Join Delphi Research today and immediately get access to our full Member Portal!
Join Delphi Research today and immediately get access to our full Member Portal!

Yenwen Feng and Nick Tong: Introducing Perpetual Protocol v2 “Curie”

Jul 12, 2021 ·

By Jose Maria Macedo, and Jonathan Erlich

Jose Maria Macedo and Jonathan Erlich sit down with Yenwen Feng and Nicholas Tong to discuss all things Perpetual Protocol, the largest decentralized perpetuals trading platform by volume. They explore Perpetual’s success to date and dive into all the details behind the upcoming Perpetual Protocol v2.

Every Delphi Podcast is dropped first as an audio interview for Delphi Digital Subscribers. Our members also have access to full interview transcripts. Join today to gain exclusive early access to every new interview before it’s published.




 Music Attribution:



Interview Transcript:

Jose (00:00):

All right everyone. Today I’m very happy to be here with Perpetual Protocol. We’ll be interviewing Yenwen and Nick alongside my colleague, Jonathan Erlich. So we’ve been working with this team for a long time in various capacities. They’ve achieved insane traction for their v1 version of their protocol which we’ll get into later. And this week they announced v2, which is a pretty big departure from v1 and also like a very interesting design trade-off. So in this podcast we’re going to be digging into all of that. Yenwen and Nick we are happy to have you here.

Nick (00:35):

Thanks for having us.

Yenwen (00:37):

Yeah. Thanks for having us.

Jose (00:39):

Cool. But before getting into all things perp, I’d love to learn more about your background and how you got into crypto?

Yenwen (00:47):

Sure. So let me start first. So I’m the co-founder of Perpetual Protocol. My co-founder and I, we actually sell down the crypto rabbit hole in 2018. And we actually, working on an option protocol at that time. So kind of like, it will be ahead of its time. if I could get myself like three years ago, I really don’t know what I was doing at that time. And the market crash we don’t really get any fundings. So actually the project failed, but we do want to stay in this crypto ecosystem and so we actually go to Binance Lab to join their incubation. So we actually working on the crypto accounting software at that time. It’s actually working okay. So we work on this for a year, but in 2019 we see contraction of Uniswap and its [inaudible 00:01:55].

Yenwen (01:55):

We found that actually AMM is quite interesting that it’s kind of like at the center of the whole the DeFi. If we can be able to AMM for all other financial instrument I think that will be more interesting. So that’s actually why we drop the cryptocurrency software and then we go back to pick up the derivative. And in 2020 we think that perpetual swap is probably the most easiest one. If we want to build AMM on top of it. So that’s why we’ll just go ahead and build AMM for the perpetual swap. Yep. So fast forward to now the change launched last December. It has been running for like six months already and the traction is quite good. We have around like 20 billion trading bargain. Totally I think that’s a really good for a new like protocol. Yeah. Everything turns out to be working well. Yep. That’s fun, man.

Nick (03:05):

Cool. Just a quick background I’ve been kind of in and out of crypto. Started in kind of 2013 part of the Mt. Gox hacks. So I kind of left for a bit came back in kind of 2016 and kind of ran a, like a friends and family kind of arbitrage fund. Joined at FinTech after that. And then more recently joined back again, predefined summer kind of exploring the space and then kind of joined Perpetual Protocol as head of strategy after that. Yeah. That’s my journey so far.

Jose (03:43):

That’s awesome. Yeah. I really appreciate those intros. And I think perps are a complicated product and I think the implementations of decentralized perps are quite complicated. So we want to start at a high level and kind of narrow down and go through v1 and then like the reason for v2. So to begin with, could you define what a perpetual future contract actually is?

Yenwen (04:07):

Yeah. Hopefully, I can explain it well. So professional contract is actually being like perpetuated by BMAX in 2016. So at the time they have, it’s actually like a beginning of crypto and, its not the beginning, but at the time Bitcoin is big, but also like you have like other old coin. But I think that at the time it’s really hard to, it’s not like right now. Most of the token ideas [inaudible 00:04:47]. So it’s really hard to have change that change. Older financial instruments. So they come up with this idea that they can have a future. It kind of like inspired every day. So it actually inspire very quick, but in the long run it actually kind of like you can chase another index. Which means that they can create easy change and then people can trade other asset that they don’t have. So you don’t really need to have the underlying. So these like inspiring everyday futures can trace that index off everything. So I think that’s a recent that they want to invent this. And also at the same time it’s a future so you can collaborate. The cost of trading this instrument is pretty cheap and you can get an average. So that’s why it gets so popular within a centralize change.

