Interview with Jeff Yan, Founder of Hyperliquid: Crypto and DeFi Are in Our DNA—We Never Compromise on Integrity
TechFlow Selected TechFlow Selected
Interview with Jeff Yan, Founder of Hyperliquid: Crypto and DeFi Are in Our DNA—We Never Compromise on Integrity
“Hyperliquid is not a ‘crypto company,’ but rather a ‘financial protocol’ that upgrades financial infrastructure using crypto technology.”
Navigating Web3 tides with focused insightsBy Jeff Yan
Translated by WuShuo Blockchain
This edition features the transcript of an interview with Jeff Yan, Founder and CEO of Hyperliquid, on the “When Shift Happens” podcast. Jeff reflects on the wealth effect and shifting responsibilities brought about by Hyperliquid’s Token Generation Event (TGE), explains Hyperliquid’s core design philosophy of “no internal order book, no discretionary authority,” and clarifies why protocol fees are automatically repurchased and burned—rather than manually timed or intervened upon.
He emphasizes that Hyperliquid is not an “encrypted company,” but rather a “financial protocol” leveraging cryptographic technology to upgrade financial infrastructure—with the goal of “housing all of finance”: enabling all financial activity within a composable, permissionless, and transparent on-chain system. Jeff also details how key modules—including HyperEVM, HIP-3 (Permissionless Perps), Outcome Markets, the alliance stablecoin USDH, Kinetiq staking, and Hyperlend lending—collectively build a decentralized, scalable, and composable on-chain financial system.
In responding to external skepticism, he stresses that Hyperliquid does not operate like a company—there is no “timed buyback” or “human intervention”; all mechanisms execute via on-chain logic. He further notes that true competition lies not in short-term metrics, but in whether one can build a trustless, globally accessible financial system.
The entire conversation centers on a core proposition: In an era of accelerating AI advancement, if the financial system does not evolve into an on-chain, programmable, open architecture, there will be no place for humans in the future of finance.
On Wealth Effects and Long-Termism
Singapore as a Startup Environment
Kevin: What’s your view on Singapore?
Jeff: It’s excellent. Truly a great place to build and ship projects. It’s very safe, highly modern, and its infrastructure runs smoothly. Many people call it “boring”—and treat that as a drawback—but to me, that’s actually one of its biggest advantages. Here you can focus entirely on building, with minimal distractions. It’s an ideal place to go heads-down.
Did You Anticipate This Scale of Wealth Creation?
Kevin: Did you anticipate such massive wealth creation at the time?
Jeff: Not at all. Honestly, I rarely project specific outcomes for particular points in time. We simply focused on doing our best work. We knew our direction and what we were building—but I never wrote down things like, “It’ll be great if we hit this quantitative milestone in one month.” That’s just not how I operate.
As contributors, so much lies beyond our control. Our job is to focus on executing what’s within our power. Outcomes certainly depend heavily on our effort—but they’re also shaped by many external factors.
That said, I was genuinely surprised by the overwhelming positive feedback later on—comments like, “This is rare in crypto,” “This is how things *should* happen—it’s just that real examples are scarce,” and “Hyperliquid shows a viable, positive path for the industry.” That felt incredibly rewarding.
Because I truly believe in this model—and in free markets. There are many external examples cited to “prove” this model doesn’t work. So when you become a counterexample—something that refutes those claims—that feels truly great.
Misalignment of Talent in Crypto—and Doing the Right Thing to Change Industry Perception
Kevin: Clearly, what you’ve accomplished is extremely difficult. But do you think enough people in crypto are trying to do the right things—the right way?
Jeff: I think there’s a widespread mismatch between what *can* be built in crypto—and who’s *actually trying* to build it.
This stems largely from crypto’s long-standing negative reputation. Compare it to fields like AI—or even pre-AI traditional finance or tech projects. If someone is highly talented, graduated from a strong school, and wants to build something genuinely valuable, I believe many such individuals today wouldn’t even seriously consider entering crypto.
Because the industry is saturated with negative examples—leading outsiders to form a stereotype: that crypto only attracts people who don’t take product-building seriously and just want to make a quick buck.
But I think that view is completely wrong. That’s not where the industry’s real potential lies.
Kevin: So how do we change this?
Jeff: By continuing to build—focusing on building what we truly believe in.
When I say “we,” I mean the entire ecosystem—not just one team. For example, the Hyperliquid ecosystem has never cared much about so-called “market meta”—nor been overly swayed by external narratives like “what should be done now” or “which sector is hottest.” Instead, we operate as a community—building what we truly believe in.
I think that may be the best thing we can do—using tangible results as feedback and demonstration.
Of course, as a permissionless network, some bad actors inevitably emerge—people who try to exploit the community or mechanisms for short-term arbitrage or value extraction for personal gain. Seeing that is indeed regrettable.
But overall, I believe our community is headed in exactly the right direction—with a strong culture. If we keep building this way, I hope outsiders gradually notice—and more people genuinely committed to long-term building will enter DeFi.
Responsibility Behind Massive Wealth
Kevin: I think many people are watching you closely now. As a founder, when you create billions of dollars’ worth of wealth almost overnight, what responsibility do you feel comes with that?
Jeff: I’m not sure whether specific events like genesis or TGE fundamentally change responsibility itself. I believe building *any* financial product carries immense inherent responsibility.
If you choose to build financial infrastructure, you must give it your all—even then, you’ll still feel like there’s too much to do. Finance is profoundly important in people’s lives. When someone uses a financial product, the level of trust they place in you vastly exceeds what they’d place in a typical consumer product.
So I don’t have some grand answer. All we can do is execute to the best of our ability—and strive for near-perfection.
That might mean slowing down. Or choosing harder, more complex ways to build systems. There are many such examples. We must always design systems around core principles: neutrality, fairness, and robustness across all market conditions.
