Kanav Kariya – President, Jump Crypto Ep #65

EPISODE SUMMARY

Kanav Kariya (President, Jump Crypto) joins the Solana Podcast to discuss his optimism for the future and the many areas in which Jump Crypto is innovating in the crypto and blockchain space. Austin Federa (Head of Communications, Solana Labs) guest hosts. 00:49 - What is Jump? 03:07 - The path to operationalizing crypto 06:00 - Optimism for Crypto 10:49 - Discovering and Building in Crypto with Jump 14:24 - Personal Journey at Jump 16:43 - What's being built at Jump? 17:55 - Reasons to want to build 19:39 - What does Pyth offer? 22:22 - Criticism about conflict of interest 26:30 - How Web 3.0 facilitates resource coordination 28:46 - Data contributors benefiting from onchain data 31:01 - Token Plans for Pyth 31:46 - Message bridging 34:48 - Wormhole, stable coins and asset tokens 37:36 - Time synchronization for cross-chain dApps 39:14 - State storage on wormhole for dApps 40:21 - Is Wormhole layer 0? 41:14 - Wrapped NFTs 44:13 - Jump's position towards NFTs 48:36 - Exciting things in the ecosystem 49:43 - Custom silicon / FPGAs 53:22 - A parallel execution model? DISCLAIMER The content herein is provided for educational, informational, and entertainment purposes only, without any express or implied warranty of any kind, including warranties of accuracy, completeness, or fitness for any particular purpose. Those who appear in the content may have a financial interest in any projects referenced, and any content herein is not intended to be and does not constitute financial advice, investment advice, trading advice, or any other advice. This content is intended to be general in nature and is not specific to you, the user or anyone else. You should not make any decision, financial, investment, trading or otherwise, based on any of the information presented without undertaking independent due diligence and consultation with a professional advisor.

EPISODE NOTES

Kanav Kariya (President, Jump Crypto) joins the Solana Podcast to discuss his optimism for the future and the many areas in which Jump Crypto is innovating in the crypto and blockchain space. Austin Federa (Head of Communications, Solana Labs) guest hosts.

  • 00:49 – What is Jump?
  • 03:07 – The path to operationalizing crypto
  • 06:00 – Optimism for Crypto
  • 10:49 – Discovering and Building in Crypto with Jump
  • 14:24 – Personal Journey at Jump
  • 16:43 – What’s being built at Jump?
  • 17:55 – Reasons to want to build
  • 19:39 – What does Pyth offer?
  • 22:22 – Criticism about conflict of interest
  • 26:30 –  How Web 3.0 facilitates resource coordination
  • 28:46 – Data contributors benefiting from onchain data
  • 31:01 – Token Plans for Pyth
  • 31:46 – Message bridging
  • 34:48 – Wormhole, stable coins and asset tokens
  • 37:36 – Time synchronization for cross-chain dApps
  • 39:14 – State storage on wormhole for dApps
  • 40:21 – Is Wormhole layer 0?
  • 41:14 – Wrapped NFTs
  • 44:13 – Jump’s position towards NFTs
  • 48:36 – Exciting things in the ecosystem
  • 49:43 – Custom silicon / FPGAs
  • 53:22 – A parallel execution model?

DISCLAIMER

The content herein is provided for educational, informational, and entertainment purposes only, without any express or implied warranty of any kind, including warranties of accuracy, completeness, or fitness for any particular purpose. Those who appear in the content may have a financial interest in any projects referenced, and any content herein is not intended to be and does not constitute financial advice, investment advice, trading advice, or any other advice.  This content is intended to be general in nature and is not specific to you, the user or anyone else. You should not make any decision, financial, investment, trading or otherwise, based on any of the information presented without undertaking independent due diligence and consultation with a professional advisor.

Austin (00:10):

Welcome to another episode of The Solana Podcast. I am Austin Federa, sitting in for Anatoly again this week. Today we’ve got a pretty special episode I think. I’m really looking forward to this conversation. I think it’s been a long time coming with a few false starts. Today we have Kanav Kariya president of Jump Crypto, or do we just say Jump at this point?

Kanav (00:32):

Yeah, Jump Crypto is good.

Austin (00:34):

President of Jump Crypto, which maybe this time last year very few people knew existed, very few people knew what you guys were doing, what you were building, what your role in the ecosystem has been. So yeah, I guess let’s just go ahead and Jump right into it. What is Jump Crypto and how did it come about?

Kanav (00:51):

Yeah, thanks for having me on Austin. So for context for the audience that aren’t very familiar with us, Jump is historically a prop trading firm founded over 20 years ago in the pits at the CME. Today one of the largest quantitative trading firms in the world. And we started a crypto division over seven years ago. It started as an intern project at the University of Illinois, where we were running a miner in a closet and building some trading infrastructure.

And today we’ve got over 150 people on the crypto team doing a lot of different things. So the way I like to describe our business is spitting it into three primary pillars. One is prop trading, which is exactly what we do on the other side of the house, we build trading intelligence and we scale it. The second piece is building and that’s the piece that I hope we’ll get to talk a lot more about on this call and it’s closest to my heart and closest to the heart of the team.

And that’s in building pieces of infrastructure, really streets and sanitation for the space and a couple of the marquee projects that we’ve really focused a lot of our efforts on have been Wormhole and Pyth. And of course, along the journey, we’ve aligned ourselves with a lot of the major ecosystems in the place, including Solana, Terra and a whole number of others in building a lot of different things across those platforms.

The third bucket is venture, I like to call ourselves accidental VCs in that we found opportunities to add value, or we had requests come in to work with partners over the last six years in various different capacities. And we found that we could be meaningful in those contexts and work with people that were solving problems for us. And that has now grown into the venture division that’s deploying across the space.

Austin (02:31):

I want to get into a lot of the work that Jump is doing as core code contributors and supporters of projects in the ecosystem. But I kind of want to start a little bit with that journey. I would say that the transition from prop trading equities and commodities to prop trading crypto, that feels pretty organic. And there’s a number of firms in the space that have also made that transition. Albeit you guys seem to have made it sooner than a lot of other firms in the industry. What was that process like of going from deciding that you wanted to add crypto to actually operationalizing that? And then we’ll get into some of the journey to actually becoming builders.