Jose (05:52):

Okay. Very interesting. And could you give just an idea of the, because obviously products didn’t really exist before crypto, if that’s correct then, but they’re like pretty much the highest traction product within crypto, right? Like why do you think it’s been so successful?

Yenwen (06:09):

That’s a good question. I think that it’s actually very easy to use. If you know [inaudible 00:06:20] rate, how it works. I think it’s that for most of trader they can understand. Like I said the transaction cost is really low. So I think traders who want to take leverage. I think this is as actually the preferred way in crypto world that you can just go to central change and then take, most of the change they offer like hundred X average. So you can kind of quieter the average you want. So I think it’s just like easy to access and then the concept is not that hard to get. So it’s really popular among the traders, but still it’s not a financial instrument for retails. Although there are lots of retail trading this, but high average means that you might get hurt. Like also your asset, but that’s popping up so good to retails, but anyway, that’s what I think.

Jose (07:28):

Okay. Cool. And now, going into perp v1 and how you manage to allow people to trade perps in a decentralized way, can you explain a bit of the vAMM architecture how that works and maybe even like how you arrived at it and what other options you considered and like some of the problems you had to solve? That’d be really cool.

Yenwen (07:49):

Sure. Sure. Sure. Yeah. So it will be a long answer. Okay. So perp v1. So we have this idea here. So it’s actually based on the new design of AMM. So we called this virtually AMM. By virtually means that it actually means a lot of virtual tokens in this model. So how it works is we actually have kind of like two components. So the first component in this kind of like occurring house or assistant that we can meet virtual token. So like trader they can put in USDC as collateral. And then based on the collateral they meet and also the leverage they want. For example, you’re putting like a hundred USDC. We actually can meet, if you want to trade like 2X average, we can meet like 200 virtual USDC or we can meet like 200 worth of virtual ease.

Yenwen (08:47):

So it depends on like long short or like depends on the market you want to participate. So, that’s the first part. So we have this virtual tokens. And then the second part is actually the pricing engine. So we can put this virtual token into like, so in perp v1, we have this SYK price engine. So if I want to take a long position we actually put in the virtual USDC and then I guess some virtual ease out. And then that virtual ease, it’s actually the one position I own. And after a while I can do it like reversely. I can put P/Es and then get our biggest VC and then that will, the difference will be my PML. So that’s actually how it works. In general, so yep.

Jonathan (09:40):

Hey, Yenwen, can you manage to sort of for example, how a market is set up in perpetual v1, how the process looks like from your side, because I think that sort of sets the stage sort of for v2 and sort of for the, yeah. For the concepts that went into creating V2?

Yenwen (10:07):

They say that you got to know how the maker work or.

Jose (10:13):

I like it. What’s the process to start a market and process v1 like how, and in perp v1 How the market actually initialized?

Yenwen (10:25):

Okay. Got you. Yep. So in perp v1 older market, I think like it has to go through the voting process. So we have this like governance voting. So all the stake holders. Someone has to create a vote first and then people can vote on that. And then once he passed the team actually get to work and then try to create a market for the soccer holders. So that’s how it works out. So we actually have lots of like voting on like which market that people want. And so after it got passed we just launched that market. Yup. So it’s kind of like a manual process at the final end. So that’s definitely something we want to make it better.

Jonathan (11:21):

Nice. And there are also some important parameters that need to be sort of defined before creating a market, right? Specifically sort of the gate parameter that sort of defines the depth of each market. Can you go into sort of that process and sort of what that entails for v1?

Yenwen (11:46):

Sure. Yeah. That’s a good question. That is a hard one as well. So like I said, so we have like this two component. The pricing component as one. We are using the SYK model, which is like Uniswap of v2. The interesting of this is that you Uniswap how they decide K is that by the decreasing provider. So the more we quickly they provide the higher the pay. So it depends on the liquidity. But because, like I said, all our tokens are actually virtual token. So we actually can meet those virtual tokens. So you actually kind of like interesting that how about we actually meet the token free market? So before we launched that market we actually meet a bunch of tokens and then putting there. Then those tokens will become the liquidity of this market. So that’s actually the thing that John mentioned that we can actually meet lots of virtual token before we enabled this market. So we actually have a process that’s, we actually detailing like in your community that so we calculate the slippage. Kind of like for this market, how much slippage we’ll want to have, for example, putting like a hundred thousand USDC. If I want to open a loan position with a hundred thousand USDC how much slippage I will have. And then based on this we will decide that X, Y, K that’s K per meter. And then we will create a market for that. So that’s the process. Like John said, it’s actually very manual. We don’t really have a, it’s a positive right now we use the slippage, but there might be other way, but we just, we don’t really have lots of things that can pick up, like using other methods, but anyway that’s what we have right now. And then that’s also something want to fix.