These principles don’t change because of a TGE. Even before that, users already trusted Hyperliquid—as a fully verifiable, transparent on-chain alternative to opaque legacy systems they’d grown tired of. That trust is the real responsibility we carry.
From “In-Crowd Belief” to Global Users
Kevin: Our last recording was roughly a month before the TGE. At that time, I did something I often use to test community authenticity—because crypto has so many fake metrics and bot accounts, a well-known issue. So I posted a group photo and watched the reaction. The enthusiasm far exceeded my expectations—I realized then that this community was stronger than I’d anticipated.
It felt like “in-crowd belief”—but more like hidden code: those who get it, get it—but many insiders hadn’t yet grasped what was happening. Then the TGE happened, and things unfolded almost flawlessly. You evolved from a tightly knit, “faith-based” community within crypto into a project backed by the entire industry.
So next: How do you make Hyperliquid a “code” that attracts *outside* users? How do you get non-crypto people to join?
Jeff: Ultimately, it comes down to delivering real value—not just for Hyperliquid, but for all builders.
I don’t think the goal is to build a “cult-like community.” What people perceive as “belief” or a “cult vibe” is really resonance with builders moving toward a shared goal, guided by clear values. Its scarcity stems from how uncommon such long-termism and value alignment are in crypto.
I’m reluctant to call it “belief.” To me, it’s simply people gathering to do the right things, the right way—upholding fairness while building an open financial system. That should be the *norm*.
If we want the world to join, we must continuously prove we can deliver value that traditional finance *cannot*. And I’m deeply confident we can.
Broadly speaking, DeFi has always carried this promise. Past issues were mostly gaps between vision and technical execution. Almost no one disputes that an open financial system benefits the world. The crux lies in execution—and avoiding damage from bad actors, short-term temptations, or self-interest derailing the broader mission.
I see only two essentials: First, execute flawlessly. Second, stay committed to doing things the right way. Get those right, and DeFi becomes a massive technological upgrade to the existing financial system. Once that becomes unmistakably clear, people will naturally join.
Do You Celebrate Milestones?
Kevin: Did you celebrate? Or was it business as usual? I ask because I recall a story—maybe from Paradigm’s Matt Huang—about a Web2 company’s IPO: champagne was ready, but the whole team drank one glass of water each and returned to work until midnight, barely celebrating at all.
So I’m curious—what happened at Hyperliquid Labs? You just said it was just a milestone, with much more ahead. Did anyone celebrate? Or was it, “This was always part of the plan,” and everyone immediately moved to the next goal?
Jeff: I thought about it… and honestly, there wasn’t any formal celebration. Looking back, I do feel a bit regretful—because there were genuinely many moments worth celebrating. But in the moment, there’s always more to do. You never find a “perfect time” to pause and celebrate.
Maybe celebration will come years later—when the entire system runs autonomously, fully mature and complete, operating as a stable financial infrastructure.
Of course, individuals may celebrate in their own ways. But at least as a team, Hyperliquid Labs doesn’t have a “champagne for every feature launch” culture. That’s just not our style.
I think it’s more a cultural thing—we’re naturally more focused on what to build next, rather than pausing to celebrate what’s already been built.
On Team Selection, Work Intensity, and Privacy Boundaries
How to Screen for “High Integrity” Members
Kevin: How do you test whether someone has “high integrity”?
Jeff: That’s an excellent question. Frankly, there’s no perfect method.
At Hyperliquid Labs, our hiring process includes rigorous technical evaluation—but crucially, we always schedule at least a full day of real collaboration. Not a time-compressed, simulated-interview environment—but actual co-working.
When you work alongside someone for extended periods—discussing problems, coding together, facing real challenges—you gradually sense who they are. There are many “soft signals”—hard to quantify but perceptible. You can never be 100% certain about someone’s character, but sometimes signals arise that make you think, “This risk isn’t worth taking.”
Also, we ensure *every* team member feels genuinely comfortable with new hires. There have been cases where some strongly supported a candidate, while others had reservations. Typically, in such cases, we decline the hire.
Kevin: So it’s a consensus mechanism? A group vote?
Jeff: Not quite a formal vote—no procedural voting system. But essentially, if *anyone* raises a moderate-to-strong objection, that usually suffices to veto the hire.
On Work Intensity and Sleep
Kevin: How much do you sleep? How much do team members sleep? Do you “allow” them to sleep that much?
Jeff: It truly varies by person. We don’t pressure people on “hours invested”—because that’s often where disaster begins.
If someone needs sleep but isn’t getting it, their output quality plummets. Our quality bar is extremely high—even when fully energized, writing excellent engineering code demands full focus.
So we operate on deep trust here. Nor do I endorse “all-nighter culture” or glorifying sleep deprivation. Personally, I likely work longer hours than many who champion that culture—but I don’t believe it should be a team-wide requirement.
Some team members do work late; others have fixed, highly productive windows. As long as they consistently deliver A+ results, we don’t care how many hours they invest.
For us, it’s not “how long you work”—it’s “how good your output is.”
Team Token Vesting and Privacy Boundaries
Kevin: One final question about the team. I promise—this is the last. Over the past few months, there’s been significant FUD, including community concerns about team token vesting schedules. Many worry that ~$300M worth of tokens will flow into the market monthly for an extended period. What’s the reality?
Jeff: We won’t publicly discuss these specifics—because I genuinely believe financial privacy is a fundamental right. That’s one reason we entered crypto and engaged in DeFi.
Just as you wouldn’t demand a community member disclose, “You hold many tokens—you must tell everyone what you do with each transaction,” there must be a clear boundary—and that line is personal privacy.
I believe protocol operations *must* be fully transparent. Every dollar’s flow must be traceable; assets in the system must truly belong to their holders; protocol rules must be public and verifiable. This level of transparency is non-negotiable for an open financial system. Without it, you get the problems we see at centralized exchanges.