Kanav (03:07):

The project started as an intern project at this thing called Jump Labs. There was a research lab at the University of Illinois and was meant to work on cool stuff with the university on working on fun problems. So alongside the crypto stuff we were doing when I was an intern, there was a VR project working with professors at the university to abstract away trading screens. And there was work on some interesting machine learning and networking problems.

And the group has grown out of that. And of course matured out of these things, but we’ve definitely strongly retained that ethos. Now I want to caveat this by saying we definitely didn’t have oppressions in being infrastructure builders. When we started the project in the lab that many years ago. It’s been a very organic and natural process for us. And it’s hard to make the instant leap from prop trading to what we’re doing today, but it’s easy to reason through the steps along the way.

As one of the earliest large trading firms in the space, we had a lot of requests from institutional liquidity exchanges, OTC platforms, and importantly projects that were looking to solve trading and liquidity related problems. And those conversations gave way to us exploring a lot of DeFi projects and a lot of L1 platform projects that shared a lot of the problems they were thinking through on complex financial system design or programming in resource consumer environments, which are very natural and germane to a quantitative trading firm.

 

And those conversations led to jamming about foreign ideas to implementing governance proposals, to maybe starting to write a little bit of code in them. And then all the way into committing over 50, 70 engineers that we have today in building through the space. And that process involves a few different steps. One, it involves the willingness for the institution at large to be mentally long the space. It requires a recognition and frankly a little bit of a taste of the upside.

It requires flexibility, which of course, prop trading firms just generally naturally just have to have. And then everything else you can just learn along the way, right? We’ve done a lot of things wrong. We’ve stumbled over ourselves a hundred times, but you’ve got to keep digging shots on asymmetric upside and with all the resources that we’ve had at the firm I think we’ve been able to make some good ones.

Austin (05:20):

Going back to you last year, Jump Crypto had sort of a moment where it decided it wanted to make itself public. You wrote a blog post that was laying out. I wouldn’t quite call it a thesis, but laying out an idea of how you view the space and the role that something like Jump could play within it. One of the things I was struck by going back and rereading this is your level of optimism in this post, right? Which is something that you don’t see from many financial trading firms. You see them seeing opportunities to make lots of money. You see them making lots of money. They’re very profitable endeavors, but you usually don’t see optimism contained within it. Where’d that come from?

Kanav (06:01):

That’s a pretty good question. So quant firms today are basically research and development firms, right? So the people that build trading systems, that build the intelligence behind trading systems are generally of quantitative background. They generally have PhDs in either statistics, machine learning, physics, those kinds of endeavors. And the people building the platforms are low latency high performance systems engineers that there are different optimizations across every level of the stack to build robust, scalable, fast infrastructure.

The environment down to the lab five years ago was about exploring this space. It was like, what does this space mean? Right. And it wasn’t about, okay, how are we going to make X billion dollars kind of getting into this endeavor? It was about exploring it. And I think it attracted that kind of people and it occurred that kind of environment.

And the leadership that stays since then has kind of embodied that. And just personally I’m a raging optimist, I believe in technology, I believe in the future, I believe in building towards something bigger. And thankfully I think the firm has shared those ideas and I hope I’ve been able to shape a lot of the culture and behaving that passion.

Austin (07:10):

Where do you think that optimism in yourself comes from? There’s a lot of things you could have gone into coming out of school. What about both, something, an organization like Jump, which is undoubtedly a great place to go work. But you stay there for a while now, you’ve worked your way up, you’re now in charge of the crypto division. Where does that sense of optimism in you come from and what makes Jump the right place for that?

Kanav (07:33):

I feel something for Jump because they had a cool internship program and they had a lab on site and they were working on really fun problems in a well resourced environment, that just made it fun and attractive. And after I had the opportunity to intern there for eight to 10 months, I kind of got a sense for the possibilities that existed. And this is the flexibility that the whole space had. And it was like, you come in, you get to make a lot of bets, you get a lot of resources. And if you make good bets, you get more resources and then you get more resources. This is the only place I’ve ever worked. I think it would be rather unique to have that kind setup. And again, no, I wouldn’t say it was a passion moment to come in to Jump and know that I would be able to build suites and sanitation for crypto. But I knew I would get to do a lot of really cool stuff, work on fun problems with smart people. And where does optimism come from?

Austin (08:25):

Yeah. I mean, you look at a space like this. It’s been through boom and bust. There’s tons of amazing projects being built in the space that end up going nowhere. And especially from the vantage point of a trading firm, right? One of the secret sauce of a trading firm is it can make money in an up marketing, it can make money in a down market, right. And that is the advantage of a professional trading operation versus a more passive trading operation. But again, like those are not usually characteristics that breed optimism. Those are usually characteristics that bleed margins, where you’re optimizing 1%, 2%, 3% here. So you can compound that over a year and it will make a marginal difference. But again, that’s not usually an optimistic space, that’s a very functional space to work in.

Kanav (09:10):

Yeah, it is. And traditionally I don’t think it lends itself to naturally just exactly this. Jump culture has kind of always been a little bit unique. So Jump also has a number of other kind of divisions that work on non-high frequency trading stuff. Historically, since about 2011 or 2012, had a VBC arm called Jump Capital that invests in growing technologies in this space. They’ve had some cool endeavors in the biospace working on automation there in healthcare.

And so the founders have generally been optimist. They definitely believe in the future. They’ve been able to take shots at things that are going on. And even if it’s not naturally germane to the trading business in and of itself, the culture itself lends itself to being able to do something like this, which is a really awesome combination of knowing how to monetize, but then also knowing how to build. Yeah, it’s been an absolute pleasure to be able to soak in from that environment.

Austin (10:04):

Let’s look at the building for a bit. I think it’s pretty open secret at this point that Jump are core contributors to Wormhole and Pyth, you’ve been very heavily involved in that process. Take me back to some of the early days there where you are internal to Jump, and you’re saying like, “Hey, we need to do more than just trade and invest in this space. I think we can actually build.” And especially you’re talking about this from the perspective of sanitation and roads and the very base level infrastructure. Crypto’s been around for a long time. I think most people coming into the space in that time horizon wouldn’t have necessarily looked at and said like, “Oh, there’s very base level features that are missing from this ecosystem.” What was that both discovery process like, and then the process of convincing everyone internally that this was worth dedicating resources to?

