Avantis: Leverage Trading for Global Markets on Base

0:00 Welcome back to another episode of the Permissionless podcast brought to you by the team at Guardian. And of course, as always, we're going deep with DeFi founders, getting to see the perspective of the DeFi markets, what their thesis is with what they're building, what got them into DeFi, how they found their way in, and of course, what is keeping them up at night, what sort of problems do they run into, and how are they addressing those?

0:26 And today we had a really high throughput conversation with Sehaj who is a co-founder of Avantis. And Avantis is really focused on building per markets for off-chain assets. Today, of course, they've got commodities, they have FX markets, and in the future perhaps even stocks. And we got to dive in really deep and get to the nitty-gritty of what the competition is in the per space, in the derivative space, how they're building, where they see themselves fitting into this niche.

0:52 And of course, we got to examine the order book versus onchain settlement oracle architecture. And we got to examine the pros, the cons, where Avantis sees themsself on that spectrum and and in a certain niche there. And then finally, we touched a little bit on the regulatory environment. And we got to hear about where Sehaj thinks that D5 is going to go and and where he's positioning Avantis.

1:16 And last but certainly not least, I think some really good tidbits of of wisdom that were dropped when it comes to building your DeFi protocol and picking the right market, picking the right place to build the right thing to build and being extremely aware of the game that you're playing when you build in DeFi. So, of course, we'll get into all that and much much more in the conversation in just a second.

1:38 But of course, my name is Owen and over three years ago, three and a half years ago or so, at this point, I co-founded Guardian. And really, we set out based on the premise that DeFi would house trillions and trillions of dollars of TVL and be a core player in the financial ecosystem of the world. We set out to build the security audit solution that would be necessary to build and ship consistently and confidently in that environment with billions and trillions of dollars at risk.

2:06 And in correspondence with that vision and mission, we've upped the Annie. We use two competitive teams of smart contract auditors, a worldrenowned lead researcher, and a dedicated smart contract fuzzing engineer to perform advanced invariant analysis and testing throughout the engagement. And really, in our experience, we found that the typical audit process is not sufficient. Projects are getting audits and they're still getting hacked.

2:29 And the Guardian model is here to put an end to this. And today we're confident that we've achieved the standard that is necessary to ship when there are billions of dollars of assets at risk. And of course already today with over $10 billion sitting in smart contracts directly secured by Guardian, we can confidently say that now it's just time for DeFi to catch up.

2:49 And so now without further ado, let's dive in to this conversation with Sehaj from Avantis. Yeah, dude. I mean, really, first of all, just happy to have you here and, you know, really, I think we could just dive in. Like you mentioned, you have, you know, a little bit of a history coming into DeFi and and you've done a lot of things that were sort of adjacent to onchain stuff in general.

3:14 Yeah, dude, I'd love to hear more about like what was your journey and what did that path look like for you? Yeah, so I, you know, I didn't buy Bitcoin in like 2013, nothing like that. I got into crypto after graduating Berkeley in 2021. So, very late when everything was moving up only. Funny story is I could not buy stocks when I was doing investment banking and so I dipped my feet into crypto and I bought a bunch of Ethereum in the worst possible place you could buy Ethereum which is Robin Hood with their spreads and with their lack of custody.

3:40 So bought it there it went all the way down and then I figured like I should actually know about the technology that I'm buying into and not just buy hype. But when I looked it up a bunch of buzzwords coming from like a nontechnical background like what is what does it mean to be permissionless? What does it mean to be centralized?

3:54 None of it made sense to me in theory. And then I just used an app. I used an app called Aave and I said, "Oh my god, like I can lend my money to some fellow DJN who can use it without my permission." That's what it means to be permissionless. Interesting. And so I got into D5/crypto because of D5 DAPs.

4:08 And so when I got into investing as a VC at Pantera, I like always had it in my mind that I wanted to build a DAP as a founder and kind of give back because I could see the whole like L1, L2, L3 proliferation. I I could tell that like this is unsustainable from a user adoption perspective.

4:24 Of course, it's very sustainable from a funding perspective because we all know where the incentives lie. But I wanted to kind of build something that people use and say wow. Okay. Like I just like I said wow to a people want to say wow to my app and that was my inspiration to build Avantis. So that's the high level of how I kind of got into defi and then now got into being a founder in defi.

4:42 Yeah. And I think that's especially relevant because, you know, way back in 2021, 2022, the problem was block space and then that was getting like increasingly solved until we're to the point where it's like, okay, really not a problem anymore with with L2s. And now it's about how do we fill up the block space with like useful, valuable transactions that actually bring value to the end user.

5:01 And I mean, Avantis is that is at the core. Well, I mean really every DAP is trying to do that, but Avantis is doing so in a way that you guys are bringing venues onchain for traders to trade offchain markets mainly. How did you see that as like the niche that you were going to go after with Avantis?

5:18 Yeah. Yeah. I think there already exists venues to trade offchain assets offchain. So the question you're asking is why build that onchain? And like to us it still is a big differentiation that you can basically create composibility. So like we have market makers powering these assets as well. So our USDC vaults generate like some level of yield.

5:36 In the bare market it's like 5 to 10%, in the bull market was like 50 60%. But this vault itself is composable. So composable RWA marketmaking vault just cannot exist offchain. That's the number one thing. The number two thing is that we just don't think onchain capital has been introduced to offchain assets in a meaningful way. All they've done is trade memes and this problem has been progressively getting worse now with you know the hit rate on memes call it is like 5% for the average.

6:01 So it's almost like you have built a casino but that's not necessarily a good thing. There's a lot of kind of correlation of Bitcoin to macro anyway. So why not trade macro assets in the same terminal if you're trading or where you're trading Bitcoin. So people on Avantis will routinely go short Bitcoin go long gold, right? And both are basically linked to USD at the end of the day.