Jose (13:56):

I just want to double tap on that. Because I think it’s a really important point to understand v2. You explained it really well. The perpetual system where there’s these two parts, right? You have the clearing house, which is kind of an accounting system that’s used for settlement. And then you have the vAMM, which is mainly used to price, right? To price this liquidity. And it’s virtual in the sense that there is no liquidity in there and there are no liquidity providers, right? Which is why this K parameter has to be set sort of manually. And could you maybe just talk a bit for people who, for whom it’s not intuitive, like what’s the impact of that K parameter and like what factors do you take into account when you set it?

Yenwen (14:39):

Sure. So for K case if we, so basically we want to have the K as large as possible. If we can cake and you want to be like a number of approach infinity. Why is that? Is because the larger the K, the slippage. The faster took a traders. So, because we want to provide like, we offer you slippage to a traders things that the platform is very efficient. So definitely the larger the better, but there is a limit. It cannot be like, kind of like a number close to infinite because if we have like such a large number which also means that because this is a AMM so every time you want to move to parts you have to create a trade. So if we have a really large K, which means that we need to provide lots of order of trades to move the price.

Yenwen (15:44):

So, which actually make it really hard for price to move. If the price doesn’t move the market actually crash. So we need to find that balance. So the larger the pay with all the slippage, but at the same time we need to file reasonable pay. That’s actually, the trading button that we have right now is enough to move the price just like ashing normally. Just like other market. So that’s just to show you how its [inaudible 00:16:12]. So one way we use it’s just like by looking at slippage. We want to keep it small, but not really that small because like, it’s kind of like too small actually. We need a big firm, or like several bit firm to push your price to kind of like to you market make I think that’s really hard.

Jose (16:37):

Absolutely. And it creates some risks for the platform, right? If your liquidity, you can sort of match external prices quickly enough. And right now, I guess that risk is being taken on by the insurance fund, but that sort of changes with perp v2. Obviously perp has grown significantly since launch you facilitated nearly 20 billion in volume. You’ve generated nearly 20 million in fees. You’re the largest decentralized perpetuals platform by trading volume. And you’re the first fourth largest protocol by fees generated overall which is massive. So what do you attribute the success of perp v1, v2 and how has this big growth been experienced from your team’s perspective?

Yenwen (17:25):

Yeah. That’s a good question. So I think there are like several reasons. Of course, I think the first one is that we’ve launched the protocol last December, which is actually a very good timing. Of course, we don’t know that, but to be honest, in February, March, actually people trade a lot. Crypto goes up like two or three, four. So I think that’s a, we need lots of trade involvement. Traders are looking for value to trade. So of course they can trade in centralized change, but they also kind of looking all their alternative, value they can trade. So definitely one thing the market is good.

Yenwen (18:12):

The second thing is that we are actually the first like protocol maybe not the first, but I think it’s more at the large scale, deployment on layer two or [inaudible 00:18:24]. So we pick a SKY. When we launch we are just like only a few. DeFi protocol launching on the side chain. We pay for the gas on the side chain. So people have experience out there and they don’t really need to think about gas. So they basically just trade and then a SKY is actually three times faster they in the east. This one time it’s not instant, but it’s still much better than [inaudible 00:18:57] network. So I think that actually helped with a lot of people. People can’t have their trading strategy they don’t need to sing about like is it too expensive for me to trade, to open this position? I think that’s really great.

Jose (19:13):

Perpetual protocol has grown significantly since launch. You’ve facilitated nearly 20 billion in trading volume. You’ve generated nearly 20 million in fees, which makes you the largest decentralized perps platform by trading volume and also the fourth largest protocol overall by fees generated, which is obviously huge. What do you attribute that success to and how has it been just experiencing that kind of hyper growth?

Yenwen (19:37):

Yeah. That’s a good question. So I think that, of course there are lots of reasons, but I think the first reason is that we actually launch it at a very good timing. At the time when we launch, we don’t know yet, but the crypto and also DeFi tokens, like have a really good round in February, March and April. We launch last December. So we actually, just like launching a very good time. So in the cycle bull wrong, lots of traders they are looking for value to trade. So that’s why they come to us. So that’s definitely like one of the reason. The second one is that when we launch we are actually one of the few that launch on the side chain or were able to solution. Like if you go to like Perpetual website we’ve launch on a style and you actually don’t need to pay any gas fee. We pay the gas fee for you. So if you want to open a trade you just like click on.