Kanav (10:50):

Yeah, the discovery process was very organic. We had a lot of inbound from people looking to solve trading and liquidity problems because a lot of people in the space, even though we were quite kind of new of our trading presence, and as one of the early trading firms that really was trying to make bigger pushes in the space. When you get to talk to awesome founders every day about all the problems that they have and get to build relationships with them, you start to uncover a lot more of the problem space that exists, start to internalize a lot of it.

And once you’ve got the opportunity to sit in that for a little bit, and I’m sure you see this today. We are much later on than we were when we made a lot of those big switches, but there’s still a lot of opportunity, right? When we were kind of ideating on the origins of Pyth, the conversation we had was, look, our whole thesis at Jump Crypto is to be as long aligned with the space as possible, right? We’re trying to get the maximum exposure we can on the space that we think is going to be explosive. And we’re trying to ideate this ways which we put that quote unquote trade on, right? The best way to put a long trade on in a growing space, and the best mode to value capture is value creation. There’s definitely a lot of inefficiencies created by hyper growth, right? And there’s room to capture those inefficiencies. But those are small in magnitude relative to the absolute value creation at play.

And then there’s a value creation capture correlation that you think about there. So if you think about it in that lens and you know that you want to be big contributors to the space and just aim to create a lot of value to both, then you start thinking about what the opportunities are within your realm to be able to engage in that capacity.

Austin (12:27):

But at some point there’s a meeting, or you have a boss who you report to, and you have to go down and sit down in front of him or her and say, “Hey, I want to spend a lot of money to hire a lot of engineers to do something that’s going to be totally public and totally open source at a firm that historically likes to stay out of the news.”

Kanav (12:46):

It was a few meetings.

Austin (12:46):

Yeah, I’m sure.

Kanav (12:46):

And it’s kind of baby steps along the way, or big steps along the way that compound into a complete shift and a big switch of that nature. We had this summit, we called the August summit a few years ago. And we went down to an offsite location and we talked about what being in this space means for us and how we differentiate. And I remember we showed up with these sheets that we went around and distributed to people. We were like, this is the toolkit that we have. This is the opportunity set in the space.

And everyone kind of had their own, things went on, but that was one of the approaches that I’ve taken. And if we believe this is where the space is going, this is the opportunity set that we can tackle. And these are the levels that we have to pull, right? And then you socialize that and you try to convince them people that there is opportunity to be had here and you get buy-in to take a first little step. And once you get the buy-in to take a first little step, and you kind of really show the big medics of differentiation in a native space, you get the buying for the next step.

And then suddenly it’s the entire [inaudible 00:13:47]. You get the whole kitchen sink thrown behind you, and then you are kind of propelling to this part that you want to be at. And that’s the whole thesis of Jump everywhere. You take bets with asymmetric upside and we throw the kitchen sink at things that are working. And a lot of the stuff that we were doing started working.

Austin (14:02):

How is that journey for you personally, going from an intern involved in a few projects now to the Jump Crypto teams over a hundred at this point?

Kanav (14:11):

Yeah. We’ve got over 150 now, hard to keep track.

Austin (14:14):

Wow. Yeah. From a leadership role, and from your own perspective, how has that transition been? What parts of it were easier for you? What parts were harder than you were anticipating? Scaling yourself is often much harder than scaling a company.

Kanav (14:28):

Without a doubt, yeah. I started in the team as an intern like you pointed out, working on software problems. I came back to the team a year later in a formal full-time capacity, working on quant problems, which was to do with predicting crypto markets, building alpha and kind of scaling that piece. And the early conversations with projects where we were trying to solve liquidity problems was an area that I got really, really interested in. And I just kind of went about trying to build that a little bit further.

Over time that led to a transition from engineering and quantitative work to more conversational business development work, just having spent years across all those functions and natively knowing how to live them has been the biggest tool that I’ve been able to build in the toolbox. Now that doesn’t teach you how to manage a hundred people, that doesn’t teach you how to propagate culture. It doesn’t teach you how to scale hiring strategy. Doesn’t teach you how to value the troops when things are low.

I definitely want to make a claim that there are many who are close to a finished product, rather than trying to be good at everything, good at every one thing, we always try to be excellent at a few things. And then by force just propel everything forward. I’d say some of the biggest lessons I’ve learned, the biggest mistakes we’ve made, definitely been in the shape of trying to shove square bags in a round hole. Where in a trading environment it’s like the only people you have on your team are engineers and quants. They’re just smart people that can solve any shape of technical problem you throw them at. When you move that towards sales and marketing and product and everything else, that all kind of falls apart.

Kanav (16:05):

And you need people that are able to natively live within specific sub domains across those functions. And that’s something that we’ve been trying to scale in. I spend basically all my time hiring and trying to focus on making sure our zero to one projects have a lot of momentum. But yeah, it’s been an awesome journey. And of course I have support from a company that’s grown to a 1500 people as the largest quant trading firm in the world and so lots of guidance and help along the way.

Austin (16:33):

Let’s talk a little bit about that work you guys are doing and actually building. So if I understand correctly, the two projects that you are mostly core contributors to is Pyth and Wormhole. Is there anything else that you’d put into that category of engagement?

Kanav (16:46):

That’s the highest level of engagement for sure. We do a lot of things across the big ecosystems of course. We can talk all of what we’re doing with Solana. We’re always trying to get deeper. We built an NFD project on the Metaplex landscape after their investment as an intern project. That was a real fun one. We’ve been core contributors to some of the projects that are coming out on the data landscape today. We’ve worked on a lot of the mechanism design that goes on, on the other one. And there’s a few other projects, but the highest levels of engagement have definitely been with Wormhole and Pyth.

Austin (17:18):

Looking at over that landscape, Pyth high frequency Oracle. But again, Oracles, they’ve existed for a long time. There’s a number of name brand ones that got their start on the ecosystem in the 2017 range. Lots of people have had ideas about Oracles over the years, some of them have worked, some of them haven’t. Similar to Wormhole, bridges have existed for a long time. Bridges are actually the basis of how any L2 works, right? Both of these are hardly new ideas I would say. What about looking at the landscape gave you guys the confidence to say, not only there’s a need for something different, but we can help build something different and better.

Kanav (17:57):

Again, just like 100% organic. In that August summit, we were looking at some of the biggest things we could do. And a big problem that everyone kind of kept voicing to us is that they don’t have access to equities data. They don’t have access to fast data so that they don’t have to have things like clawback mechanisms and all these different things that LPs don’t get direct on every turn, right?