6:19 It's like XUSD and then it's BTCUSD. The denominator is the same. You're betting on the rise or fall of the dollar. So why not introduce more assets? That was kind of the thinking. Yeah. And so I mean definitely there with the composability of DeFi I think that's where DeFi really has an edge and the exciting thing about that is it's compounding every year that more apps are built on top of chains.

6:38 We get more DeFi primitives. You know first there was an a now we have you know Uniswap and we have Uniswap v2 v3 v4 keeps getting better and better. I mean that's something that traditional financial markets will just not be able to compete with. Take it at 5 years out 10 years out. that becomes like the huge advantage and but there is some advantages that still lie in traditional markets like you know centralized order books with with the the latency and the like no front running and stuff like this.

7:05 Absolutely. Yeah. How how are you guys thinking about that because especially competing with these centralized solutions is that something that comes up a lot? Are you guys thinking about this or like how does that sort of come across in in your road map? Yeah, I think as first-time builders, we did not think about it to be honest with you.

7:19 And all we thought about is like people care about decentralization and all that stuff. Nobody really cares about that as much as the benefits of decentralization. Whether that's composability, whether that's better incentives via token, whether that's more assets offered, they care and obviously they care about security. Whether you're offchain or onchain, they care about security. So to answer your question, UX matters a lot and yes it matters and we think about that all the time.

7:41 Luckily for us, Base is also thinking about that. So as you know bases is centralized obviously in some ways even though they you know settle proofs on Ethereum but they are taking that one step further with flash blocks which you guys might have heard of which is hopefully going to improve block times and I'm curious to get your take here as well but from our perspective as a per deck going from like 3se secondond block times to like you know sub 2cond block times to sub 200 millconds 10x improvement in theory sounds amazing and anything beyond that like you can improve to 50 millisecond or like 10 millconds but it's okay like to the human eye like it's a second and and that fine you know but right now on Avantis given we have fully

8:19 Onchain execution there's no order books it's like you are putting the intent of your execution on chain and an onchain keeper bot is also executing the transaction it's a 4 to 8 second delay in finality which is which is a lot that is a lot right so we do think about it all the time and we are very excited about flash blocks I'll just say that yeah I think there there is like the base chain layer aspect which is the block times and the finality or I think like in Base's case like the soft finality at least and then there is like I think in my opinion the harder part to attack from like a fundamental design/archchitectural perspective for these like perpetual derivatives types of products is the pricing and with this Oracle model yeah

8:58 You have that two-step model where it's like you commit an order you wait you know whatever the Pyth some protocols have it like a 60-second time frame some of them like five minutes and you get whatever price the oracle puts you at but it's executed you know maybe like a minute later or something like And fundamentally that like forces in from a design perspective like from the architecture you're already kind of behind what an order book could be able to give.

9:22 But there are some interesting products that we're just starting to see come to market like Pyth Lazer. I don't know if you guys have played around with that a lot. That seems really exciting. I haven't sort of dove into exactly how that comes to fruition, but starting to see it getting used and and a lot of teams are moving to an order book model.

9:38 Like you know a lot of teams are exploring them and I think especially Hyperliquid was the one that really set off trend I suppose and now everybody's everybody's having a look like it it seemed to almost be an order of magnitude greater for you and it that really is what spun out their growth and like the the parabolic returns that they got from that.

9:58 Have you guys thought at all about like the order book model? How do you guys contrast that? How do you see that playing out with respect to the Oracle model? Yeah, such a great question. So autobooks very hard to bootstrap. You need to have extremely good BD with market makers. Extremely good take flow from the start which requires some level of nuance around like who your backers are your go to market is how you design the exchange.

10:19 That's autobooks for you but then very good to scale up right very economical to scale up very capital efficient to scale up on the other hand like very easy to bootstrap. You know, you can have a million dollars in the liquidity pool and and taker can take 100k trade with very efficient slippage because you can design that synthetically, but then very hard to scale up because you just run into capital constraints.

10:38 It's very hard to net out OI, manage risk, all those things, right? And find the capital that is sticky. So, we are at a phase where we have about 10 million TVL. We've done about 4.2 in volumes across 30,000 traders. So, we're growing. But the one thing we realized is we still want to stick to AMMs because of two reasons.

10:53 One is that we found ways to make the AMM more efficient like 10x more efficient. We can do 10 mil OI on like 1 mil of TBL level of right there's there's technical ways around it. The number two is with like underlying improvements on flash blocks order books versus AMFMs doesn't matter. You just you know like placing a trade and it will be executed within a second or two at most.

11:11 And so for us to cater to like a whale who does like a million dollar Bitcoin trade, it's actually better to be on an AMM, right? not do all of that BD execute against the AMM in like a flash and just make sure risk management systems are in place to charge appropriately, right? The third thing is we cannot list RWAs as effectively and quickly as we have if we did not have an EMM and again price things correctly and and all that stuff.

11:33 And the fourth and final thing is from an innovation and composability perspective, AMMs are still better in my view. I know that Hyperliquid has made a vault and the vault itself is can be made, you know, like a ERC 4626 and just give the yield to the vault and it's composable. But through onchain walls with composibility is still on the emm side.

11:49 And sorry I know I said I've said so many things now but the last thing is we've launched a new product called zero fee perps where you open like a large trade. You pay zero opening fees close zero slippage. You pay zero borrow fees. So no holding costs right no net cost. You only pay from your profits.

12:05 So this is a new new order type that we've created that's really taken off. It's in private beta with like thousand wallets and we're seeing like about 20 30 40 mil volumes a day. Monetization is obviously not great because people mostly lose on per, but people love the fact that they can swing a trade and just like try their luck, like try the hand.

12:20 Like if they lose, they contribute to buffer. If they win, we take a fee. So there's inherent house edge for us and we're aligned with the trader. We're telling them like if you win, we win. If you don't win, we don't win. Like it's that's as simple as that. So can I do this on order book? Probably not, right?

