Fireside Chat with Hyperliquid Founder Jeff Yan - TOKEN2049 Singapore 2025

Yeah. So Jeff, like starting it off, why does Hyperlid work so well? Like your view very broad strokes. What is it that makes it tick? Well, I think there there are many perspectives like who who you ask, right?

The maybe for a user the answer is that the chain finally and the application really kind of are not fighting each other, right? I I think before maybe before hyperlquid there were there were some gaps in the UX for a pure just like UI user. You don't care at all about the tech. If you show them a centralized exchange and you show them a decentralized exchange, just go through the flow, there's it's pretty obvious that there there's some, you know, extra clicks needed and you got to, you know, wait for things to happen before you can do other things. And so, um, I think a big part of that flow was nicely uh, abstracted away from the user in in building the infrastructure from scratch.

So like a simple thing like you know basically delegating trading permissions to another signer was something that you know it's kind of no-brainer in retrospect everyone does it now but I do think it was something that needed to intentionally be built and yeah maybe for builders the answer is different so you know depends I think I did have have asked you this before but like you went into a very competitive space you know per per dexes per exchanges and uh you know you decided Ed, hey, I'm going to like out compete and do this better. Was there like a moment or a insight where you that gave you the confidence that you thought, you know, hey, I I can do this better than what's out there? I think the space looked competitive from the outside, but I think the problem with crypto is that it's so if you follow the the capital, it's it tells one story and if you follow the what's really going on like the product, it tells a different story. So we were very lucky that you know you know before before um building we were actually autom doing automated trading ourselves and so it was pretty obvious to us that there was there were still a lot of gaps in in DeFi like like even as an automated trader when you don't need to go through the UX there there were still already a lot of a lot of things a lot of inefficiencies in in the market due to structural issues with how DeFi worked And so that plus like once we spent some time looking at the the user experience side which I felt was like even you know yeah another level of undeveloped I think that gave us the confidence that there was you know capital was not telling the right story that was not actually in its final state and I still don't think it is. I think there's still a lot to do.

So job's not done job's not done. So your team and I mean again I want to kind of share with the audience how you actually build it. Um but the core team is still like 12 people or something.

I think 11.

Yeah.

But the team is not just just the core team. I I do want to make that clear. It's like I think the line there's often a clear line between the core developers of a protocol and the application developers on top. I do think the line is very fuzzy for Hyperlid and we want to make it as fuzzy as possible, right? We want it we want to we we do want we we do want people when they think about the team to think of all the builders and um yeah I give this example a lot but like spot trading is a core part of every verticalized centralized exchange product that I know of but on Hyperlid it's it's built independently and there can be multiple different teams bridging different assets minting different assets is issuing different assets And so I really think like like look at any other protocol those people would be a department in in the team and we have the privilege of just having talented builders self-organize and compete in the market and just uh execute.

That's that's all true. But to give you some credit too, you designed the incentive structure that like the the core team has designed the incentive structure that has incentivized the community to build on it too, right? Um I mean did you think of yourself as like an expert in incentive structure design going into this?

Definitely not. We have like a natural respect for the power of incentives. I think it's really it's really easy to underestimate that when you look at markets like markets are just a small set of rules and then everything everything that follows is a coralary of of those rules and I think this is kind of like the mental model of economics broadly like I never studied economics but it I I think that's roughly how people think about the world like that's that's how that's the only way you can model a world and so yeah incentives incentives drive behaviors and so I think there like there's like the technical design of incentives and then there's the sort of philosophical design behind them. The end user is like such an important component of the system and there's a lot of coordination that goes on and I think a lot of focus there but it's sometimes it's hard to forget that ultimately like why a market exists is to take user A and user B and kind of connect them in in the most efficient way possible. And so these end users are like the the bread and butter of the the system like really like if you treat the end users right then the whole thing works out.

And so that's not really like a technical thing but it it really informs a lot of decisions like down the pipeline.

Yeah. you know, like sure there's all the incentives, the economic incentives that you're expert at from just being a market maker for so long, but I would also argue there's like this softer touch that Hyperlid has around the ethos that I think is, you know, equally compelling um or motivating to many of your builders or community members. And I do find it I still find it kind of uh pretty fascinating that such a competitive hyper financialized product can still have such like a soft and welcoming community. And I do not know how you did that. Well, that's not us.