Yenwen (20:47):

You don’t just sign a transaction and then we will actually send that transaction using our system. Because you don’t really need to think about the fees that you’re going to pay. You don’t need to think about that is it worth it to place this order? I think that actually helped us a lot from the usability point of view. So that definitely helped. Yeah. And yeah. I think the third thing is that we are bringing in. So to AMM that we don’t really need to have a heavy market maker to really go in and then have that market. Actually, at the time we were first launched. We actually kind of like launch like one or two market per week. So that’s the speed that we can launch new market. So we can attract lots of traders, but at the same time it doesn’t really create lots of overhead to the team or like we need to go out and find market maker who are willing to trade that market. So I think that’s definitely going solve the recent that we have this round.

Jose (21:58):

Nick, do you have any thoughts that you want to add?

Nick (22:01):

The only other thing I’d say is kind of a little bit more kind of on the product side of things in terms of kind of the product design and the approach that we’ve taken. I guess two of the big things that we really believe in is one kind of accessibility to financial products that normally like retail can’t access. And then secondly is kind of how do we simplify or make these products a little bit more understandable so that retail users can use them. And you can kind of see that in the product design. And I guess even the UI and the user experience that the users have gone through. And a lot of the feedback we get from the community is actually they really like the experience that they get through it. And it’s even small things like because we leveraged by economy users don’t need to go into MetaMask for example, and change their RPC over to textile. They can just trade straight through Ethereum even though the entire protocol is kind of on the X side chain.

Jose (23:00):

Yeah. Absolutely can confirm the product’s great really enjoyable experience. And you were kind of a trendsetter with your side chain move obviously we’re seeing the success of Polygon now. So with perfectly well, we kind of went into the design a bit and like how to initialize the market setting K. And I think that sets a nice context for talking about what were the problems with v1 from your perspective and what was the biggest learning, I guess maybe if you can touch on kind of the long skew issue and how that problem came about and then we can go onto v2 and how it solves it?

Nick (23:36):

Yeah. The long short skew issue is probably the biggest one that we kind of, the unintended consequences that we found. Effectively, the reason for that is in order to kind of move the price based on the AMM design, you need people to basically take either a long or short position. So you can imagine when we started Ether. We launched it the Ether market back in December. I think Ether about 800, it’s like 2K or thereabouts now. So you can imagine there’s a lot more longs than there are shorts. And so what happens is in the case where there’s like, let’s say funding payments there’s an imbalance of longs and shorts. And so what happens is the insurance fund basically acts as a counter party and it either receives or pays out payments just depending on kind of the market conditions and kind of the market price with respect to the index price.

Nick (24:33):

And so you can imagine then if you look at the balance at the insurance fund it doesn’t keep going up and goes down maybe due to like a black swan event. It kind of just oscillates. It goes up a little bit then maybe it goes back down and goes back up and goes back down. And that’s kind of a risk that we really wanted to address. And we think we’ve addressed in kind of v2. So I’d say that’s the first point. The second point is probably it’s not a problem per se, but it’s another thing that we’ve kind of really focused on with v2, which is around kind of capital efficiency. So the pricing engine that we currently have in v1 is SYK. And if we think of it in like an order book setting it’s as though we take equal amounts of liquidity and we kind of place it among the entire order book versus I guess to oversimplify you look at liquidity on order book it’s normally like a bell curve that’s kind of more concentrated around the market price. So that’s the other thing that we were kind of trying to solve as part of v2.

Jose (25:36):

Awesome. Yeah. Yenwen when do you have anything to add there or?

Nick (25:40):

No, I think that’s perfect, that’s actually the thing that we actually spend a lot of time working with you guys and trying to get it. Yeah.

Jose (25:49):

Absolutely. Yeah. It’s been a lot of fun. And so obviously in v1, the insurance fund was taking a lot of that risk, right? If fact it’s having to kind of balance out the market. In v2 one big difference is that you introduce like liquidity providers, right? And maybe I’ll just let you sort of explain v2 from your perspective and how it works for our audience.

Nick (26:15):

You go.

Yenwen (26:17):

Okay. All right. So for v2, I think that if you remember that I said so like perpetual assistant actually we have like two components. So actually the first component remained the same. So we take the USDC a collateral and then make virtual token, but we actually swap out the pricing engine. The pricing engine used to be SYK model. And then we swap that down and then we’re putting like Uniswap v3. So you Uniswap v3 they have this new design that actually, like there’ll be crypto provider or maker can decide that which range they want to provide their liquidity with. So they actually open up lots of design space that the maker actually can do a lot of, their own strategy it’s not like just a sweat, evenly close that curve.

