Hyperliquid: The AWS of On-Chain Finance
What AWS Did for the Internet, Hyperliquid Is Doing for On-chain Finance
Every decade or so, a company rewrites an entire industry — not by launching a killer app, but by building the rails others ride on. They don’t just scale themselves; they let thousands scale on top of them. In the world of cloud computing, that was Amazon Web Services. It didn’t just host apps; it reshaped how internet businesses were built and scaled. The result? A new era of builders and a business that captured massive value by powering them.
In 2024, AWS generated over $100 billion in revenue with an operating margin of ~37%, making it Amazon’s most profitable division. At a conservative 10X earnings multiple, it would be worth over $1 trillion as a standalone business. But the numbers tell only part of the story. AWS powers the infrastructure behind modern software, from Netflix and Zoom to Stripe and Figma. It’s not just a service; it’s a tax on the compute economy.
I believe Hyperliquid is playing that role in crypto — emerging as the foundational infrastructure for on-chain finance. It started as a decentralized exchange(DEX) focused on perpetual futures, or perps—no-expiry contracts that give traders long or short exposure to an asset without owning it. (Think: futures without expiry). Since then, it has quickly grown into a high-performance blockchain platform, offering deep liquidity, ultra-low fees and native programmability at the base layer.
On-chain finance, in this context, means rebuilding the entire financial stack— not just how assets trade, but how they’re issued, held, and settled — all on blockchain rails. The goal is a transparent, decentralized system that operates 24×7 without centralized gatekeepers. Hyperliquid is rapidly emerging as the platform building the core infrastructure to make that vision come alive.
In just its third year, Hyperliquid has already generated $260 million in revenue with 97% margins, with a lean team of 10 and is on pace to reach $780 million annualized by year-end. That kind of trajectory is exceedingly rare. Only a handful of companies in history have scaled this fast, this early.
To understand what Hyperliquid is really building, it’s worth reflecting on how AWS itself came to be, and why it was built that way.
The Philosophical Origins of AWS
AWS wasn’t born from vision alone. It was born from constraint.
By the early 2000s, Amazon’s engineering teams were all bumping into the same internal constraint: to launch any new web service, they had to go through a centralized infrastructure team to provision computing and storage resources. That meant delays, friction, and teams reinventing the same infrastructure again and again.
As Brad Stone explains in The Everything Store, around that time, Jeff Bezos became obsessed with a book called Creation by Steve Grand, the developer of a 1990s computer game called Creatures. The game let players nurture a seemingly intelligent organism on their computer screens. But Bezos wasn’t just drawn to the game; it was the philosophy behind it. Grand’s approach to designing intelligent life was simple: focus on designing simple computational building blocks, called primitives, and then sit back and watch surprising behaviors emerge.
"If Amazon wanted to stimulate creativity among its developers," Stone writes, "it shouldn’t try to guess what kind of services they might want; such guesses would be based on patterns of the past. Instead, it should be creating primitives — the building blocks of computing — and then getting out of the way."
That idea changed everything.
Instead of guessing what internal teams would need, Amazon built flexible primitives: compute and storage. These became EC2 and S3 — foundational services that abstracted away the hardest parts of backend infrastructure. Developers could now move faster, experiment freely, and scale without permission.
It wasn’t just a technical breakthrough. It was a philosophical one.
Amazon saw the curve: more developers, more online-first businesses, and rising demand for scalable compute. But instead of dictating what should be built, it bet on enabling builders to figure it out. EC2 and S3 didn’t prescribe the future; they created the conditions for it.
And like all great platforms, AWS captured value by enabling others. The more builders it empowered, the more central it became.
That same principle — empower builders, abstract complexity, and let ecosystems compound —is quietly shaping another platform. But this time, it’s not just about building faster; it’s also about letting builders earn directly from the platform.
So what exactly is the infrastructure Hyperliquid is building?
Liquidity
Hyperliquid started with a focused wedge: building the most liquid perps DEX. In doing so, it seized over 75% of the decentralized perps market in just three years, outpacing rivals like dYdX and GMX. It follows a familiar pattern seen in monopoly-scale businesses: start with a small market, win fast, expand relentlessly. Amazon started with books. Facebook with Harvard. Airbnb with conference-goers. Hyperliquid did it with perps. Now it’s expanding into spot, lending, and other financial primitives—on the path to becoming the central liquidity layer for on-chain finance.