I mean that's that's the community. So we we just we just do what we do. And I I think it's we're very fortunate that you know in this very past dependent world that the I think the right people came at the right time and you know it's it's we we can only do what we do. We're only we're only you know a handful of people and we only have so much reach and we're actually very bad at you know even even getting the reach that we maybe could have. But the community kind of carries carries the torch forward right and I think that's we're very fortunate about that.

like like you're right. I don't I don't think in every possible outcome this necessarily would have happened. And I think there's there's very mercenary capital and attention in the space. And yeah, we're just we're just very lucky that the community is is as it is.

Okay, let's get um I mean I love that. Yeah, but let's being maybe a little bit more technical. Um, per uh where did you first like find or hear about or discover uh the perp product? Do you remember back then? It's it's hard to remember or it's easy for me to remember, but it's not talked about much just how revolutionary BitMX was back when it first came out.

I think they they really went zero to one. A lot of respect. I think I think a lot of crypto just owes you know owes a lot to just like them pioneering technology the the technology of per I think realizing it and kind of building it out at scale. And so I remember I think the first time I saw pers was just looking at this exchange that was doing so much volume and just looking at like what what what were people actually trading there and it wasn't actually the underlying it wasn't I mean there were some futures contracts being traded but by and large they were trading this other thing that was just so novel and I was yeah I remember reading the docs and being pretty fascinated by it. I don't think it clicked immediately that it was so in such an impactful like innovation, but I did I do remember thinking like wow this is like so inefficiently priced.

Like I remember that like if you looked at the dated contracts they were like all over the place and the like if you looked at the funding rate on the pers and the sort of premia and discount on the on the dated contracts they were very out of whack and uh you know my instinct was like well like this is probably a lot of smart people probably think about this and I'm sure it's efficient and I just don't understand why it is but but I do think it actually just was not efficient and then uh I think I think I checked a year later and it was all in line. So uh that was pretty cool. Yeah, I mean I had the, you know, you and I both come from traditional finance backgrounds and uh I had the same experience. You saw this per thing and you're like, "Wait, how are they doing that? How is this derivative tracking the underlier?"

Um, and then it's this weird funding rate payment thing that's actually pretty exotic when you think about it. But the retail trading experience is amazing because you just buy this thing and you can buy it with leverage. You can go long or short. Um, and you never have to roll it like a futures contract. So, um, yeah, I've got I got an angle to push on here, but like maybe we can spend a second talking about your take on pers versus futures versus options.

Um, particularly with this kind of view of hyperlquid as the blockchain to house all of finance, like where do you think those products uh excel or or or have limitations? Yeah, that's right. I think housing all finance is not it's not just a meme. It is it is really the the goal. And we are pretty we like to we like to be pretty unopinionated.

So I do think if if housing all the finance means options dated futures like I think these things will all exist um on hyperlquid. the the thing we've seen so far is that is there's both thinking from it academically and just looking at the data it doesn't seem obvious that like that much more than pers are needed. Okay, so maybe I'll take a step back like like you said like retail by and large in crypto I think has demonstrated that pers are the superior product and I think there are two main reasons for that. one is uh there's no fragmentation between different contracts whether it's like different strike prices different expiries or for dated futures um you know rolling on the uh settle near the settlement dates. So there there's just there's just one product and that's really powerful for just concentrating all the liquidity in one place giving the end user the the best price when they want to execute in size.

And the other thing is that it's just a lot more comprehensible to the end user because you can actually trade per without knowing about funding. And I think that is sort of the ideal like that's kind of the abstraction of leverage that the holy grail of leverage and and yeah so I think I think for those two reasons both for retail and for people who just want like delta 1 exposure especially if it's over a short time frame I think there's just there's likely no better product. uh don't want knock on wood like they you know don't want to yeah but but I I think there'll always be evolutions but to me it feels like per such a clear evolution over options and futures for these specific market participants. Okay. So, to tell the audience, right, I I was doing some chat GPT uh research ahead of this, just sort of sizing um what options on Robin Hood look like.

Um and I think something Jeff and I are aligned on, um options trading, uh equity options trading, most most users, I don't think they're interested in trading the volatility of equity options. They're just looking for leverage. Um and in traditional finance, you can't get leverage particular that easily, unlike with the per product. Um, and just to give you some math and kind of be bullish on this space, uh, Robin Hood trades about 200 million contracts a month. Um, which if you make some approximations around notional and delta, I think that looks about like 50 billion notional a day.

Um, uh, give or take. Um, and I don't know my take, I don't know what Jeff's take is, but my take is, uh, equity per would be a better way to to trade a lot of this type of leverage exposure. Um, so, you know, I don't know that you can comment on that or would want to.