Yenwen (27:16):

So I think that’s actually helped to solve lots of like all our, the problem we’ve mentioned before. So yeah. That’s actually a, for the v2 the change that we have, but of course it’s kind of like overview because I think there are lots of details around this. One other thing that I do want to mention is that because of, we have these two components. So actually the liquidity provider or maker they can provide liquidity with an average. A kind of like they can provide average of 10x. Which means that they can of course are 10 times small fees, but definitely you can resend if they have to think about that. But it definitely created a loss of leverage. So that they could a provider in this situation.

Nick (28:06):

I’d add kind of two other key pieces that we’re kind of planning on doing as part of v2. And the first one is I guess just general enhancements to kind of get as a feature parity in terms of centralized exchanges. So we currently run isolated margin only. So allowing cross margin, allowing multi collateral. Its different collateral types. Again, it opens up the design space. And then the second piece, which I think is actually quite interesting that we won’t have obviously in centralized exchange is permissionless markets. And so as Yenwen alluded to before, like the way that you kind of put a market in and create a live right now. It goes through a governance vote.

Nick (28:48):

And there’s kind of some permission in that sense. And the permission is necessary because all of the markets basically share the same insurance funds. So if we just listed any old market and it kind of went to zero, it would kind of drag down the entire protocol and itself. And so we’re introducing private markets and the beauty of that is anyone can kind of create a market whatever they want. And that entire market itself is isolated. The insurance fund is completely isolated. And what that means then is that, if that kind of goes down it doesn’t affect the main protocol. So those are the other two pieces I think that were the key points to kind of v2.

Jose (29:28):

Yeah. Those are super interesting. So obviously you kind of replaced the price discovery model, right? The pricing model with, rather than SYK now you have Uni v3. And so that eliminates a lot of the problems with kind of like setting K and the long short skew that results from that, because now you have these LPs that can come in and effectively make markets, right? You have like this maker that you didn’t have in v1, that’s kind of taking on some of the risks and also some of the rewards of the insurance fund. Yeah. How do you think about sort of attracting those liquidity providers because you haven’t had to do that before, right? Previously all you needed was kind of like traders and liquidity providers in v3 are kind of like traders, but yeah. How do you think about that side of things?

Yenwen (30:16):

Yeah. That’s a good question. Definitely, there are several ways like trying to attract a liquidity provider is definitely, we will have like liquidity mining when we launch. We give out token rewards for people who want to provide liquidity, but in the long run, actually, I believe that if you think about like traditional market maker. They all like always centrally change and then they market make and they actually lost money by market making. And we actually have this model open up to like old traders. This meal, like Uni v3 kind of a model actually is kind of like slipping, order book and AMM. So it’s actually have lots of similarity with the limit order book.

Yenwen (31:16):

The traditional market maker or like anyone they can come in and then just find your strategy. For example, try to balance your own short position. You should be able to earn enough fees enough for you, for your participation. And also in our model we actually just a Uniswap or like SushiSwap. Liquidity provider actually got most of the fees compared to a centralized change. Centralized change [inaudible 00:31:48]. And we maybe like, I don’t know like five based point or something like that. Or 25 baseball, but here I kind of like, you get all the fees. Most of the fees. So you actually, we encouraged you to kind of like build the strategy that works. Personally, I think there are lots of opportunities in this area and lots of things actually are building like on chain strategy for everyone. So I believe that there should be like lots of opportunities in the makers side.

Jose (32:23):

Yeah. Yeah. That’s awesome. And competition is obviously heating up in the decentralized derivative space. You have several decentralized perhaps platform, either launching or announcing new versions like dYdX has an order book model, right? On top of Starkweather, MC decks, I believe is launching on top of Arbitrum. How do you think about like the design trade-offs there and kind of where you sit?

Yenwen (32:46):

Yeah. That’s a good question. Yeah. I’ll go with the first, so dYdX they have like maybe an order book. I think that’s more of like a traditional centralized exchange way. So maybe like more decentralized. Starkweather is pretty cool. Not a lots of project builds on Starkweather because it will be difficult to build on top of that, but definitely that’s an interesting choice. MC Decks is actually more like AMM so their design is kind of like a constant kind of a Bill curve style. Kind of a cookie provision that more people can provide liquidity using their Arbitrum. I think we are actually sitting in the middle. So we are a combination of a Uni VC and virtual tokens. So we actually have some of the characteristic from may be order book.