Start any on-chain market without liquidity and it dies. Simple as that. Liquidity isn’t binary; it’s a spectrum. It refers to the degree to which assets can be bought or sold without significantly moving the price. The deeper the liquidity, the more confidently market participants can transact and the more reliable and trustworthy those prices become. That matters because pricing is foundational to an efficient market. Efficient prices make everything else work better: investing, lending, risk management, and trading. Every major financial product—spot, futures, perps, stablecoins, lending protocols— rest on the assumption that the base assets are liquid.
Hyperliquid has already built the kind of deep-liquidity infrastructure that outpaces every other DEX and rivals even the largest centralized exchanges. In May 2025, it cleared $248 billion in perpetuals volume, averaging ~$7 billion a day—roughly 10% of Binance Futures, the dominant player in crypto futures. Spreads remain tight at ~0.12%, allowing large orders to move with minimal impact. Even after recent airdrops, volume held steady, suggesting the flow is real.
And it has already moved beyond perps: In just four months, Unit Protocol, Hyperliquid’s generalized tokenization layer, has driven $6 billion in spot-market volume on Hyperliquid. Over $390 million in assets have been bridged from their native chains (BTC, ETH and SOL) into Hyperlquid. If Unit succeeds in tokenizing traditional assets next, Hyperliquid could become the venue where deep liquidity meets every asset class — on-chain, permissionless, and global.
Liquidity is to financial markets what users are to social networks. The more there is, the better it works, and the stronger the pull for others to join. That’s why “liquidity begets liquidity” isn’t just a phrase—it’s a moat. With over half a million users and ultra-low fees, Hyperliquid gives traders little reason to go anywhere else. If you want the best prices, you go where the liquidity lives. Once a platform becomes the place where markets happen, it becomes the habit.
Hyperliquid didn’t just attract liquidity; it engineered the infrastructure for it. A transparent, low-latency on-chain order book processes 200,000 orders per second with one-block finality, backed by a robust pricing engine that updates every 3 seconds using CEX-blended oracles. It opened the rails to market makers—no permission needed, public SDKs, continuous rebates—and bootstrapped depth from day one through HLP vaults, where anyone can passively provide liquidity by depositing USDC. The result: a system where both pros and the crowd can plug in instantly, and liquidity never has to wait.
But while deep liquidity is the foundation Hyperliquid nailed early and the moat others now struggle to cross, it’s not the secret that truly sets it apart. That comes next.
The Secret Is in the Stack
Some of the greatest businesses in history are built around a secret—a hidden truth that others overlooked or dismissed. Airbnb is a classic example: the idea that people would let strangers sleep in their homes seemed absurd until it became obvious. Secrets are powerful because they let you build in the white space, where competition is low and leverage is high. As Peter Thiel puts it: great businesses don’t compete; they escape competition. Hyperliquid has one of those secrets. Everything we’ve discussed so far—revenue, growth, liquidity —is impressive. But the real edge lies deeper, in a structural choice most protocols never attempted.
Hyperliquid’s secret? It fused an exchange—the most profitable business in crypto—with a high-performance layer-1 blockchain. It may sound obvious in hindsight but it was counterintuitive, technically daunting, and almost no one tried it. Most teams picked one lane: they built exchanges or they built chains but rarely both. And the few that attempted both kept them siloed. Binance launched BNB Chain. Coinbase built Base. FTX paired with Serum. In each case, liquidity (from the exchange) and programmability (from the chain) lived in separate worlds-divided by trust assumptions, latency issues, and architectural friction.
Hyperliquid flipped that script. It combined the two most proven products in crypto into a single, vertically integrated system: HyperCore, its built-in exchange engine, and HyperEVM, its smart contract chain. The first-order result? Liquidity becomes programmable. The second-order effect? Programmable liquidity opens up a radically new design space where on-chain apps can tap directly into deep liquidity pools without sacrificing speed, trust, or composability.
This fusion isn’t additive; it’s multiplicative. It’s the first time in crypto that deep liquidity and programmable logic have shared a native execution layer. Much like how Apple fused hardware and software to redefine the smartphone, Hyperliquid fuses liquidity and programmability to redefine what’s possible in on-chain finance.
Programmable Liquidity in Action
It’s one thing to claim liquidity is programmable. But how does that work—concretely, at the code level? Enter precompiles and builder codes.