Yeah, to comment. I think the I view basically like per as the best way for the best way currently known and known to work at scale for a bunch of people to basically create a very liquid market for price discovery and speculation on a single number when that number is continuous basically evolves continuously over time. I think that's like if that's if that if if the thing you want to do boils down to that then pers are a great a great instrument a great mathematical formulation to achieve that goal and I think if you just look at what tradi does a lot of it really does boil down to that and I think the so like specific examples aside I just I just think there will always be a need in the world to allocate capital efficient ly and to price things correctly. And I'm I'm just so bullish on everyone basically realizing that pers are kind of getting the green light. Both both sort of checks out mathematically and the the institutions are realizing the regulations are are sort of like coming to terms with like hey this is actually an innovation a technical innovation that we can we can apply broadly to markets and make make things more efficient.

And uh that that is something I'm super excited about and I think Hyperlid has has an important role to play there. Um and yeah, I don't know if you want to talk about that.

Well, no, I want to like I think you said something pretty insightful that I would summarize as a lot of Tradfi is betting on when whether a single number is going up or down.

Um and PERS are a very beautiful way, very beautiful product to bet on that number. Um, and right now pers are crypto prices. Um, but here we segue in talking about HIPP3 where that goes. But basically, if you have permissionless per deployment, hyperlids coming for Tradfi, right?

I wouldn't I wouldn't frame it that way. I think it's I think of it more as DeFi is is not a something something coming in like a force a force to come in and, you know, eat Tradfi's lunch or something. I don't think I don't really don't think it's healthy to think of it as a as a one or the other, but rather I think the DeFi is is it's basically a a tool. It's a it's it's an upgrade to the existing technology that the traditional financial system happens to be using right now. I think there's there's not a decentralized system and a traditional system.

there is just one financial system and

I don't think

yeah I don't think I don't think it's productive to think of it that way. I think what we all want is efficient markets for everyone and that that leads to more prosperity for the world and I I think at least at least from my perspective what hyperlquid is building is something that you know users shouldn't care who's coming in and deploying markets they should care about the quality of of the markets and the liquidity and uh I I think it will happen. I I see this shift this mindset shift kind of happening over the past few months and I hope it it continues.

Yeah. Yes. And per are a beautiful way to they're a beautiful product.

Ignore Tradfi DeFi. They're a beautiful product that retail investors love and that Hyperlid is very good at building um HIPP3 um permissionless per deployment. Like I I mean first of all maybe describe it for a second for the audience of like what's happening there. Um and then like what are you bullish on? What are you excited about?

Um, yeah.

Yeah. So, it's basically a way for anyone to deploy per markets without reinventing the wheel. The wheel here being a blockchain, assets of value, secure on the blockchain, um, consensus, a global consensus around what the state of the blockchain is. And then the things that are unique to Hyperlquid, you know, performant natively implemented order book margining systems. Um, it's there's a lot that goes into building a PER's back end.

And I do think that if if PERPS are great, but in order to in order to really provide a good product for your users, you need to build all of these things. Then I think most people most existing institutions probably don't view that risk favorably. But if it's like, well, here's a decentralized ledger, which is an upgrade because it allows permissionless global access to the markets that I'm deploying and I don't even need to think as a deployer. I don't even need to think about how the order book and margin works. All I need to think about is the spec of the market I'm deploying.

Then all of a sudden, the I think the work you need to do goes down several orders of magnitude. And I think that kind of that's just kind of core to how we think about how the decentralized financial system of the future should be built, which is that people should not be reinventing the wheel. They should be they should be basically like the infrastructure should be perfected by just like once and then it should be it should be neutral and it should allow anyone who's competent to come in and leverage that infrastructure to see their vision through.

Yeah.

The the competent point is actually kind of hard to do with some of these permissionless systems here.

Yes. Um, do you have like what's the strategy to keep it keep it clean? Like one of the things that I think you you care deeply about is like there's not a bad bad user experience here.

Yeah, I think it's it's important to have the the two which are a bit at odds which is that the system needs to be neutral. I think this is like very core to Hyperlquid's ethos and to all of DeFi's ethos which is that it should it should be neutral. should be permissionless and it should it should be kind of welcoming like you said is a bit of a subjective thing. Um but on the other hand the products need to be good and I think I think how how this played out with spot was was super cool. Um like the the primary spot assets that are traded on hyper liquid are built by a completely independent team called unit and you know they didn't actually launch immediately after spot came out.

I think it was many many months later and they have their own network basically securing these assets. Um I'm not going to get too much in detail there but you can definitely check them out. But if you go on hyperlquid.xyz for example it shows like you know if a user wants to trade bitcoin they don't need to know all this behind the scenes like it is just the interface can abstract it away for the user. And I think for HIPP3, what I what I think probably will happen is that users will gravitate towards interfaces that abstract things away well.