The fundamental thing with financial oracles is that they’re used to settle risk transfer. They’re used to set a price at which two parties exchange value. And if that price is latent or slow or not accurate, one side gets left folding the bag. Now, DeFi, the way protocols are constructed, the side that gets left holding the bag is either the LP that’s contributing to the protocol or the protocol stakers or a key stakeholder in building the ecosystem.

And the takers are able to take all that value. If you are going to build something that’s going to house all of OTC, if we’re building something like synthetics for example, and your protocol stakers are taking the other side of every trade that happens on S-Oil or SSNP, you need to make sure that’s the right price. Otherwise you’re just going to get up the way down to zero. When we were ideating on what the biggest ways we could contribute is let’s contribute our data. And the first idea was in let’s start, let’s go and figure out how we bring together a network of people to build an Oracle.

It was how do we contribute our data, right? And we browsed through the category of solutions. We had all the conversations. We spoke to dozens of investors and builders in the space. And there wasn’t an easy way to slot in high fidelity financial data, into existing Oracle solutions. And so we spoke with some of the founding partners of the Pyth program and came to consensus that there was an opportunity here. And that led to the first step and we just kept building sets.

Austin (19:39):

In your mind, what is it that Pyth offers that other Oracle solutions don’t offer?

Kanav (19:46):

Pyth is a very hyper specialized tool for high fidelity financial data, specifically financial data for settlement of risk transfer, right? If you think about the way the market data landscape looks today, it’s different across asset classes, but there is a class of people that have access to high fidelity, streaming price data that they can legally distribute and make available to a protocol, create like an Oracle program.

One you need access to very fast financial data, which is hard to get and even harder to have a legal right to distribute. You want to make sure that the people who are publishing the prices are the real owners of the data so that you can set incentives for the data to be accurate, right? If you are staking the value of a third party aggregator, their third party aggregator has no skin in the game. That’s one of the other kind of fundamental things that you have to think about.

And third, you need to acknowledge the fact that a price is not absolute. A price for Bitcoin has about 20 liquid trading venues that are distributed across the globe that can often be fractured, that can often have all kinds of different idiosyncrasies. And that being able to accurately determine the price on most relevant venues and build a dispersion is really important. If you think about kind of all those things together, you want very fast access. You want a broad range of access of independent sources, not reporting from the same source.

You want very high liveness and uptime of course, and you want kind of good legal clarity that that price can continue to be distributed because you don’t want the application to suddenly get turned off when the regulator says, “What’s going on?” And those are the kind of key things that Pyth has really focused on very heavily to build that piece of infrastructure and Solana was the perfect opportunity. Before Solana there wasn’t a way to create a high fidelity fast Oracle. There just wasn’t a need for it and there wasn’t a platform for it, right. And so all those things just came together.

Austin (21:49):

One of the criticisms that you’ll hear about Pyth is that because of its structured model here, where the people providing data are permissioned at this point and are also like firms that are professionalized trading operations themselves, that there is an inherent kind of conflict of interest in that system. With any system in blockchain, you have to assume everyone is trying to cheat, everyone is trying to extract the most value possible. How have you gone about setting up incentives to make sure that the users of Pyth and the contributors to Pyth are not at odds with one another?

Kanav (22:27):

Yeah. I think you made a totally fine point there in that we are building for byzantine systems, right? And so that’s the kind of incentive design you’ve got to keep in place. I’ll frankly say I think that claim is a little bit ludicrous for a few different reasons. Once you peel back the onion just a little bit, and I’ll talk through some of the reasons why.

Austin (22:43):

Let’s peel back the onion.

Kanav (22:44):

One, you’ve got to first understand that the amount of value that can be created in actually pulling something like Pyth off successfully is dramatic. And the forms that are building this are now incentive aligned to make that happen. But two, this is an open sourced protocol, it is decentralized, and you can look at exactly what the inputs are, how they’re being aggregated and what their resort in price output is.

Three most importantly, there are about 50 financial firms that are submitting independent price data to this article to construct final outputs. And these financial trading firms aren’t friendly with each other. This is the very first time that a group of highly adversarial trading firms, banks, exchanges, and ODC players across the entire space have come together and said, “Let’s go build a piece of infrastructure.” And one, I think that needs to be celebrated a lot, it’s a huge win.

But two, the trading firm, there are 50 global financial trading firms contributing their proprietary prices directly to Solana on the Pyth program today. We have realized that these 50 comprise of between 60% to 80% of global asset class volumes at this point, given the network of participants that have aggregated around this protocol. When you are that big of market share that you’re covering that kind of breadth, the participants in the protocol themselves are on the other side of each other’s trades almost by definition. And so who’s manipulating the price against who? Let’s kind of just start there.

The system of incentives that set up in this taking protocol, you can read through this on the Pyth white paper has some really intelligent aggregation algorithms that put all this data together, that identify the quality of each of these independent data publishers that then sets out a mechanism to aggressively punish providers that don’t have good prices. And good prices can mean I published a malicious bad price. It can mean I have slow prices. It can mean I published, I had a bug, it can mean anything.

The incentive design mechanism is meant to reward data providers that are not honest, but that have great data. And that’s a fundamental difference in how system designs, we’re not kind of rewarding agreement, we’re rewarding prediction. And so you are rewarded for correctly predicting the price that would come up rather than for rewarding agreement between parties, and which can both have different kind of models and can both work in different ways.

But there is almost no possibility for one collusion across these landscapes, given the composition of the people in the network. And the incentive structure again is obviously explicitly set up to discourage that. Third, all these forms are heavily, heavily regulated. I spoke about 20 years of its reputation and a giant, giant business behind kind of making a lot of this happen. And we’re definitely incentive aligned to make this thing as successful as it can possibly be.

Austin (25:39):

The Web 2.0 world and the rise of FinTech apps has largely taught people that organizations that claim to be on their side often aren’t. There’s very legitimate reasons from a market making perspective that during the game stock run up and squeeze, users of Robinhood and other FinTech applications, their trading was turned off. Now, there’s a bunch of really good backroom reasons for why that might have happened. But the effect is what matters to the retail trader, which is that they were using a platform that they thought gave them equal access to a market, that platform did not provide them equal and neutral access to a market.