But conversely, how contributors manage *their own* tokens is their personal asset. It shouldn’t be subject to public scrutiny—or my duty to judge or disclose.
Protocols must be transparent—but individuals must retain privacy. These aren’t contradictory.
On Responding to FUD, Automatic Burn Mechanisms, and Removing Human Discretion
How to Handle FUD and External Attacks
Kevin: Your approach to FUD and external criticism is interesting. Many assume Hyperliquid just ignores it and stays silent.
Jeff: That’s not quite right. Early on, we *did* lean toward non-response. But later I realized my instincts on PR aren’t always correct—so now I defer more to professionals better equipped to assess and handle it.
One thing we deliberately adjusted: When FUD arises, my instinct is often, “This is clearly false—truth will surface on its own.” But I’ve gradually realized that mindset isn’t always right.
Practically speaking—if false information is spreading *and* clarification is needed, we *will* respond. We address specific issues directly—not ignore them wholesale.
Of course, we won’t react to every voice. The key is judgment: Does this mislead users, erode community trust, or distort protocol understanding? If yes, we engage head-on.
So rather than “ignoring FUD,” we’re learning to respond more maturely and strategically.
Kevin: Does this FUD affect you personally? Be honest.
Jeff: It depends on the type of FUD.
Over the past six months, crypto markets have been highly volatile—often downward. Certain events caused real harm. And since Hyperliquid is one of the few truly transparent trading venues, much discussion naturally centered on us.
During that time, I *was* deeply affected. I acted almost with a “mission”—to clarify technical details. Because these questions are deeply personal—many people lost substantial money. Emotions ran extremely high.
But I also saw competitors deliberately misrepresenting Hyperliquid from different angles—downplaying what we got right, using negative narratives to deflect attention from their own issues.
That makes me angry—because they clearly know what they’re doing.
Yet that anger ultimately fuels motivation—to explain more clearly to users *why* Hyperliquid is designed this way, and *why* transparency matters so much.
For example, some centralized exchanges cite analytics sites claiming, “Look, Hyperliquid has more liquidations than us.” But if you read those platforms’ documentation (though few do), you’ll find they may only report “the first liquidation per second,” without disclosing full data—or explaining why only that subset is reported.
At Hyperliquid, *every* order, cancellation, and liquidation is on-chain and publicly verifiable. Comparing “partially disclosed data” against “fully disclosed data” inevitably yields misleading conclusions.
This is fundamentally information asymmetry—or outright deception.
What’s even more frustrating is when influential voices—with massive followings—amplify such narratives, magnifying their impact exponentially.
Our voice may be smaller and narrower in reach. But precisely because of that, we need to speak loudly, clearly, and repeatedly at critical moments to state facts.
I won’t pretend this doesn’t affect me. It does. But what it brings is more responsibility—not retreat.
Automatic Buyback & Burn vs. Discretionary Buybacks
Kevin: Many support you, but some very smart industry insiders criticize Hyperliquid. A common critique: You repurchase tokens daily—but should pause buybacks when prices are high and increase them when low. What’s the core flaw in that view?
Jeff: The core flaw is—Hyperliquid isn’t a “discretionary buyback program” at all.
Many interpret it as corporate behavior—but that’s a mistaken analogy. Hyperliquid is a *rules-based protocol*, not a company empowered to make subjective decisions.
Take Ethereum’s priority fee: It’s now directly burned. You wouldn’t ask Ethereum developers, “When ETH price is high, shouldn’t Vitalik invest those fees elsewhere instead of burning them?”
Because that’s not a subjective decision—it’s embedded in the protocol’s rules.
Same with Hyperliquid: Fees generated by the protocol are automatically converted to HYPE and burned per preset rules—not decided ad hoc by the team (“How much today? When to stop?”). The entire conversion logic resides in on-chain execution.
I think this misunderstanding arises because many centralized or semi-centralized exchanges *do* run discretionary token buybacks—where companies decide timing, volume, and pauses. For those structures, discussing timed buybacks is reasonable.
But at Hyperliquid, fee conversion and burning is automated on-chain logic.
It’s like on-chain TWAP order execution: When you submit a TWAP, no one sits manually deciding when to split orders or send the next one to the order book. That’s part of the protocol’s execution logic.
Similarly, how fees are converted and burned is purely on-chain execution logic—not human operation.
So the essence is—many mistake a “protocol-level automatic burn mechanism” for a “corporate-level timed buyback strategy.” These differ fundamentally in design philosophy and governance logic.
Why Removing Discretionary Authority Is Essential
Kevin: Why is “no discretionary authority” so vital for Hyperliquid?
Jeff: Because Hyperliquid isn’t a company. That’s another common misconception.
When I say “we” or refer to Hyperliquid Labs, we’re just a very small team. Our role is building a small—but critical—component of the broader Hyperliquid ecosystem. We are *not* “Hyperliquid itself.”
People habitually map everything onto traditional finance analogies—because learning new concepts is hard, and fitting new things into existing frameworks feels familiar. Yes, some parallels exist—but many don’t apply at all.
Hyperliquid isn’t a mapping of any existing entity in traditional finance. It’s not a company, not an exchange company, not an organization structured to “create value for shareholders.”
Our vision is for the financial system itself to run on-chain.
Yes, the protocol creates value and feeds it back to the native token—that’s positive. But this logic differs entirely from a company returning profits to shareholders.
A company is a centralized entity—with management, a board, and subjective decision-making authority. Hyperliquid is a neutral platform—a financial infrastructure layer upon which finance is built.
Put differently, it’s like the internet for information. The internet isn’t a company; it doesn’t exist to profit “internet shareholders.” It’s a neutral infrastructure protocol enabling free information flow.