12:33 Are we open to order books? Yes. But I'm seeing the kind of instruments that we want to build are like easier to do on AMM somehow. Yeah, I like the analysis there. I think it's also important to know what niche you're going after because an order book serves very well like the high frequency, the people who really need that low latency, but there's also a whole section of traders which don't really care about that.

12:55 And like let's say I'm just I just want to take a position on gold. I don't really care if it's from an order book, from a price feed. If it takes, you know, a couple seconds for it to settle, that's fine with me if it's a longer term position. So, I think that's something else to to keep in mind is like building for the exact niche that you want to carve out in the marketplace.

13:14 Yes. And you don't need an order book for a lot of that. And like you said, there's a lot of overhead with like professional market makers and stuff like that. So, that totally makes sense to me. And then also in terms of like this new fee model, I think this was this is one thing that seemed like the biggest differentiator to me.

13:29 I mean, of course, besides the fact of of what you guys are are listing and you're focusing on off-chain assets, but I think this is starting to become a big part of the game is like who can reduce their fees to zero or who can incentivize trading volume the right way, etc. Because at the end of the day, it becomes like a volume game, which is like how much volume do you have impacts how much you can incentivize LPS, how much you can pay out to them, which it becomes like a virtuous cycle.

13:54 So, I think this is like an interesting offer to traders, and it definitely does. It has like a almost like a like a risk-free type offer where it's like, "Yeah, if you don't if you lose money, you you don't pay, and if you do, then we just take a portion of of what you want or something like this."

14:10 Exactly. Does that apply to the fees that go to like let's say the LPS? Like I'm assuming you guys have some kind of like a borrowing fee that that applies to every fee, borrowing fees and such. So right now the core protocol like fixed fee per as we call it or regular per as we call it opening fee closing fee borrow fee right and we just take like 40% of the opening and closing fee we don't even take borrow fee borrow fee goes completely to the LP cuz we think that again they are they are the ones lending the capital so they should get all all of it liquidation revenues we share liquidations obviously add to the buffer as well our buffer has been net zero basically cuz like traders have won lost etc but

14:46 Over the last like 6 months we've had like net zero buffer and we've had zero instances of like share price decrease increase in the last 6 months. In the first 6 months as we were like growing like the growing pains of like understanding risk, we had like very huge share price gains and then like a lot of like but a high sharpie ratio, right?

14:59 So a lot of volatility on the way. So if you look at our analytics, you can see like year to date like share price growth is like this and then it goes goes in a flat line above, right? Because it's like more steady and slow with no P&L distributions only fees. So yeah, high level that's the economic split and then for the new order type it's interesting.

15:16 So we split win fees 60/40 as well. So when a trader wins, win fee goes to them. And the same thing with liquidations, we share liquidation revenue 50/50. That's it. I see. Yeah. That's almost similar to like the referral model where it's like you take a portion of the fees and give it back to like the refer sending it back to the user in this case when they lose.

15:33 Basically, that makes sense. I'd love to get you guys on the platform and and and get your feedback live on like because you guys have used GMX synthetics probably Hyperliquid. It'll be good to see what you guys feel when you like actually use this new like order type as like DJs, fellow DJs. Yeah. How how high can I max my leverage at?

15:49 How deep in can I get? 250x. Enjoy. Oh, okay. Yeah, that'll do. Another differentiator, right? Like a market maker does not want to take that risk of like kind of counterbalancing such a high leverage, but we we can like our LPs and us and market makers collectively can take that risk. So nice. There's another that's another kind of differentiator.

16:07 But yeah, I was just going to say it's it's a really cool and interesting idea. So, like with these new type of offerings that you're coming out with, is this how you're attracting new traders to the platform or like what has your strategy been to like, you know, going from zero to one as a founder, first- time founder, like bringing traders to Avantis?

16:22 Yeah, I think that's been a big one. I think building on base and then we were the first DEX to be backed by Base and like so I left Pantera and I started this because I I definitely was very bullish on Base when they had like 20 million TVL and they had just launched. Aerodrome had just launched.

16:34 I just I just like had a thesis that it's it's I've had enough personally of L2s and and L3s and L1's like I think there be one or two that take all the lion share which in today's world seem like it's going to be Hyperliquid Solana base potentially Arbitrum also very close definitely very close and yeah and in the future like you know as Coinbase grows more with deregulation they bring that traffic to base and so my thinking was like I'll build a good product I'll build innovative taker products this includes zero fee per that this includes RWA and then Base will bring the distribution of course we will as well with marketing, but base will be the kind of core engine.

17:08 That was the thesis and that thesis has mostly been the same. But of course, there's a lot of changes and a lot of nuances. And we've realized that, you know, people don't necessarily care about like where the product lies. They care more about what it is and can they get there within a few clicks. people can't still get to DeFi and especially base L2s from fiat in a few clicks.

17:26 It's it's like you have to be onchain and that market is very limiting. It's a hundred billion dollar market wise and so that's an unfortunate thing. But I think as again as regulation improves as we can get regulated as an exchange hopefully that would be the the end goal right build on chain but be regulated and like serve offchain customers that would be the endry then we can do better.

17:44 So hopefully that answers your question like first- time founder how am I thinking about it it's changing and evolving over time. Yeah I love that approach it's in that framing I I think most DeFi protocol founders don't really think about it first. you chose the market first which that being like the the base ecosystem and then you know everything is is downstream of that if you choose the right market you choose the right ecosystem everything gets much much easier like launching the same thing on like a I don't know like a nosis chain or like some that's going to be so much harder definitely definitely yeah yeah so working backwards for the market and then crafting like okay now what offer do I want to craft for this market and you know you've already made

18:21 Everything so much easier easier for yourself starting out from base at like 20 million in TVL. I don't even know how many how many multiples of that we are now, but it's a good wave to ride. So, I think there's such a a sound way to approach building DeFi that I don't think a lot of people actually think about, but you know, seems to be just like something that implicitly you kind of already thought about from the start, which is cool.