Programmable liquidity starts with how Hyperliquid is architected. At its foundation are two parallel systems: HyperCore, a high-speed trading engine that runs a central limit order book (CLOB), and HyperEVM, a general-purpose smart contract layer. They run side by side, but crucially, they share the same global state, maintained by the HyperBFT consensus algorithm. That shared state enables deep composability: contracts on HyperEVM can read from and write to live trading logic in HyperCore.
At the heart of Hyperliquid’s programmable liquidity are two foundational primitives: precompiles and builder codes. Precompiles are native system functions that smart contracts use to interact directly with HyperCore—place trades, manage orders, pull market data. Builder codes are the strategies and applications developers write using those functions.
Think of it like a gaming console:
Precompiles are the built-in buttons—jump, run, shoot.
Builder codes are the custom game logic—what happens when Mario jumps or collects coins.
Put together: precompiles are the buttons, builder codes are the game. Together, they let developers build live strategies and products — directly on liquidity.
Most platforms stop at programmability. Hyperliquid goes a step further; it wires in monetization.
Monetized Composability
In the crypto world, composability is often celebrated: developers seamlessly build on top of protocols like Lego bricks to create products more valuable than the sum of their parts. But when it comes to capturing value from what they've composably built, the mechanisms have historically been ad hoc, fragmented, and limited.
Early DeFi platforms like Yearn Finance offered a glimpse of what's possible. Strategists could deploy yield-generating smart contracts that others could use and earn a share of the fees. But these systems were narrow in scope, confined mostly to yield farming and lending, and lacked scalability. So while composability has always been a powerful ideology, scalable value capture from it has remained elusive.
Not with builder codes.
With builder codes, developers can build strategies, products, or interfaces that route orders through Hyperliquid and earn fees based on the volume they drive. This isn’t a theoretical construct; it’s embedded in the protocol. Every time a builder’s code executes a trade via HyperCore, it can programmatically claim a portion of that trade similar to how a SaaS developer might charge for API calls. The fee, set by the builder, is deducted directly from the trade flow.
That unlocks a whole new world of monetizable products. Builders now have the freedom to go beyond passive vaults and narrowly defined strategies. With Hyperliquid, they can launch entire products with built-in revenue streams.
Consider three possibilities:
A regional exchange in Southeast Asia focuses on what it’s uniquely suited for—licensing, local marketing, and customer support—while using Hyperliquid for trading infrastructure. No need to bootstrap liquidity; they simply add a spread and route trades via builder code.
A mobile wallet adds a swap feature powered by Hyperliquid’s liquidity. The wallet owns UX, custody, and mobile optimization—what it does best—while the builder code handles execution and earns a fee on each swap.
A social trading app lets users mirror the strategies of prominent traders. A builder code replicates the trader’s positions in real time, routing everything through Hyperliquid. As volume grows, so do the creator’s earnings.
In traditional finance, major brokerage firms like Interactive Brokers and Robinhood operate similarly. They manage customer acquisition, retention, support, and compliance but don't run the exchanges themselves. Instead, they route user orders to third-party liquidity venues—exchanges or market makers—and earn a cut on the transaction. Hyperliquid brings that model on-chain, at scale, with enough granularity for builders to create everything from simple trading strategies to multi-billion dollar businesses.
By giving builder codes baked-in monetization, powered by precompiles and backed by deep liquidity at order-book scale, Hyperliquid quietly invokes Charlie Munger’s line: “Show me the incentives, and I’ll show you the outcome.” Hyperliquid is redefining how builders can earn, combining its deep liquidity with the expansive design space of programmable finance.
It’s already catalyzing a Cambrian explosion of strategies, applications, and businesses—all tapping into Hyperliquid’s liquidity engine. We’ll see a microcosm of capitalism and competition unfold in real time. Many experiments will fail. But enough will succeed — emerging as the billion dollar businesses of on-chain finance. And through that Darwinism, the ecosystem becomes richer and more robust.
This is exactly what AWS figured out two decades ago. EC2 and S3 were foundational computing primitives but their real power was economic: they let anyone build, scale, and monetize in the cloud. Hyperliquid’s precompiles and builder codes are playing that same role in on-chain finance: low-level primitives that abstract away infrastructure while enabling builders to create new products, new businesses, and new revenue models.
While builder codes are designed to natively capture value for builders, Hyperliquid’s approach to value capture extends all the way up to the protocol layer, where it captures value far more effectively than most Layer 1 blockchains.