Likewise, Hyperliquid—as an ecosystem, network, and protocol—aims to become the infrastructure layer for finance.
Once you grant a core team “discretionary authority”—like deciding when to buy back or pause—you effectively revert the system back to “company mode.”
We chose rule-based, automated, immutable mechanisms precisely to ensure the system remains neutral and trustworthy—not dependent on any team’s judgment.
If the goal is a true on-chain financial system, removing human discretion isn’t optional—it’s foundational.
On HyperEVM, Permissionless Perps, and Financial Onboarding
The Essence and Significance of HyperEVM
Kevin: Last time you were on the show, we discussed your aspiration to become the “AWS of liquidity.” Let’s dive into what’s happened since. Start with HyperEVM—what is it? Explain as simply as possible, no jargon.
Jeff: Simply put, HyperEVM lets developers deploy smart contracts originally built for Ethereum directly onto Hyperliquid—largely unchanged.
Think of it as a “blank canvas.” Developers can deploy contracts and run logic there, just as they do on Ethereum.
On the surface, the experience mirrors Ethereum. Innovations accumulated over Ethereum’s past few years—dev tools, contract standards, app patterns—are all reusable on Hyperliquid.
Kevin: What’s your honest assessment of HyperEVM’s current performance?
Jeff: That’s a nuanced question.
First, I think a common misconception is that HyperEVM “hasn’t succeeded.” If needed, I could list many examples where it’s performing exceptionally well.
More importantly, look one layer deeper: HyperEVM isn’t just “copying Ethereum’s stack.” Its unique value lies in letting smart contracts *directly invoke Hyperliquid’s native on-chain primitives*—like native liquidity and on-chain infrastructure. That’s critical.
I think this narrative hasn’t been fully grasped yet. HyperEVM isn’t just another standalone EVM chain. It’s more like an “entry point”—a direct interface connecting to Hyperliquid’s native liquidity and on-chain infrastructure.
So evaluating HyperEVM shouldn’t compare it to isolated, self-contained EVM chains. It’s not a siloed ecosystem—it’s a portal to Hyperliquid’s core capabilities. Viewed this way, its significance changes entirely.
Why HyperEVM Is Underappreciated
Kevin: Put simply—is it that people haven’t truly understood HyperEVM yet?
Jeff: I think some understand it—but that understanding hasn’t entered mainstream narratives. Take a concrete example: For a long time, critics asked, “Why hasn’t Circle deeply integrated yet?”
The issue is that HyperCore implements a full suite of native financial primitives—from account balance ledgers onward. Its architecture differs significantly from general-purpose public blockchains. For companies with mature infrastructures tailored to general-purpose chains, integrating directly with HyperCore is extremely difficult. That’s where HyperEVM shines—it provides a bridge.
For instance, native USDC minting and burning logic can be handled via Circle’s existing Hub-and-Spoke network: burn on one chain, mint on another. HyperEVM is one of the supported chains.
But to grasp this, you must understand the architectural intent. It’s not merely “launching an EVM environment”—it’s enabling external infrastructure to integrate *standardly*, while deeply interacting with HyperCore’s native capabilities.
The problem is—when integrations work *this* well, they become “invisible.” UX is seamless—you don’t realize it’s powered by special HyperCore–HyperEVM interaction. Hence, it’s hard to generate “explicit success story” narratives.
Many ask: “Why haven’t I heard HyperEVM success stories?” Partly, major applications are still under active development. Partly, successful integrations succeed *so* natively and naturally that they don’t seem “special.” If infrastructure truly succeeds, it feels inevitable. That, too, contributes to being underestimated.
HIP-3: The Significance and Challenges of Permissionless Perps
Kevin: Another initiative you launched last year—initially hard for most to grasp—has now yielded concrete results via HIP-3: Permissionless Perps. What does this mean? Explain as simply as possible.
Jeff: Simply put, Perps (perpetual futures) are among crypto’s most important innovations. Though theoretical foundations existed in academic literature, BitMEX proved their practical viability—demonstrating them as a more efficient price discovery tool by concentrating liquidity in a single, never-expiring contract.
HIP-3’s core idea is simple: Crypto imposes no inherent restrictions on perpetuals. So if an asset is liquid and already has futures or options markets, it *should* support perpetuals too. We’re not claiming perpetuals are superior to all other derivatives—but users *should* have the choice.
I’ve long believed that if traditional finance widely adopted perpetuals across more asset classes, a substantial portion of trading volume would migrate to Perps—because they’re among the most direct, efficient tools for expressing views—especially with leverage and two-way trading.
HIP-3 in one sentence: Anyone can deploy perpetuals on Hyperliquid—i.e., “Permissionless Perps.”
But behind that sentence lies immense complexity—because no one has ever built a fully permissionless perpetual deployment system before. Core challenges include:
- First, can the protocol entrust a permissionless deployment system with ultra-critical functions like real-time oracle updates?
- Second, perpetuals tightly couple with margin systems. Making them permissionless introduces countless edge cases and risk scenarios—requiring extremely rigorous design.
- Third, how to maintain system robustness and security amid open deployment? These aren’t solved by tweaking a few lines of code—they’re deeply interwoven. Plus, builders must invest time to understand a novel mechanism and commit to building on it. That takes time too.
Frankly, before HIP-3 launched, many doubted its feasibility. Even we couldn’t guarantee 100% success. But logically, it *should* work.
And when an ecosystem gathers capable developers—and users who truly align with its ethos and help ignite new network effects—innovations like this *should* succeed precisely in such environments. Hyperliquid offers fertile ground.
Fundamental Difference Between Hyperliquid and Centralized Perp Platforms
Kevin: Can you explain how what you’re doing differs from what others are doing?
Jeff: I hesitate to comment on “everyone,” but many non-crypto attempts at Perps follow this logic: An exchange already running crypto perpetuals says, “Let’s add price feeds for other assets and offer perpetuals for them the same way.”