18:40 But, yeah. So I mean building on base have you you talk about like integrations and composability. Do you see any particular advantages that you get on base? Any particular protocols that you already support with integration or like something you you've got on the the road map? Yeah road map. Yes. well I'll talk about existing integrations first then I'll talk about the road map.

18:59 Existing integration superform. So we're like basically integrated on superform. They're like a yield optimizer. They like listed Avantis very early on and then half of our TVL actually came from superform. That was one thing. And then we give them points. We get points from them as well. That's one easy example. Those are like the kind of partnerships I love because it's low effort, high ROI.

19:17 Like how hard is it to list a 4626 vault when you're already dealing with all those vaults? But then we have some front ends that support us on the per aggregation side of things that are coming out like Symphfony, Cadence, previously known as Cadence and then we have like a partnership with Morpho. I wouldn't call it a part like an integration with Morpho where in our in Avantis you can just borrow capital using your ETH USDC.

19:38 Plug that USC into the vault, double points, right? Or you can just like plug that USC into trading, be a DGEN. So, not too much is borrowed from within the integration. It's only like about 150K of USDC that's borrowed from within the Avantis tab. But it's just again, it makes it easier, right? So, when people are running out of USC, we say, "Hey, borrow USDC using, you know, Morpho and just takes you to the borrow tab."

19:58 Simple things like that, you know, mostly. Of course, Pyth has been very supportive in our journey because they keep marketing the fact that we are using their feeds to generate a lot of volume. on the RWA side they obviously will will care about that and then in the future we'll also partner with other oracle providers which might be a bit more offchain but just faster or look at options like laser in terms of future integrations we're looking at basically more like on the vault side so like basically yield trading basis trading with with spectra or pendle where right now the math vaults are there's two vaults and within each vault you can also lock your capital for a certain amount of time get an NFT and get boosted rewards boosted XP and boosted yields That's

20:36 How it works today. But in the future, we'll combine these vaults into one so that we can also go like the Jupiter drift Hyperliquid route of like kind of having one mega wallet powering all trades and then that mega wallet can be it can have its own basis trading like principle and yield tokens with spectra or pendle and then also that vault token can be used in say like a moon well on the lending side.

20:53 So I want to do more stuff on the on the supply side now given we have enough like demand side integration that's high level. Yeah. Yeah. Absolutely. It's it's like there's no one integration that will like completely change a protocol, but if you have a thousand of them, then it just stacks up and it becomes like, oh, this is the hub for liquidity provision or perpetrating and stuff like this because you have all these all of the features that you could want, all the integrations.

21:19 I think that's when an ecosystem really starts to thrive is when they get all of the builders on there building in unison and everything that they're doing, all the liquidity they're attracting, all the yield they're generating compounding and sort of just boosting the entire ecosystem makes a great offer for somebody to bridge and you know put their capital into the ecosystem and I think Base has been doing a great job of that.

21:40 Also the guys on like Sonic, they have been doing an amazing job on that. I think they've historically been very good at that. Totally. But yeah, stuff to learn from for L2s and and people bootstrapping ecosystem. And yeah, it's it's definitely a part of the consideration when you're building on on a certain chain. Speaking of that, what has been like the most popular feature or like what's the most popular like feature that brings the users to Avantis?

22:06 Do you think it's going to be this fee switch for free trades? What what do you think it's going to be? Yeah, right now we we're pretty excited about zero fee perps. We've been like planning this for the last 6 months from idea to audit till like the MVP is what 6 months we've been working on this and right now we're at a stage where we're like figuring out the economics of it and like how will our market makers react to this order type where people can just swing along and then will they take a short like will the LPs like take heavy risk here like we're trying to optimize those things right now so it's a very risk focused discussion but yeah consumer wise we've seen some good traction here I definitely feel like

22:41 The way we've built it out psychologically as well like promotes it a fair bit. Like I would use it as a founder myself which is like I think a good sign. Yeah, I think so. We expect zero fee ps to be the driver and then I think people who trade rwas will come here anyway because again in terms of liquidity and stuff this is what they know on chain.

22:58 That number is pretty small. We have to grow that a fair bit. I think those two things should basically be the drivers and then we we're already thinking of new derivatives, new assets to list basically just expanding beyond crypto and RWA like what can we do with leverage? That's kind of the theme. So I won't give away too much but that's the theme of it.

23:15 Yeah. That's the beauty of like synthetic trading venues is really all you need is a price feed. Exactly. Structure and then now you can list everything. It becomes very very scalable in that sense. And then also regarding the fees I was interested is there any inclusion of like price impact? Is there any difference in price impact with this fee update that you guys are pushing out or is it just applying to like open close fees?

23:35 Yeah. Yeah. So the way it'll work is that right now we offer zero slippage on Bitcoin in the fixed fee model and the reason we can do that is number one trade sizes are not that crazy high and number two there are fixed fees associated with opening those and borrow fees as well in new order type where there's no opening fees, no fixed closing fees, no borrow fees.

23:53 There will be some slippage but it'll still be very optimized. It won't be like charging you an arm and a leg because think about it from the taker's perspective. They want to take a 250x degenerate long trade and you charge them like 20 like 10 bit slippage. They're down they're down like 25%. Right there on their on their collateral.

24:09 So, of course, you are giving them the privilege of 250x, but you have to be still very slippage optimized while managing risk. And so, the way this new order type will work is that we'll have some optimized slippage. So, like maybe three, four, five bips for like a half a million dollar trade, which is very fair if you think about it.

24:26 It's very fair, right? But then when you have the win fees, the lower the ROI, the higher the profit sharing. The higher the ROI, the lower the profit sharing. Why? Because again, if you're scalping in and out and you found some like kind of either an oracle delay or you found some like profitable trading strategy, we'll charge you like high amounts of your profits as a fee share.

24:45 But as if you hold the trade organically, that gives our market makers time to hedge. That means we'll charge you a lower profit share on a higher ROI, right? So that's kind of the the thinking around how this new order type will work. And again, I'm keen to get you guys' feedback to just talking about it. Yeah.