I think when people look at something like Pyth, it wouldn’t be crazy to say that, well, the same incentives that made us think that Robinhood was on our side, could also be applied to Pyth. What is different about the Web 3.0 space and the construction of something like Pyth in your view that makes that not something someone should worry about.

Kanav (26:37):

Web 3.0 is fundamentally any means of resource coordination, and it facilitates that by, one, facilitating the export of trust. And the export of trust is actually one of the big reasons why the whole Robinhood debacle went on, right. They basically ran out of margin requirements in order to continue to clear trades on one side, since it was so directional.

And there is this massive web of intermediaries that set up all throughout traditional finance for the express purpose of establishing trust as the FCM, the DCM, the clearinghouse, all the other three letter acronyms. And all of them exist to make sure that when a match occurs on any platform that actually settles into a financial trade.

In crypto the match is the execution. And that’s facilitated by the fact that you can export all the trust of executing a piece of code onto Solana, onto Ethereum, onto the blockchain itself. And that’s unlocked this completely new means of resource coordination, which makes things like Pyth possible. It means that you can explicitly lay out a system of incentives in a closed loop fashion. And regardless of who’s uploading the code, or who’s proposing designs or architecting any of this, everybody is independently participating according to the incentives laid out very plainly by the program itself.


And that means DRW and Jane Street don’t have to trust Jump when they decide to publish prices to pay. That means they look at the program that’s running on Solana that they can read. They look at Solana’s trust model and decided they can or don’t trust Solana as a platform. And then contribute to the platform that then self executes and lives on its own terms. And the fact that we can allow different kinds of state to compose in a trustless fashion is the entire revolution Web 3.0, that’s basically what the whole space has been building for the last 10 years. And that’s what makes Pyth possible, it simply was not possible before.

Austin (28:32):

What does something like Jump or Jane Street or anyone who’s a data contributor to Pyth, what do they get out of it? What is their incentive apart from any rewards that might be generated from contributing data. How are they then going back and using this on chain data in their own operations?

Kanav (28:51):

There’s a few elements. And so one, it is fundamentally a two sided marketplace, right? It has data publishers and it has data consumers. And the other interesting thing like Uber did for taxi cabs, where it created a marketplace where cars could now come online, created this marketplace where data that was once latent came online.

Jump is publishing its own trades to the Pyth network. That is IP that it has the legal rights over, has only just been a cost center so far, and now has the opportunity to get monetized. And that’s the same for all of the trading firms that sit in the network. It’s a lot of people to turn cost centers into potential elements in the marketplace and that bootstraps the supply. The consumers of the data obviously are paying for this extremely created highly robust set of data inputs that then get aggregated. And that creates kind of flows in one direction. And then like your regular two sided marketplace, it accrues value, right?

All the data publishers today in Pyth have some sort of stake of asset interest in the thing succeeding. And there is a set of incentives that then rewards them for the correct participation going on with fees, rewards, all those kinds of things. And all that is in gross detail laid out in the white paper and we can go over some of that. But the off chain applications and some of this stuff is also quite interesting, right?

So if you look at kind of back office systems around the world at forms like Jump, you don’t need microsecond level access to financial data, but you need that for your trading engines because otherwise you’re playing at a disadvantage related to the field. But in order to make sure that your clearing prices have happened correctly in order to make charts in order to do something like a trading view, in order to get on the Bloomberg terminal or to be on a ticker somewhere, all these applications are now easily facilitated by subscribing to something like Pyth, that’s living on an open kind of blockchain area. And so a lot of the off-chain use cases are getting more and more interesting I think over time. The fundamental value is in creating the pricing source for on chain data. And this is kind of like an awesome thing that just falls out of it.

Austin (30:56):

That’s a really interesting way of thinking about both the incentive alignments and the rule that the data providers versus the data consumers play in the market. Are there any token plans for Pyth?

Kanav (31:07):

Yes, there is a token plan for Pyth. You can read all about it on the white paper, no comments on timing or anything of that at this point. And that’s going to be a networking governance decision, but I’m sure in the near future.

Austin (31:16):

Transitioning over to Wormhole, which is the other project that Jump is heavily involved in as a core contributor of the code. When people look at wormhole, I think it’s very easy to look at it and say, asset bridge, multi chain, cool, fundamentally utility. The first thing I noticed when we were talking about this and looking through it is this whole component of allowing different smart contracts on different blockchains to communicate with each other. I think most people understand how asset bridging works. Can you talk a little bit about this whole concept of message bridging?

Kanav (31:51):

Yeah. And this also kind of goes back to your question on, how do you decide that there’s an opportunity here when bridging is something that people have talked about for a while? When we were kind of ideating with everybody else on kind the Pyth’s team and the network on how Pyth goes across chain. Hendrick and team were building Wormhole as Solana Eths token bridge on the hackathon project at [inaudible 00:32:17].

And I called Hendrick and I asked him, “Look, is there a way to generalize this thing so that we can get Pyth messages across?” We’re building this Oracle thing on the best, fast, scalable censorship resistant message bus we can, but we want to get it to all the other ones that operate on a slightly different resolution. And through the course of that conversation, we came to a conclusion that enabling generic message bosses to allow this cross chain composability in a much more high dimensional fashion than just the token bridge word was a massive opportunity set that had to be filled.

And so when we launched last August as a completely generic message bus. And what that means is that any piece of state that is created or lives on a blockchain can be included as a message that then gets communicated to any other blockchain environment. And so if you think about Oracles, you think about a governance board, right? Uniswap passes a governance board on Ethereum, produces workloads on a lot of different chains. The outcome of that governance board has to, in a secure, reliable fashion, be communicated to all the other geographies that Uniswap lives on. That needs to be encoded as a message.

And so Wormhole has outpost contracts on every chain that is deployed, it is deployed over eight chains today. The outpost contract just listens for a message that is sent to that contract and the Wormhole network of guardians attests to that arbitrary binary block. That block can then be picked up, relayed to any other blockchain environment, verified that is coming attested from the homeowner network and then decode to do anything arbitrary and interesting. And so generic message process have really exploded over the last year. We’ve seen so many awesome applications being built on it. And I think we’re just kind of scratching the surface, right? There’s a lot to do here.