You see this across centralized exchanges—and some decentralized ones. The core idea is: Same platform, same architecture, same team—listing perpetuals for more assets.
Kevin: So what’s the *real* difference between your approach and theirs—and why should users care?
Jeff: End users may not need to care about underlying implementation. But if we truly want to move *all* finance on-chain, we must let domain experts—those who deeply understand specific fields—deploy and build these products.
Look at traditional finance: Each country has its own exchange system; beneath that, asset classes diverge; different assets trade on different exchanges with distinct rules; even within one country, asset classes have different trading interfaces. Drill down—and you’ll see finance is inherently highly specialized.
Yes, some complexity stems from traditional finance’s inefficiency. Crypto eliminates some intermediaries. But finance is vast—impossible for one team to build entirely.
Many large financial firms pursue another path: maintaining a centralized “super-platform,” consolidating users’ entire financial lives in one place—via acquisitions, partnerships, or internal development. It’s top-down integration.
This model advances faster, reduces fragmentation, and may offer better early UX. But it’s fragile—problems often stem from excessive centralization of power. When everything’s centralized, risk is centralized too.
I believe end users will choose the system that grants them true financial sovereignty—one globally accessible, sufficiently liquid, and offering required functionality.
To build such a system for global users, decentralization must occur across multiple dimensions. Decentralized ownership is one key principle—central to Hyperliquid. Equally vital is decentralizing *who builds the core products users interact with*.
That’s our fundamental difference: Not a centralized team listing perpetuals for all assets—but building an open system where domain-specific builders deploy markets they truly understand and excel at—on a shared underlying protocol.
Why Moving All Finance On-Chain Is Essential
Kevin: In one sentence: Why is moving the entire financial system on-chain so crucial?
Jeff: Because if we don’t, the financial system will fall behind technological progress. The world moves too fast—legacy finance can’t keep pace or meet user needs in a rapidly evolving environment.
Kevin: Why is Hyperliquid’s approach better?
Jeff: This isn’t just Hyperliquid’s approach. Hyperliquid embodies DeFi’s enduring values—we’re just a community attempting to execute them in a way with greater potential for genuine success and mass adoption.
Core values include: Finance shouldn’t be controlled by corporations—but should be a neutral foundational layer, like the internet.
You mentioned AWS—an interesting example. Though AWS is company-controlled, it functions *like* neutral infrastructure. No one asks, “Why choose AWS? Isn’t that a biased choice?” It’s just default infrastructure.
Finance should share that trait—a neutral, globally accessible infrastructure. People shouldn’t face layered approvals and barriers due to nationality, identity, or origin in developed nations to access finance.
Additionally, finance must be deeply programmable—not fintech bolted onto an unprogrammable legacy system—but built from the ground up on programmable primitives. Base components themselves are code—smart contracts—interactable directly by humans, agents, and programs.
If finance is to keep pace with technology, it must become a neutral, programmable, globally open infrastructure—like the internet. That’s why moving *all* finance on-chain is essential.
HIP-3 Data Performance and Silver Trading Case Study
Kevin: If this approach is better, data performance should reflect that. Can you share key metrics proving Hyperliquid operates perpetual markets correctly?
Jeff: I still emphasize—we’re extremely early. The future is never guaranteed; short-term data alone can’t prove long-term success.
That said, existing data already reflects the power of permissionless protocols—when you empower deployers to freely express creativity and execute visions, results become quantifiable.
A representative example is the silver market. Recently, silver and other metals experienced extreme volatility—multi-day “10-sigma” moves—drawing intense attention from both institutions and retail traders.
On Hyperliquid, related markets launched via HIP-3 (primarily operated by first deployer XYZ) now account for ~2% of global silver trading volume in global price discovery.
2% sounds modest—but in the context of the *global silver market*, it’s astonishing. More importantly, these HIP-3 markets launched just months ago. For a nascent on-chain perpetual market, this growth is remarkable.
And this achievement wasn’t driven by core team trading. The team built primitives—HIP-3 itself required heavy engineering. But how these liquidity infrastructures are used depends entirely on builders choosing to deploy atop them.
This is more proof of builder quality: They chose Hyperliquid because only this platform enables building the market structures they envision.
Kevin: You said data is starting to get interesting—but still early—so the future isn’t guaranteed. Why say that?
Jeff: Because treating a currently evidence-supported idea as “immutable” is dangerously rigid. The world changes too fast.
Some principles are relatively stable—for example: Finance should be open to all globally; finance, like information, belongs in open systems; users should hold financial sovereignty. These won’t easily shift. But over-fixating on *any specific implementation* makes it fragile.
A key starting point in designing blockchain primitives is: When builders and users across diverse directions flag the same issue—“We want to do X, but the current system can’t”—and those feedbacks converge, that intersection often reveals a foundational capability worthy of protocol-layer inclusion.
When the protocol layer exercises restraint—selectively choosing *what must be native primitives versus what belongs in generic EVM or application layers*—the system becomes more elegant and resilient.
Thus, the protocol avoids embedding specific business models or market judgments—leaving “opinionated decisions” to builders. Builders adapt swiftly to real-market shifts, while the base layer remains neutral.
All HIP proposals—including HIP-3—embody this logic.
Spot Trading Function as Price Discovery Hub
Kevin: We previously discussed spot trading—a feature not built by the core team, but by Unit in 2025. What does this signify?
Jeff: 2025 saw several critical launches—and during early stages for these assets, Hyperliquid became the primary source for spot price discovery. And—to my knowledge—this marks the *first time* a decentralized exchange has become the dominant spot price discovery hub for a new asset.
One example is XPL. During critical phases, Hyperliquid led in on-chain spot market depth and volume.