25:00 And I I'm pretty sure of course you guys also have like probably like open interest caps on individual markets. So it's like you can limit your your risk there. It's not like anybody can you know go and manipulate the the price on let's say Binance and then just open like a billion dollars of volume on on your on your guys' Exactly.

25:19 Exactly. Like it's like very small right now. Basically in the grand scheme of things it's like $4 million $5 million on Bitcoin like two three million on Solana. So it's not really that crazy and we've not done the AMM optimization right now for the sole purpose that like there's a lot more why we can utilize even at just 11 12 million KVL.

25:37 Yeah, it's a really interesting like almost inverse network effect that like as you scale and you get bigger markets, you by definition have to have worse price impact to avoid getting arbitrageed with these price feeds is something. Absolutely. And of course your price impact can scale and you have you know the right algorithms so that they're paying the appropriate amount when they get up to that larger but yeah and I know you guys are working closely with like chaos labs and they do a lot of the work on on that analysis.

26:03 What kind of involvement does does Chaos Labs have in like the configuration and and what kind of help do they give you there? Yeah, so we worked with them like in Feb of 2024 or like Jan. This is back we were launching mainet. And the reason we worked with them is I'm not sure if you've noticed, but Avantis does not have funding rates, but we do have something called loss rebates.

26:23 And so it's like easier to explain over just by sharing my screen, which hopefully I can do. The way it works on Avantis is that we have something called a loss rebate. And if you're balancing a market and if I'm bringing skew from like 48 up to 50/50, I'll get a 5% guaranteed rebate on my losses for like basically taking this trade, right?

26:41 So, as you can see, my collateral is 10 bucks and I'm getting a maximum rebate of 50 cents on that by taking the short side. It's pretty cool. This is basically like an option on your per. So, and which doesn't expire. So if you if I open this trade, this trade will you know if I end up in a loss, I can like basically close it and pocket this and then be head on the other side.

27:00 Because of this, you know, market makers like Krock have basically like gotten a fair bit of loss protection over the last like year roughly. You know, kind of since this is so new, our whole idea was like let's build something that can be an alternative to funding rates and give both directional traders as well as market makers a reason to hedge, innovative way to hedge basically for them.

27:20 And of course there's pros and cons of innovation. The pro is that you come up with something new. It's 0ero to1. It's exciting. The con is that you have to explain to market makers how it works. They have to they have to model it in the back end for for them. So we we've had both good and bad views on loss protection.

27:35 But net net it is something new and it's something that you know like market makers love. Takers also love when they get a chance to actually use it because it's always arbitrageed away to be honest. But with this zero fee order type the one good thing is that there will be a lot of opportunities to arbitrage. So basically given it zero fees, you can see that like if I swing a very large trade size, say like a million dollar trade size, I'm saving on like five six hundred bucks in fees and then six bucks in like borrow fees, slippage is still like fairly decent one way.

28:02 And so this will create a skew and then that skew can be arbitrageed away by a market maker. And our goal is obviously like how do we ensure that the MM can take that size and not be like basically affected by a taker who's just opening and closing because there's no fees to do so. So we've had if you can see like basically Ethereum we're having we're seeing a lot of lot more like trades on the zero fee order type and you can see like it give it's very tempting to do 250x and just get blown away right so that's the high level like basically either you choose like a regular order type and then you pay the fixed fees you pay the borrow fees but then if you truly believe you're a good trader

28:35 You end up keeping much more of your profits in the long run you can see in our P&L calculator as well like if you're like a zero fee order type or like a like a regular order type trader you're saving money on fixed fees is right. But then if you compare to like the fixed fee model, you're keeping less of your P&L the more price actually goes in your favor.

28:53 Right? So you can see 2125 for zero fee and then 2500 for like fixed fee. And yet we're seeing so much volume on zero fee per because again at the core of it per traders are loss averse just trying to create new ways for them to like use the platform and trade while economically balancing all of it on the back end.

29:08 Yeah, that's very cool. So with the with the the loss rebates, those are essentially it's kind of like stashed positive impact that you realize when you do get a loss in the future. Very cool. Very cool. Yeah, that is definitely interesting to think about and I can see the sort of like gamification where where the traders really like it and then yeah, I could definitely see where market makers are like hm we've never modeled something like this before.

29:32 Everything we do is like always like net new and then we're like left figuring out the aftermath later. But we always think like okay let's let's change it up. let's see why something else can work, you know, like it's just cool to like try something new. But yeah, of course. Yeah. And I mean, if anything, it's better for market makers than like a typical positive impact would be because you see like a lot of exchanges just give like positive impact for balancing and you get that right away.

29:55 But for the market maker, it's like, okay, now you have to make a loss to even realize this. So yeah, in my mind, it's even better. Yeah, it's like I mean, and they won't always make a loss, right? they might have to book a profit because somebody because the skew has changed and it's in the SA to kind of follow that.

30:10 But when you do make a loss, the loss the dollar value multiplied by the percentage can scale with the loss, right? So the loss if there's a big market crash and the market maker was loss protected like congratulations, right? You gamma scalped a very good amount back. Yeah. And do you see a lot of the volume coming from arbitrage or coming from people taking like a delta neutral strategy to take advantage of these?

30:30 So before it was like 30%, because only a couple of assets are hedged on the platform like majors and memes are hedged. L1's, L2s, altcoins like they're not hedged right now. So it was like 30% hedging volume. Now it's down to like 5% because of zero fee perps and how like MM are getting scared. And so we are convincing them like hey guys it's okay, chill.

30:49 Like don't balance all the trades. Balance the longer termish ones like the five minute 6 minute plus trades. Forget the scalpers. We'll just take house edge on them. Yeah. Or hopefully house edge. Right. We'll see how that pans out. U so far it's been net neutral to the house. Yeah, that's the that's the high level. Yeah. And of course I think another really interesting aspect about you guys are specifically t tackling offchain markets is the market hours.

