Hyperliquid is a name that basically everyone knows now. Its on-chain revenue consistently ranks in the top three. After being named by Trump, combined with the explosion of RWA trading, Hyperliquid's platform token HYPE has repeatedly hit all-time highs, approaching $100. The ecosystem is becoming increasingly complete. Leveraging the HIP-3 mechanism, it has produced giants like TradeXYZ; HIP-4 has also officially launched, beginning its foray into prediction markets.

This is Hyperliquid founder Jeff's first time on a podcast. In the podcast, Jeff explained Hyperliquid's mechanisms and operating logic, as well as some of his views on regulation.

TL;DR

· Hyperliquid is a protocol. Its slogan, or rather its aspirational goal, is to "contain the entire financial system." Simply put, it is a neutral base layer where finance can ultimately happen.

· We see the Hyperliquid protocol as a step-function improvement over "markets that are already electronic but still fragmented."

· We didn't see this as building a trading platform, but as building an entire network. At the time, we examined existing blockchains, and none could support the throughput and latency needed to "contain global finance" without making unacceptable compromises.

· At the base layer of a global financial system, everyone must reach consensus. If we can't even agree on the ledger of account balances, or the basic state of the system, then everything else is off the table.

· Liquidations are rule-based, determined by the clearing system, not by someone's subjective discretion; what's unique here is that the clearing system is entirely on-chain, with the HyperCore execution module autonomously executing clearing and liquidation mechanisms.

· The real beauty lies in the modular separation of responsibilities: the protocol provides the rails for execution and clearing, while deployers specify contract parameters and maintain oracles; this modularity is unique to on-chain financial systems.

· Auctions occur approximately every 31 hours, because 31 is coprime with 24, allowing reset times to rotate across global time zones; the price declines linearly from a high point, and the first bidder to accept the price immediately wins the seat, making it naturally resistant to wash trading.

· Look at this year's tech company IPOs: enormous wealth has been created, but the process has been heavily gatekept. When companies list at $1 trillion, $2 trillion, $3 trillion valuations, the vast majority of wealth creation has already been completed in private markets. This is a fundamental flaw of the existing system.

· The protocol enforces mathematical boundaries, not subjective market opinions; regulation naturally falls on the layer where discretionary control exists, which is the deployer exercising control. On Hyperliquid, anyone in the world can audit in real time, in a provably mathematical way, that the sum of all user balances exactly equals the amount of USDC held by the smart contract address.

Vision Positioning and Regulatory Status

What is Hyperliquid? What does "accommodate the entire financial system" specifically mean?

Hyperliquid is a protocol. Its slogan, or rather its aspirational goal, is to "accommodate the entire financial system," and I believe the community is serious about this. Simply put, it is a neutral underlying layer where finance can ultimately take place.

Decades ago, electronic trading was a step-change improvement over open outcry trading on exchange floors; similarly, we see the Hyperliquid protocol as a step-change improvement over markets that are already electronic but still fragmented.

Trump previously publicly mentioned wanting to bring Hyperliquid to the United States. Did you know about this in advance? What is the real regulatory status for U.S. users?

I did not know in advance, so that was a huge surprise to me. Hyperliquid Labs develops on this protocol. Perhaps it is necessary to break down a few parts clearly: the protocol itself is a neutral blockchain operated by a set of permissionless validators who jointly maintain a single, consistent global view. Hyperliquid Labs is a very small team that builds technology for this blockchain.

As for what the CFTC means when it says it wants to bring this protocol to the United States: in fact, U.S. users trading derivatives on this protocol violates the terms of service of Hyperliquid as well as many other front-end interface services.

However, I believe the current administration and the CFTC want to take a more proactive stance toward on-chain financial infrastructure. This is a technical judgment: markets can be improved, and the United States should not take an extremely restrictive position, but should instead find ways to incorporate this technology into the existing financial system.

Underlying Architecture, Consensus Mechanism, and On-Chain Settlement

What roles make up Hyperliquid's protocol architecture? How do validator nodes and the protocol collaborate?

From a high level, the protocol is a messaging framework, similar to HTTP or TCP, where different validators agree to communicate according to specific rules.

Compared with pure information-format protocols, what is interesting about blockchain protocols is that: under certain assumptions—for Hyperliquid, that two-thirds of the network's stake is non-Byzantine and operating according to normal rules—you get very good properties, and these validators can each operate independently.

The entire network will reach a globally consistent world view. This is the foundation of finance. At the base layer of a global financial system, everyone must reach consensus. If we can't even agree on the ledger of account balances, or the basic state of the system, then everything else is moot.

How does a decentralized validation network reach consensus and eliminate the single point of failure risk of traditional regulated trading platforms?