Yenwen (33:47):

And also some of the characteristics from AMM. For order book the liquidity provider, or maker, or market makers, they can pick the strategy they want. They can place there, kind of like even older on the curve. So any place they want. So it’s actually a loss like in the order book, but at the same time the system is still AMM. So you need that trade to move the price. And we kind of like, if you’re a look at liquidity provision you can kind of get a sense of like the slippage. And also we don’t really need a market maker first to kind of boost liquidity. We can work with like strategy providers like Charm or visor to provide people strategy wise. There is certain amount of price liquidity there that we can launch our market. So I think that’s kind of like the trade-off. It’s not about [inaudible 00:34:50], but different like a solution have like different trade-off.

Jose (34:55):

That makes a lot of sense. Jonathan, did you have a question or?

Jonathan (35:00):

Yeah. I was wondering like from your perspective sort of what the main advantage was of adopting sort of this Uniswap v3 style model versus a centralized limit order book. And from what I understand it sort of in allowing retail participants to provide liquidity and to participate as market makers in an easier way, and wanted to sort of understand if you agree with that and yeah, how you see I would say like?

Nick (35:41):

Yeah. Like a hundred percent agree with that. And I think we touched on this point before, but kind of I think one of the trade-offs with a centralized exchange is in order to get like a good trading experience, you need like an experienced market maker. A it’s quite expensive. B it takes time. C you kind of have to then lend them inventory for them to kind of trade against. And so that means that you kind of exclude a lot of assets that can and can’t be traded. And Uniswap kind of yes, they have all the major pairs that do a lot of volume, but like they’ve done phenomenally well through kind of the long tail of the assets as well. And so having this AMM model and having composed, like being able to be composable on Arbitrum means we can kind of build an ecosystem that then allows anyone to kind of build on top of purpose, kind of like a base layer Lego. And effectively then launch whatever markets that we want. And it means that like we don’t, for those that are less technically inclined we’re still being inclusive of them because maybe we have like a default on chain strategy that they can use or one of the partners has a default strategy that they can use. And so everyone can still kind of participate in this ecosystem and it’s not limited to a small set of users.

Jonathan (37:03):

That’s very interesting. Nick, could you get a bit more into sort of the, like how this model allows a perfectly to sort of to scale into do a long term of asset markets on the other hand could you dive a bit deeper into sort of how you see compostability playing within the decentralized perpetuals market and how a model like Uniswap v3 sort of helps perfectly to sort of be compostable within the ecosystem?

Nick (37:37):

Sure. Sure. So I guess with the long tail of assets, it’s kind of what we touched upon before which is around kind of private markets. So the plan right now is staked perp or users who kind of who staked perp receive stakes perp. And then they can stake that into the private markets to effectively act as a backstop and obviously receive a percentage of the fees there. So they increase their risk slightly, but they receive a portion of the fees there as well. What that does then is once there’s enough kind of stakes perp in that market, that market can launch it’s a private market. All it needs is a price feed and you can then kind of launch whatever market you want. I guess the beauty of this is again, we can then start capturing assets and launching kind of perpetual assets a lot sooner than kind of other players where traditional you might need a little bit of track your record, protocol’s going to be live for a little bit. Then you kind of have to raise funds. You’ve got to then find a market maker and so forth and several.

Nick (38:45):

So the time to actually get a product live, with the time actually reduces quite significant. So that’s probably the first part of your question. The second part around kind of composability and the ability to kind of build composability. I think there’s a couple of benefits with Uni v3 as well in there. There’s lot of research that’s kind of been going on into the space as well. And so we’ve seen at least the ability for all the partners to kind of port over let’s say if they’re building out a market-making strategy for a Uni V3, for example, it’s not too short of a hop to kind of import that over to perk and build on top of it. I think it also, the composability piece means that the design space actually opens up significantly. So before where you might need like a fund might has to have a lot of its structure connected to a lot of kind of disparate systems. Now it’s a lot easier for kind of engineers to kind of plug and play. Like I want to pull in compound and I might pull it out and I’d like a fixed interest market. Plus I might pull them perps and then kind of create some crazy strategy out of that. There’s so many kinds of possibilities that we haven’t probably even thought of that are now available kind of with this composability.

Jonathan (40:04):

Yeah. Yeah. Totally. Yeah. I agree. I think there’s a lot of sort of creativity to be discovered within the decentralized perpetuals market that we haven’t seen yet and that the launch of protocols like perp v2 are going to allow, so yeah. Really exciting.