Not Just Value Creation, But Value Capture
The chart above tells an important story: Hyperliquid, barely three years old, tops the list of all Layer-1 blockchains by gross profit in May 2025, outpacing long-established platforms like Tron, Ethereum, and Solana. The insight here is simple but critical: the best businesses don’t just create value; they capture a fair share of it. Hyperliquid makes that possible by design. While most L1s rely on gas fees — a model often at odds with usability — Hyperliquid embeds a CLOB at the protocol layer, giving it a direct claim on economic activity through volume-based fees. Since all trades clear through that same engine, Hyperliquid captures a predictable share of every transaction as revenue. And because the CLOB is the gravitational center—from spot markets to builder codes—every participant is incentivized to route volume through it, fueling a flywheel where volume begets revenue.
Crypto has long debated where value accrues: at the protocol layer (“Fat protocol thesis”) or the application layer (“Fat app thesis”). But a more useful lens may be to move beyond categories and ask a simpler question: regardless of whether it's a protocol or an app, is there a direct and durable mechanism to capture the value being created? In many cases, value-capture mechanics matter more than stack position. Many protocols and apps create real utility but without capture mechanics, that value often leaks away or relies entirely on speculative reflexivity. Stack position alone isn’t enough; it amplifies capture only when the mechanism exists. Hyperliquid stands out because it combines both: strategic position and native, scalable value capture.
It’s clear that Hyperliquid captures a fair share of the value it creates. But how does that translate into HYPE, its native token? Here’s how: Since launching the token in late 2024, the protocol has directed 97% of daily revenue, roughly $1 to $2 million per day, toward token buybacks, accumulating over 24 million HYPE, now worth over a billion dollars. That makes it one of the most consistent , large scale buyback programs in modern finance, executed daily and directly at the protocol level. On top of that, HYPE staking offers tiered trading fee discounts and governance access, and the new HIP‑3 proposal requires staking 1 million HYPE to launch permissionless perpetual markets, further locking capital and creating structural buy pressure. In short, Hyperliquid’s design ensures that as the protocol earns more, more HYPE is bought and more demand is created for holding it. That doesn’t guarantee investor returns, but it tightly links protocol success with token strength.
But even the best design means little without a team that can execute—relentlessly, precisely, and at scale. That, I believe, is Hyperliquid’s greatest edge.
The Edge You Can’t Copy
Some founding teams are like the proverbial 100x engineer — their impact isn’t just 100% better than average, it’s 100x. Jeff Yan and the team behind Hyperliquid are one such case. A Physics Olympiad gold and silver medalist, Harvard graduate, and former quant at Hudson River Trading, one of the world’s top high-frequency trading firms, Jeff didn’t stumble into this space. He earned his stripes by building one of crypto’s most dominant market-making operations in 2021. What followed was Hyperliquid. With an elite team of 10 builders, led by Jeff and his co-founder Iliensinc (also a Harvard classmate), they have executed with a level of pace and precision rarely seen in crypto. They ship faster and cleaner than teams fifty times their size. In just three years, they’ve built the most performant perp DEX in the world—with zero VC funding — earning monopoly-scale market share, shipping relentlessly, and advancing a broader vision for on-chain financial infrastructure. Every product decision—from latency to UX to protocol design—reflects a depth of thinking rarely seen at this scale. If past execution is any indication, this team is not just capable of delivering on the vision, they're likely to outpace anyone else attempting it. Teams like this are rare. When you see one, you go full Druckenmiller: you bet, and you bet with size.
Timing the Curve
Writing about AWS in 2005 would’ve sounded bold—maybe even reckless. It was just three years old. Cloud computing wasn’t mainstream. Startups still bought their own servers. You’d have to believe the internet would keep growing. That developers would build online-first businesses. That Amazon — a bookstore then — would out-execute every tech incumbent. That nobody else would catch up in time.
That’s exactly the kind of bet this thesis asks of you.
To believe Hyperliquid will become the AWS of on-chain finance, you’d have to believe a few things. That finance is going on-chain. That more assets, public and private, will be tokenized. That the regulatory fog will lift. That builders will choose programmable liquidity over static rails. And most importantly, that the Hyperliquid team will keep executing at an elite level.
That’s a long list. But that’s what a risk premium is—it’s the price of being early on a curve that most people don’t see yet.
You don’t get AWS returns by waiting for AWS certainty.
As for me? I’ll take this bet any day of the week.
Disclaimer: Nothing in this essay constitutes financial advice. Always do your own research and evaluate risks before making any investment decisions.
As of this writing, I own HYPE and it is trading at ~$41, with a circulating supply of 334 million and a market capitalization of ~$13.9 billion.