In a sense, spot trading most closely resembles Satoshi Nakamoto’s original “peer-to-peer electronic cash” vision. Because spot trading is fundamentally on-chain ledger asset transfer:
One party transfers the quote asset to another; the other returns the base asset. Order books and local matching mechanisms are merely coordination tools facilitating these on-chain transfers.
If you asked Nakamoto how Bitcoin should trade, I’m confident he’d say: fully on-chain. Of course, Bitcoin’s base layer may not suit complex matching logic—but trading itself should be peer-to-peer.
At Hyperliquid, we’ve built many things—but this “native on-chain execution” ethos is core to our DNA.
I believe spot trading is an excellent validation scenario. When all pieces align—performance, order books, UX, liquidity—users prefer transacting in ways consistent with crypto assets’ underlying technical logic.
Most people trade spot crypto assets because they believe in crypto and DeFi’s vision. And after years, finally trading held assets fully on-chain—without sacrificing UX—feels like closing a loop.
From technical ideal to real-world implementation—back to the original peer-to-peer concept—it’s a “full circle” feeling. I think that’s incredibly cool.
On Early Entrepreneurship, Nonlinear Financial Primitives, and On-Chain Finance
Jeff’s Early Prediction Market Project Failure Experience
Kevin: You’ve posted videos on X saying that in 2018, “young Jeff” built a prediction market project—around the same time as Kalshi. Why didn’t it succeed?
Jeff: Many reasons. The simplest and most honest answer: We weren’t ready yet.
The idea itself was cool. From an infrastructure perspective, we got many things right—like off-chain matching with on-chain settlement. Given the tech landscape then, that was a reasonable architecture.
But building a successful product involves far more than “having a good idea.” Ideas constitute a tiny fraction of the process. Execution, product refinement, market timing, user education, liquidity bootstrapping, psychological resilience—all matter equally.
Back then, it wasn’t obvious anyone truly wanted on-chain products. Early on, market sentiment was euphoric—prices rising, everyone excited. But by late 2018, the market crashed sharply.
Crypto’s cyclical nature is now familiar—we’ve seen bull markets draw crowds, bear markets leave near-empty rooms.
But if that’s your first cycle—and you talk to users only to find almost no one willing to try your product—the blow is crushing—especially after investing immense time and energy.
Still, that experience delivered deep learning. We truly grasped what long-term commitment and mindset building a successful financial product requires.
In a sense, the idea never died. Some teams launching around the same time persisted—like Kalshi. They clearly possessed the conditions and resilience to make it happen—and are now seeing years of effort yield results and broader mainstream adoption.
Seeing these companies succeed, I’m genuinely happy for them. It shows many ideas aren’t wrong—just a matter of being ready, acting at the right time, and persevering the right way.
Outcome Markets: Nonlinear Expression Tools
Kevin: Eight years later, you’re back in prediction markets—but more accurately, Outcome Markets. How do they differ?
Jeff: This returns to our “primitives” design philosophy. At Hyperliquid, we aim for protocol-layer primitives that are as small and self-contained as possible—simple in structure, clear in performance, yet maximally versatile.
Outcome Markets begin with a premise: They’ll become the core tool for expressing “nonlinear views” on-chain.
Compare first: Spot is the most basic asset form. You hold it; trading completes asset transfer on the on-chain ledger—a foundational primitive no one disputes. Perpetuals (Perps) are more contested—but increasingly recognized for their price discovery and capital efficiency value, making them worthy universal financial tools. HIP-3 represents a staged realization of this idea.
Though spot and perps differ greatly in mechanics—perps enable leverage and higher capital efficiency; spot represents tokenized real assets—they share one key similarity: Both express *linear views*. Whether using 1x or 100x leverage, your P&L changes linearly per unit price movement. Sufficient for many—but some users truly seek *nonlinear outcomes*.
Examples: A contract paying $1 or $0 depending on whether an event occurs (classic binary); holding Bitcoin long-term but wanting downside protection if price falls below a threshold; wanting capped return structures. These are nonlinear views—impossible to replicate precisely with spot/perps combos. That’s Outcome Markets’ purpose.
Outcome Markets are essentially “fully collateralized contracts.” Both parties deposit collateral upfront, redistributed based on outcome. Settlement can be binary or continuous—but crucially, *no liquidation risk exists*. Users deposit capital, await settlement—no margin calls, maintenance ratios, or funding rates.
Kevin: Like options or prediction markets—similar examples?
Jeff: Yes—options and prediction markets are the most intuitive examples. Both feature nonlinear payoff structures.
Additionally, more intriguing applications exist. For instance: Can markets aggregate opinions—not just trade prices? Can market mechanisms “bubble up” critical topics? Can subjective outcomes drive capital-backed consensus? All naturally fit within Outcome Markets’ framework.
Perpetuals remain vital price discovery tools—but are relatively complex financial primitives. Outcome Markets provide a more foundational, universal nonlinear expression capability.
Simply: Spot solves “owning assets,” Perps solve “linear views + leveraged expression,” Outcome Markets solve “nonlinear view expression.” Together, they form a more complete on-chain financial expression system.
The True Meaning of “Housing All of Finance”
Kevin: You often say “housing all of finance.” What does that mean?
Jeff: Its core meaning is that finance shouldn’t be fragmented—that’s what “all” signifies. You shouldn’t need one bank account, then transfer funds to another for purchases—or open new accounts for specific products. Ideally, users’ entire financial lives should unfold in one place. I believe “composability”—integrating financial activities into a unified system—is a massive breakthrough.
But this isn’t unique to crypto. What crypto means by “housing all of finance” is more crucial: This platform shouldn’t be controlled by any centralized company. It should allow many diverse teams to build. When someone builds a new app or protocol, it should seamlessly integrate and compose with others’ systems—creating true network effects that constitute a financial system.