And I think this is a very I basically think that the free market will solve this problem. Like just because thousand people can come and deploy their markets doesn't mean a thousand people have equal opportunity to kind of be in front of the end user, right? you kind of ultimately the users go where trust is earned and trust is hard to earn and very easy to lose. So uh I think the incentives around all these things will sort themselves out and so I don't know what the final picture looks like but I think there will be it will be permissionless to deploy and there will be very high quality applications um you know in the in the bullish case I guess. Love that.

High quality pers on most things bringing all of fin housing all of finance. Um builder codes. So I mean I think builder codes are super cool because again it's yet another way for you to scale hyperlquid with kind of incentivizing others to to build on your platform. Um maybe it's worth just giving the audience an overview of like builder codes and also what's working well there. um and what you're excited about and maybe we could go back and forth on like where they're not working perfectly.

Yeah, happy to.

Um builder codes are another very simple idea. It's similar to HIPP3. It's like how can incentives and the protocol be designed to let builders really run with their idea and like set them up for success. And so the idea behind builder codes is that anyone can create interfaces and as long as you know as long as a user opts into it they can charge the builder can charge a fee on top and it's a simple idea but done permissionlessly and at scale I think you get a lot of really cool network effects like like recently Phantom um launched their purpose product and you know you know Phantom's a super super good team. They're obviously like very cryptonative and understand a lot about the space and their users, but they they didn't previously have a per product.

And it kind of goes back to the same point about HIPP3. I think if Phantom were weighing well, should we should we build a PERS deck that I don't I don't know I don't know their team personally and their strengths, but there's very good chance they would say, oh, like it's not worth our time because this is kind of does not synergize with what we've what we are good at. But with builder codes, the question now becomes, well, can we do we have a unique distribution and do are we are we uniquely positioned to offer an interface to users that adds value to them? And in this case, the answer is yes. They have uh amazing distribution across all of crypto and they have a mobile app that is, you know, beloved and they have they simplify trading experiences for users.

And so this is such a natural synergy where they don't need to they don't need to go and rebuild the per stacks. they can focus on their part which is which is a you know very hard problem in and of itself and when the two things align then you know hyperlquid benefits because there's more volume and more volume creates more liquidity and generally the you know just you know more users and network effects and for Phantom their users have a great experience because they when they want to trade massive size they don't you know they they they have the peace of mind that their order is hitting uh the most liquid decentralized liquidity in the world and I think yeah that that maybe that's like a concrete example of the good case and you know happy to talk about doesn't work. I also want to say I think it's so cool when you combine that with HIPP3 because you can like imagine just to theorize that there's a new mobile app feels like a fintech app that has some new per products that is getting paid but using your back end. Um it's pretty wild pretty cool. It's a neutral back end.

So we just think of infrastructure as encompassing a little bit more than just consensus and literally agreeing on some bites in the world. like we we view for finance to go on chain it needs to be performed like I think people forget just how much traffic processes um and for that to happen just as a practical matter infrastructure needs to include liquidity and by here I mean you know per performant order books margining etc like these things are um very very hard to build on a on a VM like it needs to be part of the infrastructure yeah so I I I think the um the part where like as many people know there's a lot of um hyperlquid copycats at this moment too and it's sort of funny cuz like in a different world they could have just built businesses using builder codes um but I guess any take on why you think some of these other folks want to own the full stack? You think it's just profit motive?

Um well I can't speak for the other teams. I do think you know it's it's a free market. I think competition is good. I do think often times probably what happens as industries mature is that in a free market people do like some people do just try to reinvent the wheel and maybe they can reinvent it better and maybe they can't but it's I think it's it's an opportunity that everyone should have and if if you go to market today and you say like you know I I don't want to build on AWS like I want I want to like put my own machines in a data center and you know figure out the electricity and the uptime for my user. and everything like that's that's a decision that everyone has and you know some there are some success stories there but by and large I think as a market matures the the sort of opportunity set shifts and so I think it's just I think it's just points to the fact that it's crypto is still very immature and early and I think um I'm happy to see these sort of maybe they're copycats today like by and large but maybe someone innovates and does something really good for the market and maybe happy live can learn from that too.

Yeah, that's the kind of humble response I'd expect. Um, but no, I think it's also accurate. Competition keeps you honest. All right. Uh, not too much time left.