Austin (34:04):

When I think about messaging, I think about how a lot of the models right now for cross chain communication of assets are a little tedious and maybe have more risk inherent to them than are necessarily required. A very centralized example, USDC, right? You can go to FTX and you can withdraw USDC as an ERC-20, as an SPL token or across several different networks. And what’s happening there largely is because the mint authority to that is centrally controlled. They’re able to issue new, quote unquote new USDC natively on each layer that USDC is supported on. Do you see the capability of developers using something like Wormhole to make that possible for fully decentralized, both stable coins and just asset tokens?

Not only possible, but already widely adopted in the Wormhole X asset framework, right? There’s over four and a half billion of assets in the token bridge today. And the word token bridge kind of has meant a lot of different things to people at different points in time, right? The old token bridges were bidirectional, state sponsored bridges that sovereign ecosystems would run to communicate to Ethereum, to get liquidity in as soon as possible.

And then if you send that across a different bridge, then you would have like a double wrapped and triple wrapped implementation and just an absolute UX nightmare. When you use something like Wormhole’s X asset framework, you retain complete path independence as you move assets across the ecosystem. Once you’re registered as an X asset, let’s take USD as an example, there’s a couple billion dollars of USD on the bridge today. It flows throughout the ecosystem using Wormhole on the back end, Terra bridge money, uses one more on the back end to expose one of many front ends to users.

When USD flows from Terra over to Ethereum or to Solana to Polygon and then to Avalanche, it retains the same representation on Avalanche that USD flowing from Terra to Avalanche directly or through any other part in the ecosystem would retain. It’s a truly cross chain native asset. It doesn’t fracture liquidity, it fungus seamlessly, and it allows a lot of cool composition.

If you look at something, now like the result in second order effects of this, it’s this theme that we’ve been calling X Dapps, right? So cross chained apps. And we’ve seen kind of the first marquee deployment of one of these apps in the form of X anchor, which is deployed on the Avalanche chain now, right?

And X anchor is just a light set of endpoints that’s deployed on Avalanche. And all that does is it lets you kind of hit some functions that then really assets and/or messages bundled or separately or back to the Terra blockchain and then trigger state transitions on the Terra site. Anchor contracts don’t need to be deployed to every chain. You don’t need to replicate state everywhere, you don’t need to stay synchronized continuously. But you allow for outposts and communications and different chains to then communicate back to the home chain using messages and assets. And now the USD that’s in the X asset standard can be deployed to X anchors everywhere. And it’s a much faster, much more robust getting strategy that has far less communication over.

Austin (37:07):

Let’s dig into just a little bit on like a technical level too. When you’re talking about X Dapps or cross chain Dapps that are communicating via Wormhole, you’re inherently talking about fractured state across multiple L1s or L2, it’s unavoidable when you’re … anything cross chain is inherently working under a fractured state model. How fast does that time synchronization need to be for developers to actually deploy something like an AMM or a club across chain and actually maintain price parody and appropriate liquidity between them.

Kanav (37:42):

Yeah, I’m glad you brought this up. There’s a few different programming models for how cross chain Dapps works, right? One is you try to state synchronize as aggressively as possible. You keep sending messages back and forth. You have allowances, risk limits, tolerances that allow your apps to communicate. And the other is this X Dapps framework where state only lives on one chain and you allow people from other chains to then interact with it.

Now, of course that also comes with its own downsides, right? If you look at something like a club and you’re trying to trigger a cross chain swap using the club from another chain, you are inherently incurring the latency of the two blockchain transactions and the finality assumptions that you want to kind of work with that. The more stateful your application becomes, obviously the more latency and risk constraints everything through. With something like a lending protocol or like a cross chain anchor, things like that. They are less stateful than something like an order book, but order book is probably the most stateful you can get right in the spectrum of applications.

And so any cross chain swap design inherently has to have some additional liquidity back then, that’s like fundamental, right? You can ask people to take risk on your behalf. You can have the protocol take risk on your behalf, but that risk exists. There’s a lot of ways to program around it and create better user experiences, but fundamentally that’s a real problem and somebody has to be compensated with that risk.

Austin (38:56):

For the X Dapp framework, are you looking to actually be able to offload compute to the wormhole level there? Or is it really just … The natural extension of this seems to be that eventually there’s some sort of state storage on Wormhole that Dapps are able to actually access and leverage with some functionally side chain compute resourcing. Are you guys thinking about that as well?

Kanav (39:19):

Yeah. The fundamental cross chain thesis is that there are going to be independent, specialized compute environments that attack their own communities, their own audiences and their own apps. And Wormhole is away for folks to leverage state that results from these autogenous environments and compute the solutions on these environments to compose.

And you can cut that in a million different ways. You can leverage Solana as a state execution machine. You can leverage Terra as your stable coin asset layer and you can represent this third thing as a NFT thing, or you can bundle them all in. But the Wormhole vision itself right now with all the genetic message capabilities that are out there, in the near term roadmap doesn’t need to build an execution layer of its own. It can naturally extend to it. I think you’re definitely kind of pointing to something that’s relevant.

But I don’t know if that’s the lowest hanging fruit given the capacities that exist in current blockchain compute environment. The vision of course is to make people, Web 3.0 users rather than blockchain users or L1 users. You basically want to deploy resources to the most relevant execution environment with the right community, that’s creating the right apps and then expose that to at a higher order to consumers.

Austin (40:24):

Would you describe Wormhole as layer zero?

Kanav (40:28):

I’m rather old school, I think of layer zeros as networking protocols and internet backbones and things like that. I think it is maybe a useful analogy for kind of blockchain audiences given how we’ve very economically can’t use the word L1, so I don’t have an allergic reaction to it, but it’s not my first word of choice.

Austin (40:46):

What would your first word of choice be?

Kanav (40:49):

Interoperability protocol. I’m not that creative.

Austin (40:51):

Yeah. Wormhole is also supporting wrapped NFTs, which is kind of an interesting concept. I think most people don’t think of NFTs as something that’s been bridged and quite frankly, the numbers on Wormhole on bridge NFTs are quite low compared to the success as an asset bridge or a messaging bridge. What was the original idea of using wrapped NFTs? And why do you think it hasn’t caught on as much yet?

Kanav (41:20):

I think cross chain NFTs as a story are just beginning to play out. So there’s about 16, 1700 on the NFT bridge itself. And again, NFTs are also cross chain fungible and composable across environments. They are also part of the X asset framework. And so X assets can mean anything. It can be in rebasing assets like STE, it can be in NFTs. It can be in fungible assets. It can mean anything else, right?