Here, a constant tension exists: One end is fragmented but highly composable; the other is potentially more efficient but centrally operated with limited functionality. These poles constantly pull against each other.
But with thoughtful design, balance is possible—retaining composability while ensuring efficiency. This principle drives much of our product and system design thinking.
Hyperliquid Isn’t a Crypto Company—It’s a Financial Protocol
Kevin: You once told me Hyperliquid isn’t a crypto company—but a financial company using crypto as infrastructure. Though “company” may not be quite right. Can you explain?
Jeff: I think it’s about distinction. We genuinely believe in DeFi values—crypto and DeFi principles are in our DNA.
But we don’t define ourselves as “building the best crypto exchange.” That’s not our vision.
To me, finance is just finance. Crypto’s significance lies in its potential to upgrade the underlying “rails” for financial activity—rebuilding financial infrastructure with blockchain, enabling more open, efficient, transparent financial behavior. That’s our direction.
If the Hyperliquid ecosystem succeeds, some may say: “This proves crypto is right—crypto is validated.” I think that’s a fair interpretation.
Others may say: “This is something new—inspired by crypto ideas but evolved into an independent system.” I think that’s equally valid.
The label—“crypto company”—matters less than whether it truly upgrades financial infrastructure.
USDH Alliance Stablecoin Mechanism
Kevin: Hyperliquid’s core team has only 11 people—but also hosts an ecosystem of the industry’s brightest developers and builders. Let’s discuss one project. I noticed your stablecoin USDH—why build it?
Jeff: Roughly six months to a year ago, the industry saw a wave of stablecoin enthusiasm across multiple blockchains. With the “Genius Act” passing, markets widely believed stablecoins would be institutional capital’s first entry point into DeFi rails—sparking high interest.
Within Hyperliquid’s ecosystem, we proposed a protocol-level concept: the “Alliance Stablecoin.” This wasn’t predetermined—it emerged from repeated discussions with ecosystem participants, culminating in a protocol-layer design.
Its core: Yield behind the stablecoin (e.g., from Treasury assets) is distributed per preset rules between the protocol itself and the stablecoin issuer. USDH is operated by Native Markets—a team including experienced institutional founders and strong ecosystem builders. I believe they’re excellent—and look forward to them building a deeply integrated, aligned stablecoin on Hyperliquid.
These stablecoins maintain highly symbiotic relationships with ecosystem apps and developers.
For example: The protocol earns yield—so HYPE token holders’ interests align with stablecoin growth; users trading with the stablecoin enjoy lower fees. All rules are written into the protocol layer—objectively and automatically executed, not subject to human discretion.
We believe this design creates powerful synergy—and unlocks entirely new application scenarios.
Kinetiq Full-Chain Staking Innovation
Kevin: What is Kinetiq—and why is it important to Hyperliquid’s ecosystem?
Jeff: Kinetiq is currently Hyperliquid’s largest liquid staking protocol. It embodies a key part of HyperEVM’s vision.
As noted earlier, HyperEVM can directly “read” and “invoke” HyperCore’s capabilities via specific mechanisms—effectively opening a window from smart contracts to the underlying core system, letting them use core functions as tools. This architecture enables fully on-chain liquid staking.
Typically, staking builds atop system mechanisms lower than the application layer. On Ethereum, leading liquid staking tokens (e.g., stETH) are on-chain tokenized assets—but often involve off-chain components or operational processes—meaning tokenization isn’t fully closed-loop on-chain.
At Hyperliquid, staking-related read/write operations—including staking, unstaking, and yield accounting—are directly callable by smart contracts via precompiles and core writer primitives. This means: A smart contract can tokenize a pool of staked HYPE tokens just as vault shares are tokenized on EVM; staking, unstaking, and accounting all close the loop within one on-chain system. This is a truly “fully on-chain” liquid staking mechanism.
Its significance: It eliminates risks from cross-chain bridges and extra trust assumptions. No off-chain dependencies, no additional custody steps—everything secured at the protocol layer. Thus, we see this as a very cool and important innovation in liquid staking.
Hyperlend’s Critical Role in Hyperliquid’s Portfolio Margin System
Kevin: Why is what Hyperlend is building important to Hyperliquid?
Jeff: Hyperlend is currently the leading lending protocol on EVM. Since launch, its TVL has grown notably—I attribute this mainly to consistent iteration and solid execution. It runs fully on EVM: Users deposit collateral as suppliers; borrowers enter the market to borrow assets—essentially a supply-demand-driven interest rate market, where borrowers pay interest to suppliers.
This is a fairly common DeFi model—but once Hyperliquid’s HyperCore portfolio margin functionality matures, this lending system becomes critically important.
We haven’t detailed portfolio margin yet. Its core idea: Traders want to use *any* assets they hold—especially highly liquid ones (e.g., BTC)—as unified collateral to trade *any* market—not restricted to one specific market. This reduces liquidity fragmentation and expands eligible collateral, boosting capital efficiency for perpetual or spot trading.
But in DeFi, safely implementing portfolio margin requires only one feasible foundation: a sound, robust lending layer. That’s why portfolio margin has struggled to scale in DeFi—it needs strong lending support.
At centralized exchanges, enabling portfolio margin lets the exchange “magically” generate internal ledger balances for users. For example, a user pledges BTC—the exchange gives them a credit line to trade USD-denominated perpetuals. It works because the exchange trusts pledged assets cover this “internal credit balance.”
This usually works—but black swan events can cause bad debt. Fairly, CEXs face extremely complex problems—and manage them reasonably well. Portfolio margin remains a key user feature.
But this “internal credit expansion” is unacceptable in DeFi protocols. If a DeFi protocol adopts it, it essentially bears the risk—anyone depositing funds indirectly exposes themselves to portfolio margin collateral failure. This violates DeFi’s core principle: Every dollar in the system must be fully verifiable and fully backed—no implicit risk premises. For decentralized protocols, this trade-off is unacceptable.