Whether there exists a regulated entity to deploy these markets is an orthogonal question. In fact, I believe there definitely will be in the future, and it's not either-or. Even if a licensed single entity has authority over the markets they deploy, you can still enjoy the resilience of an on-chain network.

If you want to operate a 7×24×365 trading platform, that's no simple matter. Any failure in the matching engine, clearing, or any third-party vendor will create a single point of failure for critical financial infrastructure.

If the underlying rails are composed of a diverse, distributed set of validators, where one or even multiple can go down at any time while the network still maintains state consensus and continues forward at normal throughput, then you suddenly have a much more stable foundation on which to build a regulated trading platform.

Take buying $1 million in Bitcoin perpetual contracts as an example — how do order matching, collateral, and on-chain clearing mechanisms work?

First, to reiterate: as a U.S. user, you should not open this position. But for any user in a qualifying jurisdiction, the core mechanism is no different in essence from futures on traditional trading platforms: there is a counterparty in the order book, and liquidations are rule-based, determined by the clearing system, not by someone's subjective discretion.

What's unique here is that the clearing system is entirely on-chain. The thresholds that trigger liquidation and the execution mechanisms are similar to the traditional "default waterfall" arrangements of U.S. regulated futures trading platforms, but it is autonomously executed on-chain.

The HyperCore execution module — think of it as a dedicated submodule of the blockchain state itself — autonomously executes clearing and liquidation mechanisms.

HIP-3 Asset Deployment, the Deployer Role, and the Dutch Auction Mechanism

How do real-world asset perpetual contracts like NVIDIA stock get listed? What role does the Deployer play, and how are oracles defined?

We use "deploy" rather than "issue" because "issue" carries too many other connotations. Deployment is a technical process through which anyone can use this network. On Bitcoin, you can only move Bitcoin around; on Hyperliquid, because it's a system built to power finance, the functionality is far broader.

Deployers come to the network and stake the required amount of native tokens, currently 500,000 HYPE, not $500,000 worth of HYPE. Staking unlocks the ability to deploy directly on the blockchain.

On top of this, deployers define an oracle system through which perpetual contracts track the single exogenous number outside the system: the index price. This is the beauty of perpetual contracts: they are expressive and powerful, yet ultimately boil down to tracking that one reference index.

In the blockchain context, oracles are published by one or more entities that track off-chain data, which cannot be derived solely from on-chain features. In the future, as spot markets migrate to Hyperliquid, oracles may no longer require external publication and instead be derived natively from on-chain attributes.

How do deployers like TradeXYZ launch contracts, and how are responsibilities divided between them and the protocol?

Deployment is as simple as submitting a single transaction to specify the contract specifications. In doing so, deployers inherit the network's existing order book technology, liquidation architecture, and integrated USDC infrastructure.

Take TradeXYZ as an example: they maintain a publisher system to push oracle prices.

The real beauty lies in the modular division of responsibilities: the protocol provides the rails for execution and liquidation, while deployers specify contract parameters and maintain the oracle. Together, they create an attractive end product for users. This modularity is unique to on-chain financial systems.

Where does liquidity for these newly listed trading pairs come from?

Liquidity is entirely driven by the free market; it is simply an order book.

Anyone placing an order must have corresponding collateral to back it. The protocol itself does not provide liquidity; it merely provides the mechanism for buyers and sellers to come together.

Are multiple parties allowed to deploy the same asset? How does the on-chain Dutch auction allocate scarce deployment slots?

Anyone can deploy; there are no restrictions at the protocol level on what a perpetual contract must look like. The protocol is a basket of tools, or primitives, that deployers use as they see fit.

However, deployment slots are a scarce computational resource, because every active order book and liquidation engine consumes network compute. To allocate these slots efficiently, we implemented an on-chain Dutch auction.

Currently, the auction takes place approximately every 31 hours. The 31-hour interval was chosen because 31 is coprime with 24, so the auction reset time rotates across different global time zones.

The game-theoretic design of traditional auctions is flawed: bidders undercut each other by a penny, withhold information, and snipe or wash trade at the last moment. Dutch auctions are much better in an on-chain environment: the price starts high and decreases linearly. The first bidder to accept that price immediately wins the seat, so it is naturally resistant to wash trading. The clearing price then resets to twice the winning bid for the next round, creating a natural market discovery mechanism that pushes the price toward the fair market value of deploying a seat.

Financial Democratization and the Wealth Barrier of Tech Giant IPOs

Do you pay attention to who wins the auctions? How does a permissionless deployment mechanism solve the problem of traditional IPO wealth distribution being monopolized by a small group?

Because this is a permissionless system, I personally do not track who wins the auctions or what they plan to launch, although the community has built excellent analytics tools that can audit and visualize this on-chain data.

Overall, I find it very exciting to see markets that have historically been restricted and walled off become globally accessible 24/7. Giving retail investors and participants around the world equal access to financial opportunities is something I deeply believe in.