The NFT story started to play out as a result of new other ones trying to access marketplaces that supported one or the other chain, right? And so you get to access as new audiences, you get to create experiences with different communities. You get to access different user bases, but we’re seeing the experiences get a lot richer. So you see something like [inaudible 00:42:00] come out recently, they got featured on Bloomberg for new cross chain staking program where they have in game elements that kind of change based on cross chain NFT staking that are different experiences with different communities. And much like the asset bridge has that kind of globalization and cross pollination of commercial kind of elements. Cross chain NFTs are globalization kind of culture. And incorporating a lot of those elements across games that live on Solana, that live on Terra, that live on other environments and just creating those kind of richer experiences.

And so we’re seeing people make NFTs on one chain, come to Solana, fractionalize them, trade them, put them back in, move them over to OpenSea on Ethereum. There’s all kind of interesting use case patterns. And so it’s definitely been less aggressively adopted than the explosive token bridge or the other generic message applications. But there are still 16, 7,000 NFTs, there are a lot of teams using it for cool and innovative stuff that we just kind of keep up out of the wood works every some time.

Austin (43:02):

Do you think that’s social? Do you think that’s technological? Do you think that’s just like the ecosystem hasn’t matured enough? I think I’m surprised how much … well, I guess surprises maybe the wrong term. People have a lot of emotional attachment to an NFT, in the same way they don’t have an emotional attachment to a Bitcoin. They may have emotional attachment to the concept of a Bitcoin, but I would be upset if I lost my particular Degen ape, even if I got a different one for the exact same value. Do you think that factors in at all to how people view the concept of wrapping an NFT, that it somehow weakens the authenticity?

Kanav (43:39):

I think for a lot of purists, it does. I think it was just so worthy, right. For the most part, people aren’t even going to realize, the large end of this consumers like buying these things, an NBA top shot or air, or any of these other platforms, it’s something on the app for them. And eventually it’s going to be extracted away as we draw to Eth, we draw to Solana, we draw to wallet, connect wallet, and it’s going to be kind of as simple as that. And so we’re always going to have purist stakes, but I think that’s going to remain within our little chamber here.

Austin (44:05):

For Jump Crypto in general, how do you view NFTs? There are obviously firms now that are dabbling and market making and NFTs. Is that something that you’ve looked at and if not, what was the decision not to enter that space yet?

Kanav (44:19):

It just doesn’t take a lot. We are looking at trading opportunities. You are looking about margins, you’re looking about what predictive offer you can have, like what the edge you can have on a traders and then how many times you can apply that edge, right? It’s just as simple as that. And even if you can get a 30% margin on something that trades a hundred million like week one, I mean, [inaudible 00:44:40] now.

But if you have a low volume asset class, even if it has slightly higher edge, and it is harder to predict and more dimensional, this is on a good researching decision. So as that volume changes, we will continue to stay on top of it. And I don’t know if these are trading tens of billions of dollars every day, and have really interesting datasets, I’m sure we’ll be trading them.

Austin (45:00):

If the market hundred X in size, you wouldn’t be opposed to it, it’s just the sizing opportunity issue right now.

Kanav (45:08):

[inaudible 00:45:08] you can’t be the richest man. It’s about identifying if there’s opportunity and executing all native there is.

Austin (45:14):

Looking at wormhole, one of the things I do want to touch on is the wormhole hack and exploit that happened a little while ago. It was one of the larger bridge hacks at the time. It was eclipsed a few weeks later by an even larger hack of another bridge, also targeting stolen Eth in this process. I’m sure that activities and projects that Jump has been involved in have had larger losses of money or similar volumes of money just based on the area you operate in. But this is one that inherently to the nature of Web 3.0 is very public. How is that like internally knowing that your core contributors to a project that suffered this kind of exploit, and also that failure is now a public failure, as opposed to maybe where it would’ve been a private failure before

Kanav (45:56):

Building is hard, building in the open is even harder. And building in a decentralized open space where there’s a large network of participants, consumers, affected people, the stakes we’re playing in, right? That’s the stakes that every DeFi application, that every L1 at every bridge and that everything in Web 3.0 that aims to do something meaningful inherently adopts and has to learn to deal with.

The hack was big punch in the gut, obviously a big financial loss as well. The fundamental nature of smart contracts is that the code and code can have bugs. And this exploit was kind of deep, deep, deep down in the stack, in kind of like Solana instruction verification account check that was missing. The auditors listed our team that has independently been one of the biggest bug bounty finders in the space missed, and code based at the opportunity to be out in the wide for seven months, kind of had unchecked.

The day of the hack, of course really, really rough. Jump is not used to being a public institution. So this was like you said, a very public kind of fallout in nature. I can’t possibly have been prouder of the way the team reacted to this incident. We kind identified it within short course of it happening. We pulled the meeting room together, identified the bug, fixed up a batch, managed to coordinate the guardian network to bring it up, bring it down, announce our intent to refill the gaping 320 million hole within an hour of the incident being reported on, and brought the bridge back up within 18 hours to end to end.

Building bridges and building cross chain is very, very hard. And that’s where the reward for it, building it right, is even harder. You don’t even make 320 million decisions very lightly, and this should hopefully signify you how much conviction and faith we have in the code base in bringing it back up in 18 hours. It should tell you about where we think this whole space is going and where Wormhole is going and where interoperability is going and what a core piece of infrastructure in that realm would mean.

Security continues to be extremely, extremely top of mind. We have a 10 million bug bounty. We have an internal red team that’s basically thinking about breaking Wormhole and our key projects every day. We have multiple audit from [inaudible 00:48:12] with lots of audits going on, pretty intense security review practices, all of which can be found publicly online. And I’m incredibly confident that Wormhole has come out more stronger from this incident. The team has come out kicking and that we’re building one of the best and most trusted inter op solutions out there.

Austin (48:32):

Looking across the ecosystem, let’s say over the next 12 to 18 months, what are you personally most excited for and what keeps you up at night? What do you still have worry around?

Kanav (48:44):

I’m looking forward to a whole bunch of things. So definitely very excited about all the advancements that we are seeing in the succinct proof and zero knowledge space. That stuff is just awesome, it’s magic. And I’m just so excited to see all the things that’s going to unlock for us. There’s a lot of interesting problems in the hardware acceleration space that need to be made to make that possible. There’s a lot of problems algorithmically that are kind of being uncovered there. And I think hopefully this conversation has lent on that we have a big infrastructure mindset. When I say streets and sanitation, that’s kind of what we think about every day. That’s what we’re looking forward to. And on what we can build to and contribute to that.