Thus, Hyperliquid’s design avoids “credit expansion from thin air”—instead, portfolio margin is fully backed by real lending markets. Lending pools directly provide real liquidity support for portfolio-margin borrowing. This resembles Aave V3—but with deeper architectural integration.
This approach is more complex—users may not perceive structural changes—but it’s more robust and scalable. Crucially: When demand surges for collateral on a specific asset, the system doesn’t raise platform-wide risk—instead, interest rates adjust—borrowing rates rise short-term, attracting more suppliers seeking yield, eventually pushing rates back to equilibrium via market competition. In short: Demand shocks cause short-term rate fluctuations—not increased protocol risk. A healthy market adjustment mechanism.
I believe Hyperlend and other lending protocols will integrate highly synergistically. Portfolio margin funnels demand from HyperCore to the EVM layer, while HyperEVM becomes infrastructure for lending implementation and asset tokenization. Lending activity happens on EVM—while supporting HyperCore—enabling efficient distribution of liquidity both internally and externally.
On Market Competition, Project Vision, and Industry Future
What “We Have No Competitors” Really Means
Kevin: Last year, I saw your banner: “We have no competitors.” Many thought, “Is this self-denial? Or do most people not truly understand Hyperliquid yet?”
Jeff: I think the key question is—*who* is competing with you?
I don’t mean zero competition. In a sense, many competitors exist—because Hyperliquid sits at the intersection of many sectors: trading, chains, lending, infrastructure, community ecosystems, etc. Finance has countless participants—many view Hyperliquid as a competitor, but I think that framing isn’t fully accurate.
In many scenarios, it’s less competition—and more collaboration.
I’m seeing this trend emerge. Different protocols are building products in Hyperliquid’s ecosystem. Just days ago, a developer told me: “We’re multi-chain deployed—but the structured product we launched on Hyperliquid performed better than anything we’ve done in the past year.” They were genuinely surprised.
Similar stories will keep emerging.
So people naturally see markets through a “zero-sum competition” lens—but reality is more “cooperative competition,” not value dilution.
Conversely, I do believe no one else is building *exactly* what Hyperliquid is building.
HyperCore’s architecture and direction stem largely from ecosystem evolution—not a preordained grand blueprint, but a three-plus-year organic progression. It hasn’t been smooth—community members have journeyed with us from day one through many ups and downs.
Any ecosystem’s development is path-dependent.
From day one, Hyperliquid upheld “no insider” principles—fairness and openness to users and developers. This culture shaped today’s ecosystem structure. I hope to uphold these values regardless of scale.
Because these principles enabled building things others haven’t attempted.
For example, when other teams launch perpetual products, few say: “We trust the community and developers to co-build the ecosystem—even with fragmentation and internal competition—because it leads to a stronger system.”
But at Hyperliquid, this story has played out many times—and will continue.
Each time communities attempt building decentralized systems rivaling—or surpassing—centralized exchanges, skeptics ask: “Can the community *really* do this?”
We just keep proving it.
So when I say “we have no competitors,” I truly mean: On this path—community-centric, fairness-driven, on-chain-mechanism-challenging-centralized-alternatives—I struggle to find an opponent with identical goals and methods.
This isn’t denying competition—it’s stating we walk a uniquely distinct path.
The Long-Term Cost of Fairness and Integrity
Kevin: You often speak of “integrity” and “fairness.” Tweeting “I’m fair” is easy—but enacting it carries high cost. Where’s the highest cost for you?
Jeff: It’s hard to answer with specific incidents—because internally, it’s never a “should we be fair?” choice.
We never sit down debating: “Is this fair or unfair? Should we do a cost-benefit analysis?” That’s not our mindset. So I can’t pinpoint one specific trade-off as the cost of upholding fairness.
But I have no doubt such costs exist.
If you compromise integrity and fairness for shortcuts, growth will almost certainly accelerate. That’s undeniable.
Such things have happened in the market. For example, FTX was a classic hyper-scaling case—growing at breakneck speed. Had it remained undiscovered, it might have masked problems long-term.
For some, this *is* a cost-benefit calculation.
But for us, it’s not a calculable business decision. Because compromising on integrity carries potentially *infinite* long-term cost—risk not quantifiable as loss, but as fundamental damage to the entire system and community trust.
For us, integrity and fairness aren’t negotiable chips—they’re non-transferable principles.
Even at high cost, this principle remains unaltered.
Hoping Hyperliquid Becomes “The Internet of Finance”
Kevin: What do you hope Hyperliquid is remembered for?
Jeff: I don’t hope it’s “remembered.” I just hope it’s continually *used*.
The ideal is like Bitcoin—becoming a long-lasting store of value, uncontrollable by any single authority, truly belonging to everyone. I’m uncertain how many crypto projects will withstand time—but I hope Hyperliquid becomes “the internet of finance.”
Why Stay in Crypto?
Kevin: Many founders and investors are exhausted. Crypto markets disappointed many over recent years—innovation seems stagnant. Meanwhile, AI is siphoning talent. As one of the industry’s most respected builders, can you give people reasons—why stay in crypto?
Jeff: From Hyperliquid’s perspective, I believe humanity has a vital task in our era.
AI is a “ticking time bomb.” It’s advancing rapidly—and will ultimately displace human intelligence at scale. Personally, I think it’s not as imminent as many imagine—e.g., current AI still can’t write truly high-quality, mission-critical code. Those truly important lines, I don’t believe should be entrusted to AI. Maybe I’m conservative—but the trend is inevitable.
Before AI reaches “self-accelerating evolution”—achieving self-improvement and rendering human intelligence marginal—we must first build a financial system *accessible to machine intelligence
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
吴说区块链