Look at this year's tech company IPOs: enormous wealth is created, but the process has been heavily gatekept. When companies list at valuations of $1 trillion, $2 trillion, or $3 trillion, the vast majority of the wealth creation has already happened in the private market. Ordinary people are excluded and only allowed to participate after most of the upside has already been realized. This is a fundamental flaw of the existing system.

Trading perpetual contracts on unlisted companies does not carry equity dividends or claims on cash flows. How do you view this trade-off?

Trading futures or perpetual contracts on an asset does not confer a direct claim on dividends or future cash flows, but the same is true of traditional futures.

I am not arguing that perpetual contracts can strictly replace direct share ownership, nor that everyone should trade these instruments. But there is a very large group of people globally who could benefit from gaining price exposure to these assets.

At the same time, parallel work needs to be done in financial education so people understand the risk characteristics. But if the opportunity to participate is completely blocked by gatekeepers, no amount of education will help. The first step must be to make the infrastructure fair and democratized.

Founder Background, HRT Experience, and Rebuilding from the Ground Up

You previously worked at the top high-frequency quantitative firm HRT and founded Hyperliquid in 2023 without venture capital funding. What was the reasoning behind that?

I went to Hudson River Trading (HRT) right after graduating from college. It was a great experience; they are an extremely successful company yet maintain the culture of a tech startup, and I learned a lot about market structure there.

When we started building in 2023, the vision was very focused: build something on-chain that people truly need. We don't view crypto as just memecoins. Centralized trading platforms offering perpetual futures have found undeniable product-market fit, proving that users value this product.

But centralized platforms expose users to counterparty risk from the trading platform and require handing over custody of assets. Our goal is to bridge this gap: provide the execution speed and user experience of a centralized perpetual contract venue while maintaining on-chain self-custody and transparency. Combining these two sets of attributes is what makes it a truly step-change improvement.

"Accommodating the entire financial system" — is that serious? Is there a possibility of introducing physically delivered commodity futures in the future?

Regarding the aspirational goal of "accommodating the entire financial system," we are extremely serious. Even capturing just 90% of global finance would be a massive success, and we are still far from there.

Physical delivery will always have an indispensable place in society. There is a misconception that our goal is to replace traditional market operators. On the contrary, our technology can serve everyone currently operating markets in traditional settings.

Whether a market is perpetual futures, futures with expiry dates, or physical delivery, the underlying needs are the same: order books, clearing systems, global distribution capability, and liquidity concentration. What we build are neutral primitives — tools — that allow participants to freely compose and build whatever they need.

What was it like writing the entire network's code independently? Why can't existing blockchains carry this vision?

We didn't see this as building a trading platform, but as building an entire network. At the time, we examined existing blockchains, and none could support the throughput and latency needed to "accommodate global finance" without unacceptable compromises.

It felt a lot like doing consumer tech in the early 2000s: you realize the underlying network infrastructure isn't sufficient, and you have to lay the foundational rails yourself, just like Amazon built AWS back in the day.

This realization meant the project's scope was much larger. And precisely because it's hard, it's all the more worth doing. We had to dig all the way down to the consensus layer, question fundamental assumptions about how distributed nodes communicate, and design a custom consensus architecture capable of maintaining a single global financial state.

What hardware configuration is needed to run a validator node? Does it require stadium-sized data center support?

There is a misconception that running this network requires a stadium-sized data center. We do not operate this network; it is run by independent validators. There are currently over 30 validators, with the top 27 by staked amount actively maintaining consensus.

The hardware requirements are surprisingly low. Recently, a validator remarked that it is truly unexpected that ordinary consumer-grade hardware can power a global financial network without needing an exclusive supply contract with Nvidia.

You cannot run it on a smartphone, but it requires about 128 GB of memory. You can absolutely buy a high-end laptop or standard server with this configuration. This network should remain accessible and cost-effective to operate, rather than requiring stadium-sized server farms.

Degree of Decentralization, Team Boundaries, and Trust Assumptions

The team rarely faces the media and has anonymous co-founders. Will the development team eventually fade into the background like the TCP/IP protocol?

I think the team is far more than just Hyperliquid Labs; it is a constellation of teams across the entire ecosystem. For example, Purr, that digital asset treasury company, has a highly visible public management team that directly engages with traditional investors. Deployers and builders are also part of this collective team.

As for developers setting protocol parameters: in the early stages of any blockchain, the software must be written by engineers who translate community feedback into code. Over time, this role can become increasingly decentralized.

The core philosophy of this network—fair, transparent, neutral primitives—does not belong to the developers. It is an embodiment of what an open financial system should look like.

Under the HIP-3 architecture, are users ultimately trusting the Deployer or the Hyperliquid protocol?