31:13 I think that a lot of the per venues don't really design with this in mind and then the listing the offchain assets is kind of like an afterthought and then they have some issues when they go and they try to adapt their architecture for like an off-chain hours thing. they start to add some of those markets. Was there any particular like how did you guys approach that from day zero like what are there any measures that you guys take for these market hours assets?

31:35 Yeah, I mean we can always like basically the way it works on amounts or most onchain purposes that keeper bots have to be whitelisted or they're open source but they have to have some kind of rules on what they can and cannot execute. So as soon as market hours like are or like offchain markets close our keeper boards are also like they know that it's back end board that knows that I cannot execute this trade even if a person places a request on the smart contract directly via the SDK and on the front end obviously it's very easy just you just show that the the markets are closed and come back again when they're open the thing you have to communicate well with the user is that that look guys I mean like you can

32:12 There will be gaps in the market so like when the markets close Friday evening and they open up Monday morning there might be like a 20 bit gap and you're on 10x leverage. So you might get stopped out of your position. We cannot control that. Even even centralized venues cannot control that. The only way to kind of fix that if you wanted to is make ours 24/7.

32:29 And for that again the feed has to be 24/7 and trusted which which pit will take some time to do but from our perspective we can always use other feeds that are 24/7. Yeah. I think the biggest challenge is like okay so the close of the market price and then the the open price based on like pre-market trading can be totally different.

32:45 And a big worry is like LPs can just like withdraw on on the news. but I think you guys have like specifically like lock up periods to address that. Is that right? So we don't have necessarily lock up periods, but like LPS that choose to lock will obviously not pay an early unlock fee and and and they they are locking because they're getting enhanced fee rewards that are claimable and not being auto compounded.

33:06 So they can just use that as as dividend/cash flow and they're getting like boosted Avantis XP which will eventually, you know, be a token. So from their perspective they won't withdraw which helps us get some stability in the vaults to be honest which was the original way that was the original intention of designing it this way. And the second thing is when the vault buffer drops below one like right now it's at.998 sometimes it's 1.02 if it drops below one there's like a 25 withdrawal fee and so you will like twice like do I want to pay this fee do I just stay in the vault and if I pay this fee then I will contribute to the buffer.

33:36 So there's like around it. Yeah. Yeah. Yeah. And I'm assuming you guys even have like probably a hard stop where it's like we have a reserve ratio where you just can't withdraw past a certain We do. We do. Yeah. If like if the OI is a certain level like you know the the vault cannot be withdrawn below that level and that's conservative because OI is like just a function of leverage but the payouts might not be the OI.

33:57 Payoffs would be like the collateral plus P&L. Yeah, that totally makes sense. The fee the increasing fee on withdraw as you get lower is pretty interesting. I don't think I've seen that. That's another thing that I haven't really seen too much with with other models, especially with like synthetics and stuff like this where they have like a staker model.

34:13 So, yeah, that that's pretty interesting as well. Cool. So, I think I mean we've gotten a pretty good idea of like the highlevel differentiation of like what you guys are are building for. What is like the long-term goal? Like what's the end product? Of course, we never get to like an end product, but like what's the end state, end experience for Avantis from LP side and also trader side?

34:34 On the trader side, I I'll probably address that first. I think a mobile app with like good distribution and you open the mobile app, it's everything is there like all assets are there. Forex is there, metals, commodities are there, crypto is there, different modes of trading them are there. Zero fee per regular per trader see and that's all that anyone using the app sees.

34:53 The demand is the app because the app is a distribution and the distribution is because we did something right on the regulatory front etc. Right? And the second piece is the LP piece. That's more professionalized. That's more like, oh, you're a yield hunter, you're onchain, you know, that capital, we'll chase for that capital, right? And then that can be done on the website.

35:10 You can like look at your options. You can look at kind of on the website what composability venues you have to further your yields and make it kind of an ecosystem play. So that's kind of the goal there. And if you like that's the kind of medium-term goal there. And then if you think about long-term like the the vision would be to kind of expand beyond like RWS and crypto into like prediction markets and sports betting that would be the house is still there just you know different instrument but like same liquidity layer for all these assets.

35:36 So like kind of global assets were the home on base using that distribution and making things more efficient and the TDL is like much higher hopefully right because of the inflow of demand and we're much faster because base is faster. So the kind of aligned vision will take a lot of work obviously to get there and everybody wants to get there.

35:52 So we'll see. 100%ed. I mean we do too. We want to see more capital coming on. I mean it just empowers more apps to be made and more integrations all this stuff and and that's a part of the composability and the flywheel that DeFi brings. So you mentioned regulation there is like a pretty big one. What do you think is like the number one blocker to getting to that state right now?

36:11 Is it regulation? Is there some other kind of bottleneck? What is it for you guys? Yeah. So honestly the way it works at least what I understand from my knowledge is that you need a lot of money to apply for licenses in the US. That's the first thing a lot of money this and most founders don't have it neither do like you know I'm not we're not bootstrapped we're backed by VCs but it's not a good use of VC money to million $2 million on licenses which you're not sure of if you'll get or not.

36:38 So you tackle the onchain space and you like basically get revenues that way. bootstrap and build a product and we be a team and then you can go with that kind of treasury and compliance runway to build this out. That's the first thing and then the second blocker is like again if you get compliance you need to have amazing fiat on-ramps and then that again more compliance because you have to then partner with a bank like and and circle etc.

36:59 Right. So that's kind of the those are two blockers basically becoming big enough that you need to actually you have the resources to do it. That's pretty much it I would say. Yeah that completely makes sense. It's kind of like the hurdle that every DeFi app has to has to get over. And obviously what will happen is only a few of them will get over that hurdle.

37:16 And I mean that's the beauty of having the experience on chain. I think I mean in the ideal world is like all that can just be circum circumvented. And to have an app that just goes straight to everybody and puts them straight on chain maybe with some like guard rail so they don't like using some like scam you know memecoin stuff like this.

