olympiad training team in ninth grade, placing among the top 50 high school students in the country in mathematics.[17] Controversies[edit]North Korea Security Incident (December 2024)[edit]In December 2024, Taylor Monahan, a security researcher at MetaMask, publicly identified a cluster of blockchain addresses linked to North Korean state-sponsored threat actors that had been actively trading on Hyperliquid, accumulating net losses of approximately $700,000 across ETH perpetuals positions. Monahan interpreted the activity not as a profit-seeking exercise but as operational reconnaissance—the same pattern North Korean groups had exhibited before other large-scale exploits—and expressed concern about Hyperliquid's validator architecture. At the time, the platform operated with only four validators, meaning three of the four would need to be compromised for an attacker to authorize withdrawals from the Arbitrum bridge contract, which then held approximately $2.3 billion in USDC.[18] Hyperliquid Labs responded on Discord, stating that no exploit or vulnerability had been identified and that all user funds remained secure. The platform noted it had not received notice of a specific vulnerability from any party. Monahan publicly disputed the adequacy of that response, arguing that the validator concentration represented a structural risk independent of any known active threat. The disclosure triggered a net outflow of approximately $250 million in USDC over two days—a significant but ultimately temporary withdrawal that represented a fraction of total value locked. No successful exploit occurred. The incident prompted broader industry discussion about the trade-offs between performance-optimized, lightly validated consensus architectures and the security assumptions required for platforms holding large custodial balances.[19][20] References[edit]
|
John Lothian News Feed
We visit more than 100 websites daily for financial news (Would YOU do that?). Read the John Lothian Newsletter.
| ||||||||||||||
- Sponsored Pages
- Sponsored Pages - NFA
- Pages with RSS feeds
- Professionals on Twitter
- Professionals on LinkedIn
- Hyperliquid Labs - Current Employees
- Hyperliquid Labs
- Biographies
- Cryptocurrency Exchange Executives
- Decentralized Finance
- Blockchain Entrepreneurs
- Hyperliquid
- Hyperliquid - Founders
- Hudson River Trading
- Hudson River Trading - Former Employees
- Harvard University Alumni
- Derivatives Traders
- Proprietary Trading
- Electronic Trading