Austin (49:19):

You said something I got to get a little more info. You said specific hardware to accelerate certain kinds of applications. The only place we’ve really seen this so far across the entire crypto landscape is ASICs for Bitcoin mining. You see GPU mining optimization, but again, nowadays I wouldn’t necessarily even call GPU specialized hardware. It’s really commodity hardware at this point that’s just deployed for a specific application. When you’re looking at the space, where are you seeing actually custom silicon or FPGAs becoming something that it makes sense to deploy?

Kanav (49:50):

Yeah, I mean, definitely for zero knowledge provers, right? So like two verification times have compressed a lot to the point where it’s pretty feasible on most blockchain environments today. But proving itself is still super, super resource intensive. That’s where there’s a lot of simple math operations that can be encoded into Silicon and into FPGAs or ASICs to speed up the process significantly. And that’s where we are seeing a lot of adopt. There’s already a lot of people working on this on hardware acceleration using FPGAs, maybe even ASICs on zero knowledge provers.

It’s a little bit of like it’s tough to say when the right time is because there’s new changes like algorithmically coming out all the time with the new advances in new papers. And so when you spend a whole bunch of time just optimizing Fast Fourier transforms. And then the next paper makes Fast Fourier transforms not relevant. It’s tough to make a decision on when the right time is, but I know there’s a lot of work already going on into it. And it’s a space that we are very familiar with and that we are also excited about. And mostly, mostly positive stuff on the regulatory side.

Kanav (50:56):

As of recently I think there’s a lot of good faith engagement from regulators around the world on setting frameworks and policies for how kind of all this stuff gets put into place. Outside of maybe China we haven’t seen anything very aggressively or handed on cutting off innovation. We even saw India now finally starting to open up. And so I feel more optimistic about the regulatory landscape than I did 12 months ago. We need a new influx of builders to keep coming and building cool experience and leveraging this technology where we’re seeing that happen. We need capital being continued to commit to this space where we’re seeing that happen.

Austin (51:35):

The inverse of that question, what are you most concerned about on a macro level for the space still?

Kanav (51:39):

Asset pricing is of course highly dependent on macro environment and that is unrelated to crypto, right? And there’s just like, it’s its own thing. And so we’ll see price movements on a different time scale. And if you see a very sustained global macro depressed environment, then we’re going to see less capital, less builders and less momentum in the space. And I think that’s probably the biggest overhang we have today.

Austin (52:03):

In the long run we’re all dead.

Kanav (52:05):

In the wrong run we’re all dead. That’s right, so let’s keep building.

Austin (52:09):

Yes. One kind of last question here, I think if you rerun the clock maybe three or four years, the prevailing wisdom in this space was not that traditional financial institutions were going to expand their vision and embrace blockchain and we’d call it Web 3.0 at the end of the day. And you’d have Twitter profile pictures of NFTs, you’d have Jump Trading building software that’s open source for a decentralized environment. And we really have seen that that is what was originally pitched as a forked parallel path of economic development.

Austin (52:42):

It’s a little bit more twisty curvy than we thought it was going to be. And there’s a lot more integration with traditional companies. As crypto has a thesis about it, that it’s moving more consumer, right? Across the spectrum you see more normies getting into crypto in one way or another. Does the existing market of specifically the United States and Europe where you see very few competitors within an ecosystem.

Austin (53:07):

There’s basically only two phone companies. There’s basically only three cell phone companies. There’s basically only four internet provider companies. Across the spectrum you see very non-competitive markets. When you look at the consumer landscape in the United States, do you imagine that we’re going to see similar patterns rolling out there as we saw in the financial industry, or we really are going to go back to that idea of a parallel execution model?

Kanav (53:30):

Yeah. I’ll strongly state that I don’t hold a heretical view of this kind of being a completely forked off parallel path that has no relevance to anything that we do today. I think it’s an amazing technological invasion that gives us tools to coordinate resources in an untrusted environment. And that’s unlocking a lot of magic.

Kanav (53:49):

But that again bleeds in with the rest of the real world, which is also big and has its own dramatic pieces of innovation and with a whole bunch of other stuff going on. I think one of the most exciting things has been kind of the global equalizer that crypto can serve to be. Yesterday we saw Polygon come out with an integration with Stripe. And these are three kids from India that had no early supporting or backing that kind of boosted the network on their own and are now competing on a very, very competitive landscape with people from every single part of the world that are very well resourced, competent teams.

Kanav (54:23):

We see [Inaudible] coming from Korea. We see teams from Australia and New Zealand over the [inaudible 00:54:28] guys. We see people from Berlin and the US and everybody competing on the same, not only the similar consumer markets, but also on the same capital markets. And there are network effects that accrue, but not cannibalistic network effects that accrue. That makes me very excited about where the space is going overall. When we talk about integration points itself, it’s going to largely depend on [inaudible 00:54:52], right? And that’s like an unsatisfactory answer.

Kanav (54:55):

But if you’re talking about financial markets, crypto is already integrated heavily into the financial markets with 15 excellent international venues that are competing, so we already have a fractured environment. That is before the [inaudible 00:55:08], the NASDAQ, the CME groups have made their moves in the space. And they’re clearly not going to be monopolies in crypto, obviously, right?

Kanav (55:16):

If you look at something like a telco and interactions with like cell networks still remains to be seen, whether like decentralized constructions of those kinds of things can be competitive. I mean, building telcos and stuff has such strong network effects and so many economies of scale. And it’s unclear whether a Web 3.0 means of accruing that value to a decentralized organization has the ability to accrue the similar kind of network effects and so remains to be seen. But I’m excited to see it play out.

Austin (55:43):

I always enjoy getting to pick your brain about where these technologies are going and the intersection of a very traditional financial world with this new global system that we’ve all been building. But thank you so much for joining us for spending some time digging into this stuff.

Kanav (56:00):

Thanks a lot for having me on Austin. This was super fun and as always, love chatting, so yeah, we’ll see you again soon.

Austin (56:04):

Thanks.

Leave a Reply

Your email address will not be published. Required fields are marked *