37:36 But I think that's the dream. Obviously, when you talk about, oh, we want to go into like the app store and we want to be able to actually do this and go to market, I think a lot of regulation gets in the way of that. So, I think it probably will end up being that, you know, only a handful of applications really get enough steam and and get to go to that distribution scale.

37:54 But I think I mean if the more that we can and I think partners like base and and having these like much larger entities in the space will probably pave the way for that and do a lot of the cost cutting and maybe we can get the cost down to a reasonable level. I mean I think there should there should be costs associated with it but it shouldn't be so prohibitive that like barely any applications can do it.

38:18 But yeah I think I think that's probably what we see as the next step for crypto and DeFi. I think we're at an interesting place now where we've tapped out most of the people who really feel like they are invested in privacy and invested in having their own custody. Those are the people who care enough to actually come on chain and open up their laptop and download MetaMask and start, you know, looking around on Twitter for what kind of apps there are, looking on Dar, looking on Bon, finding Avantis and and using it on base.

38:44 I think we've sort of tapped out the addressable market of those users. Absolutely. It's about how do we monetize that enough and and get a good enough product market fit from there to expand and and like actually draw in the the rest of the financial ecosystem. So yeah, I think we're in a very interesting spot right now and the regulatory environment like you're saying will it will decide what that ends up looking at like.

39:06 So interesting couple of years we've got here. I mean we're rooting for you guys. We want to see more more of these apps make it into the app store, make it onto onto mobile phones and make it into the hands. and it's all about utility. like how do we actually give a good offer to to the market and and get it over the hump there.

39:22 Yeah. So, I mean that's that's super insightful to see like what you guys are sort of battling with on the longer time horizon. What is it on the dayto-day for you guys? Is it just shipping new features? Is it you know crafting like your your offer to traders working with market makers? What does it look like for you guys dayto day?

39:36 Yeah, so I think as we head into like basically decentralization like launching the Avantis Foundation, it's quite the lift. It's it's a heavy lift. We're a team of like 11. So like you know everyone has specific roles and then I am kind of orchestrating a a lot of it obviously as one of the founders. The day-to-day is more like in the next two months launch a bunch of cool campaigns for our traders stabilize Europs like manage risk on that try to get more market makers on board onto the exchange do a lot of cool marketing unique marketing on like Twitter and like new channels to to even reach people onchain.

40:06 I know that it's tapped out but then we haven't tapped into the full potential of the onchain market which is a very big market. Hyperliquid is doing thousandx the revenues we are doing right or like a magnitude of like 500x or something crazy. So there's still a lot of room to grow and like prove that we've built something really special and then obviously working with base giving them feedback on stuff like flash blocks and like helping kind of progress that because we know it's a it's a big product unlock for us as well and we're one of the earliest like testers of of of that technology.

40:32 So I think day-to-day is more like just running the ship and like running the machine and making sure it's not breaking. And then given this new launch, we're like kind of stabilizing it and like making sure we understand what traders are doing and their trading patterns and stuff. That's the day-to-day. Yeah, definitely. You mentioned Avantis XP a couple of times.

40:49 Maybe you could tell us like all the ways, you know, people can can earn that right now and maybe like what's coming in the future too. So it's very simple like all roads lead to XP. So basically you can like trade, refer, you can put your money into the vaults. All of that has XP and all of that is XP that is mapped out in a way that it really values your contribution uniquely.

41:06 What I mean by that is a trader is not the same as a liquidity provider, but they're both very special and needed in a marketplace. And so we have liquidity XP, then we have trading XP, which are not funible with each other. That should give you a sense that we think about those two entities separately. Like season 2 will end which is our final kind of point season before we decentralize.

41:24 It'll end on May 31st. So there's still like about two months left to go and then we'll just we have a bunch of campaigns planned even after that from an incentives point of view like liquid incentives we have op incentives grants trading competitions especially for zero fee perps like high leverage traders inviting them funny thing is we also give XP for liquidations cuz we understand it's hard and even though like you know it doesn't contribute that much to our revenue but like we we give a lot more XP for liquidations cuz it's just motivating for traders to know that there's some rebate coming for them in the future.

41:52 So yeah. Yeah. And it keeps them coming back, right? Even after you get liquidated, you you still feel like you're indebted or at least you feel like you want to keep using it, keep the score going up. I think a lot of people have done a lot of research and put a lot of effort into the gamification of tokconomics, but I think there's much more to explore with the gamification of applications and like using the application.

42:15 And I think the XP model that's pretty interesting. Did you guys take any like inspiration from anywhere with that or is that just totally home? I think XP has been around for a while. Points are offchain and they've been around for like a year, year and a half now. No inspiration really. When we like crafted points, we just thought like what do we value as the protocol?

42:32 What do like what kind of behavior do we want to incentivize? Like more organic taker behavior, not like spammers. We don't care about spammers on the platform. And so just want organic trading volumes and like you know there's a lot of people like if you look at our like 3 4 million and the volume yesterday was half a billion yesterday and I kind I'm not a big fan of that to be honest with you because I know what people are doing.

42:51 They're like spamming zero fee post per they think there's XP coming and like I'm about to tell them that it's not you know like it's not coming chill out like trade organically. Give us good feedback on what we've built. I mean give us feedback on would you use zero fee pers? Are you a loss averse trader? Does the slippage optimization make sense?

43:07 How do we compare to the platform? So getting that type of feedback is what we value and those actions that kind of you know lead to us building a sustainable exchange. Yeah, I love that grassroots approach because I mean I don't even know how many billions of dollars we've seen just get lit on fire with like all these different incentives to just bring liquidity over to a chain and then bring it back the next week after you've collected.

43:27 The grassroots approach is the right right way to go there to build like loyal users. I think that's what really matters in the market today with you know as all the per venues they start to list more assets they start to you know cut down on fees and you know they're all trying to compete there. having a loyal user base that's sticky is like incredibly valuable.