We'll go maybe a little bit over. Um, stable coins. So, um, in addition to being like CEO and CTO, I also think you're a marketing genius. Um, because the whole USDH uh, governance process I thought was incredibly bullish marketing event for Hyperlid. Um but maybe just actually tell us what uh what a hyperlquid aligned stable coin looks like and what you think what advantages where where you think that can go.

Yeah. So going back to housing all the finance I think we we we focus a lot on pers today and I do think people think of when they think of hyperlquid they often think of pers that is changing a bit right they you know there's the EVM there's you know there's like a culture being created with with NFTTS I think um so I think that is changing but by and large people think perance is much larger than than per than trading itself even so I think stable coins are obviously a very hot topic I do think some of that is potentially, you know, frothy, but I do think there's some substance here, which is that Tradfi is embracing crypto, crypto rails. I think um similar to how we think Hyperlid should be infrastructure to let the existing financial system upgrade its technology. I think that kind of viewpoint is becoming more consensus at this point and stable coins are just a very digestible way for massive institutions who are already at scale to feel safe about kind of like dipping their toes and in into this like new this whole new like um unfamiliar technology stack and and I think there's good reason because stable coins are a simple thing to explain. And I think the value proposition is very clear and the and the sort of existing adjustable market is very clear which is that the you know by and large is a dollar stable coins and the US dollar is is the world's reserve currency and it is in and of itself uh a very important financial product and I think stable coins are kind of like an upgrade of that and so uh yeah if hyperlink was going to house all the finance it needs to really be the place that the stable coins that sort of win the stable coin wars or whatever are going on.

Um they they they should be built on Hyperlquid because this uh this is the platform on which transaction volume should happen and sort of um I I see a lot of really big businesses being built around payments and sort of upgrading banks. I think these are just like basically like really allowing global access to a great product which is the US dollar that everyone kind of there's genuine demand for um I think is massive and the idea behind it the alignment I guess sorry it's like kind of a long answer um is that hyperl right now versus where it should be I think there's still like a massive massive gap and um to get there I think basically Hyperlid needs to leverage its very unique distribution which is kind of cryptonative um liquidity as sort of like demand for liquidity at size like this is like a very unique distribution um for just genuine demand for finance and leveraging that to actually build these like massive payment products and sort of neo bank products I think is it's not a short journey but I think the idea the idea behind the alignment protocolized alignment is to make that path clear and then I'm really excited to see um amazing teams step up and kind of like embark on that journey.

You know, it's interesting your answer is not totally what I would have expected like it wasn't like hey we need like a stable coin that earns us the interest for hyper for hypeolders. Um the answer my my take was like we need to house all the stable coins cuz we want to house all of finance not all of them but that fine

that there if there is one that makes sense to build a hyper liquid which I think there are at least there's at least should be at least one like massive business to build here but probably many that these uh yeah that that there should be a path forward basically.

Do you worry about the user experience at all cuz like right now today one quote currency is kind of nice. Um, I mean, we've talked about this before, but like once you have multiple quote currencies, do you worry at all that that's just gonna confuse more of the normie users?

Yeah. And I will say like that, you know, one currency now it's it's not I think there's a bit of a meme here, but I think Circle actually does a a super good job and like they, you know, Hyperlink was like very very different from other blockchains that um sort of multi-chain solutions plugin do. So I I do think it really is just like the the fact that Circle can embrace Hyperl Hyper EVM as a first class chain in into the CCTP network. I think it's just like you know it's it's it's humbling to us like that that it it sort of it indicates the level of adoption and and demand. So I I don't think there's like anything wrong per se with with the existing system.

It really is uh yeah it's it's just it's just zooming out a lot. Like we're not we're not trying to microoptimize anything here. It's it it's just like there are just massive businesses yet to be built by anyone and we think Hyperlid is it's very aligned with the the vision of Hyperlid and so yes there'll be some fragmentation but also I don't think I think similar to the answer to the the question about whether it's spot permissionless deployers or um HIPP3 you know deployers being permissionless I think I think there are similar effects here where um if things are designed right then the the free market kind of like figures things out and the end user can simultaneously benefit from it having been and continuing to be a very competitive and open system while still having the sort of the concentration benefits of liquidity.

Yeah, I think there's something very thoughtful and very insightful maybe to spell it out with the for the audience a little bit. You know, I think some of these hyperlid front ends are going to pick a stable coin um because of builder codes or incentives or alignment. And um you got the free market now figuring out which the right product or or or kind of permutation of features is um all in a way that benefits this incredibly neutral um backend infrastructure. So guys, thank you Jeff. Um thank you for doing this and um Hyperlid