Jose (40:31):

Yeah. Absolutely. I definitely agree with you on the trade-offs of AMMs versus central limit order books where I think AMMs are going to be able to tap into a lot more retail capital, right? Because there’s sort of an easy way, kind of that strategist. Middleware layer. There’s an easy way for retail to just onboard liquidity onto these strategies in a way that there isn’t with a central limit order book, right? Where it’s much more difficult to have those unchanged strategies. And so you end up having to have both the market makers and the capital providers be the same person whereas with Uni v3 you can kind of separate that out, right? You can have a market maker, like Charmer or Visor, and then you can have a capital provider separately. So I think you end up tapping into more capsule which is super cool. In terms of scaling part you were obviously super early to this. You bet on MadTech kind of and onside chain design before anyone else did. And obviously that’s seen a lot of traction. How are you thinking about scaling perp like long-term, are you thinking about other chains? Are you thinking about any layer twos? Yeah. What are your plans there?

Yenwen (41:34):

Yeah. That’s a good question. We have being actually SKY for some time. And I think because the new design of perp v2, we actually need to use Uniswap v3. So we have PPL on the chain or a side chain or a little too that has Uniswap v3. So I think that the decision that we have right now is probably go to operation first, because Uniswap announced that they are going to deploy Arbitrum and then actually deploy it. So I think that’s probably the first one, but I think we will definitely get other layer two or like side chains that Uniswap are going to deploy from. So that we can utilize that.

Jose (42:26):

Very cool. Okay. And we’ve seen a lot of growth and adoption for decentralized spot trading protocols, obviously like Uniswap kind of started that whole thing. But we haven’t seen the same trend materialize yet for decentralized derivatives trading. So in the centralized exchange world derivatives make up multiples of spot volume, right? Whereas in the decentralized world they’re still a fraction. Do you think there’s anything about the decentralized world that makes this the case, or is it just like a case of these products getting better? And what do you think needs to happen for the decentralized derivatives to start catching up?

Yenwen (43:05):

Yeah. Maybe I can go first. I don’t really like 100% have to answer. I think we have like the protocol has been running for six months. But far what we learned, I actually think that the usability is actually the main issue that scale. People compare us to exchange institutions, like FTX. If the trader is got free king on the decentralized properties like promotional as all this. It probably will stay on FTX no more, its better UX. Like it actually has more liquidity there. So I think we still have a long way to go. That’s actually why we decide we want to have this v2 is trying to get, most of feature on FTX, bring those features to Perpetual Protocol.

Yenwen (44:05):

Like close margin like a market caterer. It used to be like if you place a trade, you have to wait like five seconds, right? Even on the SKY. So I think that things are actually improving like most of months kind of like, but I personally believe with Arbitrum, new varieties, like risk a feature that’s kind of on par with FTX. I think that we should be getting more traction than before. So to get back to your question is I think the product is not mature yet. That’s what I feel. Nick how think?

Nick (44:49):

I completely agree. I think like we’re all kind of super crypto native here, but if I try to explain it to kind of non define native or non crypto native people, it’s normally like, “Hey, you have to trust your money to this Chrome extension and it’ll hold all of your money.” And you kind of tell people that and they have a second look at like you, “What do you mean?” And I guess until us as an ecosystem we can kind of solve this problem. This kind of usability problem. I think that’s going to be harder. I think we’re definitely going in the right direction. And it’s just going to take time as a whole for the ecosystem A, to kind of educate users. And B, to kind of improve on this experience as well.

Jose (45:37):

All right. So we’ve seen significant growth and adoption for decentralized spot trading protocols. Obviously, that started off with Uniswap. And it’s kind of grown since. We haven’t seen the same thing materialize in the same way for decentralized derivatives. In the traditional world, decentralized, sorry, in the traditional centralized world derivatives volumes are a multiple of spot volumes. Whereas in the decentralized world they’re still like a fraction. So why do you think that is and do you think that there’s something sort of intrinsic about the decentralized that makes derivatives less attractive or is it just a question of these products improving to catch up to their centralized counterparts?

Yenwen (46:19):

Yep. That’s a good question. Of course, we only have six months of experience. It’s not like 100% for sure, but personally I think it’s just like in our product is not there yet. So lots of our users they come in and they compare to us like centralized change like FTX. Which is kind of a really good, centrally changed. And they compare the features, they compare the speed. Even if we all at SKY every time you want to place a trade, you have to sign and then after signing you have to wait like five seconds. Compared to FTX is like within that second. We still have a long way to go. That’s actually the recent that we putting lots of features like trying to be at par with FTX. So we want have closed model. We want to have like USDC collateral trying to bring more features from like this centralized change. So to go back to your question, I personally believe that it’s just like, the product is not mature enough. We will get there. Just like trying to, we’ll call it and trying to get there.


I think as an ecosystem as a whole we’re still kind of a little bit immature in kind of the user experience side of things. I think all of us here are super crypto native, but if you take on a non crypto native people and you try to explain it to them, you’re basically telling them to kind of trust all their money and kind of this Chrome extension effectively. And so that’s kind of, usually people kind of do a double-take. And so I think as we kind of educate users more on kind of the benefits of DeFi and I guess in Ethereum and kind of decentralized finance in general. And as we kind of build our products as an ecosystem to kind of get to that better use experience I think that kind of liquidity and that kind of trading volume will happen that we see in centralized exchanges.

Jose (48:26):

That’s awesome. And I guess to finish up just like a really big question. Obviously, it’s hard to predict things in crypto. I don’t think anyone could have predicted where we are now a year ago maybe even six months ago, but how do you see the decentralized derivatives market looking like five, 10 years from now? Do you think there’s going to be crypto native products, obviously paradigm released their sort of perpetual options, product recently, we’re seeing some projects built around. Yeah. How are you seeing it? What are you personally excited about? Where do you think it’s going?

Yenwen (49:00):

Yeah. That’s a good question. Actually, I agree with you that there will be like crypto native deriviative product in like five years. Actually, its just like perpetual swap. So perpetual swap itself it’s, of course it’s a swap. Of course the same concept probably happened in like, sometime in the traditional financial market, but itself it’s actually a crypt native parlor. That’s the first one. And then the perpetual option you mentioned, like a Paradigm really did a good job of like trying to, nobody talks about this in a traditional finance world. Lots of difference from crypto, like financial DeFi to traditional finance. I think one of the differences is actually treatment by engineers and it’s 24/7. So people just trying to, for engineers, we want to have something that you can put out there and then you grants. Just like watching, right? So no central parking, like taking care of that. So I think it’s actually have lots of different thinking than traditional finance. So definitely they will be like more part of like this like perpetual swap. Perpetual option, perpetual insurance swap, I don’t know, but definitely I believe they will be like something new in this, patching world. And then that also will be the most popular one in the future. Yeah.

Nick (50:40):

I would echo that I think we’re in the very early innings of kind of composability and what it can do and what we can do with it. And because most products are still so early it’s kind of a lot of the focus is still around kind of building the products themselves rather than kind of how can we plug these Lego blocks together and how do we kind of build new and different things that we kind of never have thought about. So I’m super excited to kind of see what this happens and kind of what innovations come out of it.

Jose (51:12):

Yeah. No, it’s super exciting for us as well. We really appreciate you coming on the podcast, talking through v2. It’s been a real pleasure kind of working with you over the last few months and seeing the advancements and your thinking. And thinking about these private markets. I think people are going to be really excited about v2. I think it’s a meaningfully better experience than v1 and kind of does really push the cross collateral and stuff, does really start pushing this to sort of a six level or a centralized exchange level experience. So I guess before we go, just where can people catch up with you? And if people are interested in learning more about perp, where should they go?

Yenwen (51:50):

Yeah. They can go to our website. So it’s perpDeFi. P-E-R-P DeFi. So there are, [inaudible 00:51:59] is there. So I think most of the development and communication hang up in this core. So welcome to joining us.

Jose (52:10):

Awesome. Thanks very much, Yenwen, and Nick, and John for co-hosting. Yeah. Looking forward to having you on again and looking forward to perp v2.

Nick (52:19):

Thanks for having us.

Yenwen (52:22):

Yeah. Thanks for inviting us. Yep. Have a good one.

Show Notes:

(2:12) – (First Question) – Yenwen and Nick’s background and what brought them to crypto.(5:09) – What is a perpetual future contract / Motivations for Perpetual V2.

(8:41) – Perpetual V1 / vAMM architecture.

(11:19) – Process to start a market in Perp V1 and what parameters are needed before creating a market.

(17:34) – Reasons for the success of Perp V1 and V2 and how intense growth has been experienced from the team’s perspective.

(23:40) – Biggest learnings from Perp V1.

(26:07) – How Perp V2 works.

(32:19) – Yenwen’s thoughts on the design trade-offs and where does Perpetual sit.

(34:50) – Advantage of adopting a Uniswap V3 style model versus a centralized limit order book model.

(40:36) – Yenwen’s thoughts on scaling perp long-term.

(41:48) – Decentralized vs. centralized derivatives

(44:35) – How the decentralized derivatives market will look like in 5–10 years from now.

(47:44) – Where to find Perpetual Protocol.