43:46 So yeah, I think that's the right approach that you guys are taking there. And I saw you guys also mention something about AI agents for trading. Is this something you guys have put a lot of thought into? Where where do you see the intersection of AI bots, AI agents with like how does it actually collide with trading venue?

44:05 So I think I think no one's given any thought to AI agents properly, even the ones that are live today. I think AI in Twitter is like just AI in crypto is just like right now at least it's a Twitter personality with a token or like data platforms that are masquerading as AI platforms but they're just data platforms that allow you to kind understand someone's identity on chain.

44:22 Kaito is like the closest thing that comes to like a crypto AI platform because it's all sentiment driven and like focuses on influence and who has the most yaps and who's like actually influencing Twitter's algorithm. That's definitely there's a lot of AI involved there, I'm sure. But when it comes to us, we want to think about like how do we take all of the data that a trader has and educate them on the platform and increase their retention.

44:43 So you can think of it as like we'll have our own version of dueling goes owl which will tell you traded this way, you traded that way, this is how much XP you got. Are you sure you're doing it the right way? Can you this this right? Or like yesterday you traded like Bitcoin right when Trump announced Aar like that didn't work out too well for you.

44:59 You know just do something so to make them feel like it's not like us versus them because it's not we're building an aligned exchange that's how we think of AI agents and then the second would be high level again we've not scoped out a lot of this because again just going step by step in the journey but the second thing would be something cool like and I I'm open to sharing ideas because ideas are like everyone has one you know it doesn't matter execution is what matters but another thing we could do is like a voice to trade AI platform like a TGbot you just like record a message you send it the trade's opened on the subway, open a,000 x long, close it at 1% stop loss, that's it.

45:32 And you just open, you're like, "Oh, actually did not close." You know, those are more experiments, but I I do think the retention aspect is very cool of like having a personalized assistant who can guide you on the platform. That to me is cooler, I think, than anything else. Yeah, I really like that. Like giving the trader the tools that they can't get anywhere else.

45:50 And I think you know when you've really struck a chord where even if they're trading on some other venue, they come and they open up the terminal for Avantis and they are you know talking with their AI agent like it's you know like your adviser or something like this. I think that's that's a really cool idea and I think fundamentally that's what AI is going to be at least for the next couple of years.

46:10 You know everybody's talking about how AI is going to completely replace people. I think we're at the point now where AI is it's a way to supercharge people and it's this is kind of similar to our approach to to what AI has to offer as well internally. we've we've been working on and we've got kind of like a v1 of like a sort of like an AI assistant for audits that we do where like on an engagement it will basically ingest the codebase and basically just serve as like an LLM that has all the context questions for for our researchers and everybody on the audit and it just it's almost like having another teammate in that it just reduces the the feedback loop from like hey I have this question I ask another human being

46:49 Maybe it takes five minutes maybe it takes them an hour to answer to I ask a an AI bot and it gives me the answer within, you know, a minute, 30 seconds or something like this. And I think that kind of a feedback loop helps you scale organizations much faster because it's all self-contained. I think the hardest part about scaling organizations is like the communication and like all the nodes in the network.

47:10 So I think AI is really it's a tool for empowering humans at this stage. I don't think it's replacing at least a lot of jobs yet. So I think that that approach is definitely fruitful. And it sounds like if you guys can get it right, like that could be that could be pretty powerful and it's also kind of a mode that's like hard to replace if you have all that data and you can get into like tuning the model and even stuff like that.

47:31 Yeah, that's really cool the way you guys are looking at that. And then finally, I mean I was just interested like what's what does the team makeup look like? Where are you guys trying to go with like hiring? Are there any positions that you guys are trying to fill that have been hard to fill? I know hiring is pretty hard in DeFi.

47:45 What's your experience? Yeah, it is. It is, man. Harden's unnecessarily expensive I think because of how few good people are there. It sucks but yeah we so team is like 11 people half and half or like 12 people half and half the ops marketing versus engineering most of the engineering is like Europe and India and then most of the marketing and BD ops is New York and Canada.

48:05 Yeah, that's the high level. And then we're we are hiring for like strong backend and front end engineers, one each ideally. And then we'll be in pretty much steady state or avant this at least. And then obviously we spin up a new entity, build something new, it'll require new people, but for the exchange, I think we're fairly good to go.

48:21 Yeah. And of course, if anybody listening is interested, I mean, we've got a pretty big like technical audience here. If anybody's interested in the vision, what you guys are building, where could we send them? Send them over to me on Twitter or via my email. I'm happy to share that with you guys. So, both work. You can also DM the Avantis account.

48:39 So, Avantis FI on on Twitter, avantisfi on Twitter. So, you can just send your resume in there or to my personal Twitter or to my email. Sweet. Sweet. Yeah, we'll definitely have those linked down below the website and personal Twitter protocol Twitter. And then, is there anything So, you mentioned there's like a season going on right now for XP.

48:57 Is there anything any particular offer? Obviously, we've got the zero fee per that's a big release coming up. When do you expect that to be out? That might be out by the time that we release this actually. I think it'll be out like next week hopefully. So yeah, gearing up for a public launch and maybe a week maximum two weeks.

49:13 That's awesome. Okay. Yeah. So I think you know if anything it's probably out now if you go and and check out the Avantis website. I think it's avantisfi. Is that right? Avantisfi.com. That's what it is. zero fee pers. So I mean you can go check it out, play around with it. Don't lose any any more money if you lose.

49:31 Yeah. Yeah. in this market, it's already gone. So, if you have any left, please try out your luck and give us good feedback. Absolutely, man. So, we'll we'll send them over. anything else? Any any anything else you want to shout out or or we can send them to? No, I think and then I think you guys covered everything.

49:46 This is a lot of fun. Actually, a lot of fun, guys. Yeah. Yeah, it was a pleasure to have you on, dude. And and really cool stuff that you're working on and we'll be checking out Avantis and all the cool stuff that you guys are are shipping in the future. Definitely. Yeah.