If this question is directed at Hyperliquid Labs, then the primary trust assumption is that the node software accurately follows its published specifications, just as Ethereum users trust the various client implementations.

Under the HIP-3 framework, the operation of a specific market is governed by the deployer, who sets the contract parameters and publishes oracle prices anchored to external assets. If a user is trading a gold contract, then their trust falls on that specific deployer to push accurate index prices. The underlying protocol always remains a neutral execution track.

Oracle Mechanisms, 24/7 Trading, and Hard Mathematical Constraints

Can perpetual futures trading volume exceed spot trading volume? What happened with the SK Hynix oracle price anomaly?

Regarding the SK Hynix incident, TradeXYZ published a detailed post-mortem. It was an implementation issue in their oracle system, not a protocol failure. They acknowledged that the design was not intuitive enough, issued refunds to users, and patched the system. Software systems across all industries encounter bugs and get patched.

As for trading volume, perpetual or futures trading volume can safely exceed the underlying spot trading volume without breaking the market. We see this in traditional markets: S&P 500 futures and ETF volumes are many multiples of the underlying spot stock volumes.

The key is that sophisticated market makers have the ability to provide liquidity against price movements that do not reflect fundamental value. If derivatives deviate 1% from fair value while the underlying spot volume is thin, market makers will trade against that dislocation, pulling the price back into alignment, thereby allowing true price discovery to occur on the derivatives.

If the underlying asset is too illiquid or difficult to price, will the protocol layer refuse to list it?

The protocol has no semantic understanding of whether an asset is reasonable or has liquidity. It cannot make subjective judgments, just as a traditional trading platform's order book software vendor does not judge whether a particular order book is appropriate.

What the protocol actually enforces are strict mathematical constraints. For example, the protocol has a hard rule: an oracle price update cannot deviate more than 1% from the previous update. If a publisher attempts to push the oracle price out by 2%, the protocol will automatically clamp it to 1% to ensure an orderly state transition.

The protocol enforces mathematical boundaries, not subjective market opinions.

When spot markets are closed on weekends (such as WTI crude oil contracts during Middle East conflicts), where do 24/7 prices come from?

How non-trading-hours prices are handled is defined by the deployer in their contract specifications.

However, for perpetual futures, having an oracle that continuously tracks the underlying price 24/7 is not strictly necessary. The funding rate mechanism will naturally tether the derivative to its underlying asset over time.

You can design a market like this: when the underlying spot market is closed on weekends, let the derivative price float freely; when the market reopens and funding fee payments resume, pull the price back into convergence. The mathematics of funding rates allow for a flexible market structure even during periods with fewer funding fee payments.

Institutional Access, Traditional Finance Integration, and the Compliance Pathway

Are there currently large institutional traders and market makers participating on the platform?

Yes. Many of the larger traders are institutions.

To be an effective market maker in a high-throughput order book environment requires considerable technical expertise and institutional-grade capital.

Could traditional asset management institutions like VanEck and State Street potentially become on-chain deployers within five years?

I don't quite agree with labeling traditional institutions as "old guard." The term "old guard" implies they will be disrupted or replaced, and that is not our vision.

Traditional asset management institutions have excellent client relationships, trusted brands, and provide valuable services. Adopting new technologies that improve their own products aligns with their natural business interests, just as traditional enterprises embraced AWS and cloud computing back in the day.

Seeing large asset management institutions become deployers on Hyperliquid in the future would be an important and entirely natural development.

Could a regulatory framework emerge in the future where "regulators constrain Deployers while the underlying protocol remains a neutral technology layer"?

That framework captures it very accurately.

Just as internet technologies like TCP/IP are used to build financial systems but are not themselves regulated, regulation naturally falls on the layer where discretionary control exists.

The ones exercising control are the deployers — they control contract specifications, listing rules, and oracles. This very naturally allows the protocol to continue playing its role as a neutral network and liquidity infrastructure layer, while regulatory scrutiny falls on the deployer layer.

What shift needs to happen among U.S. regulators to allow U.S.-based users to access without needing airport compliance workarounds?

First, if you're American and using the airport, you're violating the terms of service; that's not something to laugh off.

For this to happen legally, regulators like the CFTC must determine how this technology aligns with their core statutory principles. And those core principles are designed to be technology-neutral.

Regulators are conservative because they carry the responsibility of protecting an existing system that serves many people. But on-chain technology actually solves some long-standing regulatory challenges.

In traditional settings, verifying whether user balances match bank reserves requires an indirect, trust-dependent audit process. On Hyperliquid, however, anyone in the world can audit in real time, with provable mathematics, that the sum of all user balances exactly equals the amount of USDC held by the smart contract address.

When "verifying reserves" shifts from an intermittent administrative process to an intrinsic mathematical property, people begin to realize just how beneficial it is to integrate on-chain rails into the financial system.

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia