EP.53 [EN] - Cindy & David: Making Drift a most-trader-friendly on-chain derivatives trading venue / 将Drift建造成最交易者友好的链上衍生品交易所
55:29 Share

Hello and welcome everyone. I'm Mabel Jiang, the host of 51%, a podcast series presented by Multicoin Capital. This show is the exploration of blockchain's rapid development across Asia with a particular focus on the perspectives, communities, and operators based in China. My goal is to bring Eastern perspective to West and Western perspective to East so you can better understand the crypto's unique market structure and how these distinct communities think and operate.

This podcast will feature a mix of English and Chinese discussions. The language you're hearing now will be the language I use for the rest of the podcast. Stay up to date with my latest episode by subscribing to this podcast. Thank you for listening.

Mabel Jang is a principal at Multiquin Capital. All opinions expressed by Mabel or other podcast guests are solely their opinion and do not represent the opinions of Multiquin Capital in any way. This podcast is for informational purposes only and should not be construed as an inducement to make an investment or relied upon as investment advice. Multiquin Capital may at times hold positions in some of the tokens or companies discussed on this show.

Hello everyone, welcome back to 51%. I'm your host, Mabel Jang. Today we are extremely excited to welcome Cindy and David from Drip Protocol to join us. Let's say hi to Cindy and David. Hey guys. Hey, thanks for having us, Mabel. Absolutely. Originally we were talking that we should do this in Mandarin, however, due to someone's limited capability of Mandarin, obviously this person is not Cindy, we have to do this in English.

No comment there. All right, before we jump into Drift Protocol, maybe you guys can introduce yourself and obviously the team, your background, why you started in crypto. Let's start with Cindy and then David. Yeah, sounds good. So I joined crypto in late 2017 towards the end of the last cycle.

I was an exchange student in Korea at the time and discovered that there were some price differences between Bitcoin in the US and Korea. So I started doing the whole ARP trade back and forth manually on my student accounts and kind of just repeated that trade over and over again. And that opened my eyes to, I guess, all the inefficiencies in crypto.

And eventually, a couple months later, I dropped out of school and just started trading in crypto full time. So I started exploring the first BitMEX perps when it came out and then started trading some crypto options as well.

A year or two later, DeFi came out, got pretty deep in the DeFi world, and then started thinking about, as a PERP trader, why isn't there a decentralized version of PERPs on-chain yet? So then when Solana came by, I had the idea to build a decentralized PERP exchange. And along with David and other co-founders, Chris and Zee started coming up with the idea for Drift.

I mean, it sounds like you took most of my thunder, but for me, slightly different background. I'm a lawyer by training, but I ended up in venture. So I was covering venture at a Singaporean VC and told to look at sort of fintech and crypto was the tangent to that. It was at a time where it was just taking off and I was just tasked with just doing research tasks on that front.

And what ended up being a one month research project just became this rabbit hole that I went down, never to return, met Cindy towards the tail end of 2017. And the two of us have just been jamming on a bunch of trading, a small hedge fund we raised in 2020.

and then started looking more closely at DeFi. For me personally, I tried to market make on Kyber Network back in the early days of 2018. There were obviously challenges around fees, around especially gas and just some of the design back then. So a lot of that didn't quite work out at the time. So when Solana and all these L2s started emerging that reduced that barrier, it suddenly presented

this amazing opportunity to finally build something that was performant. And I think just an internal litmus test that the two of us always have is we always want to build something that we'd also train ourselves. And Drift is, I guess, the culmination of what that looks like.

Cindy just mentioned something very interesting. You said you're wondering why there weren't any on-chain version of PERPs. I guess at that point, it was probably not very clear to a lot of people why do we need a on-chain DEX for perpetual trading or any type of derivatives trading. But based on what you have felt and have been thinking lately,

Why do you think on-chain derivatives is better in some sense compared to some of the centralized trading venues?

Yeah, for sure. I think a lot of the thinking came after exploring DeFi and all the lending protocols, all the AMMs, and thinking that there is an obvious benefit to trading on a decentralized context because, first of all, everything is permissionless. All of the markets that you see on Uniswap, anyone can list a market, right?

And whereas compared to the listing process for essential as exchange, you needed to get approval from multiple different parties, hire a market maker, pay them a fee to make the market. And we just thought that the designated market maker, which actually benefits traders rather than just extracts value from the platform, was really appealed to us as traders.

I think that was one thing, permissionlessness of DeFi. The second was really transparency of prices.

I think on a lot of centralized exchanges, some user data is intentionally concealed from public. And we know that in a decentralized context, anyone can see anybody else's trades on chain. And we think that this information should really be open to anyone to trade on, to use, instead of being traded against in a closed context.

which we've seen evidence of happening in a centralized context. So I think it's quite ironic that given how much perps have blown up the most liquid crypto derivatives empirically, that they're mostly still traded in a centralized context. So yeah, I think those two, transparency, permissionlessness, were really what appealed to us as traders, and we wanted to bring that into the world. That's amazing.

So on top of what you just said, I guess what exactly or what else inspired you to build Drip Protocol to the world? Either David or Cindy, I don't know which one you want to take this.

I think an interesting experiment that I've personally been curious about is having decentralized organizations try and govern elements of an exchange. So to have users, or rather owners actually, be given the token ownership. Owners also be decision makers in the protocol has always been very curious. I feel like

especially in traditional or even just Web2 sense, there's always the saying that customer knows best. And I think that's an interesting proposal around could the customer or group of customers and owners be the best decision makers for the ultimate direction of protocol? So I feel like

That removes sort of an intermediary layer of customer typical complaints to a team who then reports it up and then changes get made. What if that owner could just make that decision with other owners and decide what the best decision was for the entire community? So that's an interesting experiment that I've always just been thinking about, whether that was possible. And I think as we are now, this is probably the best way

gathering ground for this to happen. So that's sort of maybe the non-product take, but an interesting piece on sort of a new type of organization that could emerge. Anything to add, Cindy? Yeah, and I think compared to kind of the other decentralized protocols out there, we hadn't really seen something that was truly fully

fully on-chain, everything from being non-custodial to also having orders settle on-chain, to also having exchange fees go to token holders and, like David said, to allow token holders to really own and govern the platform. So yeah, those were the couple of things that we wanted to see in a decentralized exchange. And yeah, I think that's what inspired us to go in this direction. Right. I guess another point you guys talked about in the past is also building a derivative stacks with

all the features that people used to from a centralized exchange. I think that's something both from the UI/UX level as well as in how to improve the capital efficiency level, which is I thought was very interesting.

We can talk about that, but before going to another direction, I'd love to maybe dive deeper into how the, I don't know whether you guys call it a V1 or just the dynamic virtual AMM, how it works and then how it differentiate from some of the other perpetual swaps on Chang.

Yeah, for sure. So I can give a brief intro about the DAM, the Dynamic Virtual AMM, and also a new feature that we just shipped alongside that called the Decentralized Limit Order Book, which we like to liken to Blob. It doesn't sound that great, but you know, it's a hybrid. We're working on it. We're working on it.

Literally, this is my first time hearing this name. I didn't know about this before this. This is insider alpha for listeners of this podcast. Yeah, at the time that you're listening, we just launched the Centralized Summit order book. So yeah, this is news. Awesome.

But yeah, so the dynamic virtual EMM is sort of a new invention that our team came up with. We like to think of it as marrying perpetual protocols, V1, VAMM with Curves, CryptoPools paper, so Curve V2.

by essentially, it's a virtual AMM, but the main difference is it uses fees generated from the protocol to adjust liquidity and adjust depth of the pool based on trading demand.

So in the classical VMM, you sort of initialize a price, initialize the base asset and the quote asset reserves. And that's sort of the slippage that traders are going to be stuck with as the trading pool progresses. And you initialize it at a certain price. So when price moves too far away from that initial price, you sort of

end up trading at the far end of the curve, where slippage is much higher.

So we sought out to figure out the solutions for these two issues that were in the classical VMM by introducing allowing fees generated by taker fees to essentially adjust liquidity as well as to adjust the peg of the curve back to the deepest part of the curve so that we can keep slippage low. And implicitly, traders are earning from being in positions.

So this, coupled with our latest launch of limit orders, actually allows traders to have all the benefits of trading against a virtual AMM that is backed by the protocol fees and also have limit orders against that. So we think that this is a pretty unique convention because no one's ever combined guaranteed virtual liquidity with the specificity of limit orders before.

So going back to kind of the, you know, we are building a DEX with all the features that a centralized exchange has. I guess we have all the features that a centralized exchange has right now. But the way that we've constructed it is in a purely crypto native novel way that we think is the best model to execute on an on-chain environment.

Because I think, you know, kind of copying the UX directly or the underlying infrastructure from just a centralized exchange doesn't necessarily work in a decentralized context. David, do you have anything to add? Otherwise, I have a few other questions to follow up. No, I think Cindy nailed it pretty concisely. So nothing further on my end.

Right. I'm curious for different markets that you guys support, how do you think about what market better trades with the limited order book version versus the original version you guys deliver with the DAM? Yeah, that's a great question. And I think one way to look at it is to compare our volumes against an on-chain order book.

So if you look at, let's say, Mango and Solana, a lot of their volume goes into the big caps. So BDC, ETH, and to some extent, SOL. On Drift, the composition is very different. We actually have most of our volume on mid caps and altcoins.

So our biggest markets right now are Sol, Luna, Atom, AVAX, even Algo. But it's a lot of the, I guess, like mid cap, more medium tail assets that I think the DAM actually really appeals to. Because on most of these, we actually can provide a tighter spread and better prices for mid cap coins compared to, let's say,

an on-chain order book where the spreads might be wider for these mid-cat vaults. I think you're comparing yourself against centralized exchanges, not the other on-chain liquidity. Because I guess when people choose between where they trade, obviously for some of the Ethereum, I guess for Ethereum, some people might even go to Uniswap V3, but I guess for some of the other L1s, they might choose between either on-chain liquidity versus centralized exchanges.

Yep, exactly. Yeah. So I think, yeah, even I think you see the same phenomena on centralized exchanges where a lot of the volume really consolidates to the major coins, the top 10, top five. But I think on an AMM or a DAM like ours, you would see, like you said, more of a distribution like Uniswap, where there are volume in mid caps and large caps, but there's also, I think, a large consolidated amount of volume in long tail assets.

So does this indicate that for order book that you guys have, the larger pairs can be better support in that sense or more people will come straight there?

Yeah, I think there's definitely a component to which the order book will allow more liquidity on larger caps and as a result, more trading volume. But we actually think because of the way that we've constructed it in terms of a flat order book that operates very similar to how liquidators operate, which is really off-chain keepers take the order matching logic

off the chain to maximize compute units on chain and to maximize efficiency, and then just settle orders on chain when the price of the dam hits the limit prices of the users. So in a way, we don't think that it will

change the construction of the dam because it's working with the dam. So we think that the user behavior will still be quite similar, but we'll probably see more maker type activity start popping up on these markets. So that should be really interesting. Got it.

The other question I had, that's kind of relevant to what I just asked. You mentioned you support other L1s like AVAX or maybe even L-Grand. How do you think about your target audience? Because one of the concerns that I had was people who would trade Polkadot, who would trade, I don't know, like Avalanche or whatever, they're not going to trade on top of Seoul. So

So how do you think about your go-to market for these alternative pairs? Yeah, I think at the end of the day, we kind of think that the traders are not necessarily the users of the underlying pairs. And I would say, yeah, the trader audience is probably more chain agnostic than the maxis, like the Polkadot maxis or the Matic maxis, right? Right.

But we've actually been talking a lot to the founders of these protocols as well and the BD teams to kind of see what we could do together. So maybe I can let David dive into some of the cross ecosystem strategies there.

I think cross ecosystem is something that one I guess has to strike a fine balance on and I mean this both in terms of just engineering Lyft from just a resourcing perspective but also looking I think this is where going second is wonderful where you can see what this sort of multi-chain integrations look like I mean if we take sushi as an example even there is a lot of ways in which you can sort of bridge and use sushi but not

Not all of that has necessarily taken off. So I think when we do that, we want to be very careful with what that positioning is, because if you look at Polkadot, you look at Avalanche, you look at some of these other chains, more often than not, you're going to have a localized version of, you know, Perp Dex or a lending platform and

at times there's going to be potentially a loyal user base to that particular chain. So rather than I think Drift saying, "Hey, we want to try and be the be-all for everyone to be a part of," we probably want to be quite selective what that looks like. I think we've tested the waters with, for instance, Matic, running some campaigns with them. I think we're in conversation with some other teams as well, which I'll save for now.

But moving forward, I think that is the direction we want to head towards. It's really thinking about, I think the real secret sauce is, can we apply this dam logic to not just Solana, but potentially other chains as well in environments that support what we have in mind? So that's sort of how we think about some of the cross ecosystem pollination.

Yeah, I guess like what Cindy said about the traders are not necessarily the user of certain chain or certain smart contract platform. I kind of echo that as in when you try to market to a group, I think that just at the end, it's,

I'm not saying you should do that, but I feel like generally a lot of these on-chain venues right now, they're just more thinking about how do I target the existing users of these chains? Alternatively, maybe thinking about how to target the general trader audiences and then think about how to deliver a best UX to these people, regardless of what chain it is on. I think that's probably the best strategy. Not sure how you guys think about this.

I guess it goes to the longer term thesis of having the dynamic AMM being more conducive to supporting those longer term markets. At the end of the day, these markets are where a lot of the retail interest is going to gather. And if you provide that in a non-custodial manner, where they can also interact with

good UX, i.e. the fees are quite minimal, then you're going to see a fairly seamless user acquisition versus bridges, for instance, or perhaps even high costs. I think a mantra that I always live by is there's nothing that kills UX than fees. So I think by removing that and those barriers, some of which can be structural or systematic, that's a very powerful acquisition tool. So

Amongst, since you guys have been up for quite a few months by now, I was wondering what are some of the interesting user behaviors? Okay, let me be a bit more specific. For example, for people who open a long, short position,

What's the average numbers of leverage they usually take? I'm just curious about that because I was wondering if the user behavior would be a little bit more conservative because they're on Chang versus when people are on centralized exchange, they feel stronger to be YOLOing.

Yeah, that's a great question. We just kind of started analyzing a lot of the user behavior on our exchange. So right now, the average leverage on the platform is around 2.5x. So our max leverage, I think, is 5x right now. And we're going to be increasing this to 10x pretty soon. But yeah, the average trader is not actually that levered.

So if you look at the distribution of leverage across different types of users, you can find that a lot of the more high frequency traders tend to use lower leverage. So even something like less than one and most of these smaller retail traders tend to use more leverage. So that's a pretty interesting comparison stat of who's trading on our platform and how much leverage they're using.

So does that mean like people who are trading on top right now, it's mostly the invisible whale, so to speak? So I think 98% of our users, we would categorize under the retail UI trader bucket. So we had about, so right now we've just hit 12K traders. Also thanks to a recent campaign we did with Project Galaxy. So shout out to them there.

And across the last month, I think we had about 4,000, 5,000 active users. We found that 98% of those were regular retail traders. And we kind of categorize them in buckets based on how much they trade and how often they trade to detect non-bot type activity.

So that's quite promising to us to hear that it's quite a big number of retail traders. But then at this stage, majority of the volume is still done by the arbitrageurs and the market makers, which actually mimics, I think, the distribution on a regular exchange as well. And also on AMMs, I think a lot of the volume is distributed.

volume that that arpes prices back to to oracle prices and that's also driven by retail activity that pushes prices in the in the opposite direction so yeah i think um there's definitely an element of uh different types of user behaviors here so what were some of the milestones that you guys felt great to share over the past few months

Yeah, it's been a crazy couple of months. We just launched on Mainnet, I think in end of November, started December. And we just hit $4 billion in cumulative volume. We hit, I think, 25 million open interest. I think we mentioned 12K users. Just listed over 12 markets now as well. So I think those are some of the key stats.

How about maybe integration with some of the other front ends? Because I figured that was also one of your major initiatives in order to get more people interacting with the protocol. So like integrating with other protocols to host us on the front end, right? Yes. So like not interacting directly from your own front end, but like someone else's.

Yeah, we haven't gone down this route yet, but it's definitely one that we're interested to explore, especially around decentralization. So David can talk a bit more about this, but yeah, definitely want to start decentralizing who runs the front end, since that's the main part of our protocol that's still relatively centralized. Yeah, just to quick add to that,

primarily the volume on our platform is just organic at this stage. I know that from our

i guess past trading life we were fortunate to know a lot of these people who are also uh big players on our platform as well so that's definitely assisted but beyond that the drift doesn't have a token we don't have any type of reward mechanism right now so it is really positive to see that a lot of this can be organically generated so i think from here on there's obviously quite a few hot and quite a few verticals we can go under i think as you mentioned the ui

to have multiple people host that is obviously great to attract a wider catchment area. I think there's also a legal lens to look at that, which I can go into in a bit without, you know, sort of going too tangential of a conversation. And then, yeah,

Beyond that, there's discussions of maybe funding rate vaults that we can run. Even another multi-comp portfolio company, UXD, thinking about could they hedge on drift with SolPurps potentially. Obviously, this does require cross collateral, and that's, I guess, the next thing on our roadmap after limit orders. But there is so many different avenues for us to, I guess, compose and go through.

around increasing both the collateral vault as well as the number of just open positions on the platform. So I think there's quite a few verticals for what this could look like. I feel like having these little orders just enables so much more. So I think at this stage, a lot of it has been internal, but over the next few months, you'll definitely see a lot more partnerships being announced. Got it.

Out of curiosity, what's the kind of distribution for Solana and Solana native coins or tokens versus some of the other non-Solana based L1s or some other tokens? What does the volume distribution look like? Volume distribution and open interest, I should say, if you have that on top of your head.

Yep. So we actually have not listed any Solana native tokens yet, which is going to be our kind of next expansion plan once we ship out a feature that allows different leverage on different markets. That's really what's blocking us from shipping like truly D-gen markets or like anything from, yeah, something, some fair launch project that just launched today or some NFT floor price index. But yeah,

We put risk management kind of first on our priority list. So I want to ship out different leverage before getting to those really, really alt markets. And so right now we don't actually have any Solana alts on our platform, but definitely looking into it. And we'll be adding that on once we ship that feature.

Right, but generally just like Solana versus some of the other ones. I forgot where I saw a number. Maybe it's like 60% of those volumes were for Solana versus like the 40% or 30% was like the other odds out one. Oh yeah, for Solana directly, it's around 60% on Solana.

Got it. Interesting. So it's actually still quite significant volume. So that's like what, what surprised me quite a bit. Cause like when David previously brought up that you guys are, are, are listening from the other L1, I was like, why would there be someone like trading these L1s on Solana? I guess like I was wrong. So definitely interesting to observe that. Yeah, for sure.

Sorry, are you going to say something? I think it also timed really well with the whole L1 boom. Everyone wanted to, long or short, different types of L1s and had a lot of opinions on that. So yeah, I think it definitely timed well with that. Got it. That's interesting. Yeah, I guess for a lot of these perpetual swaps, it's always about listing something timely and something on the topic, and then you'll get a lot of volumes. That's for sure.

Yeah, it really was, I guess, the flavor of the week. It was sort of the sentiment of a lot of conversation on the likes of Twitter and even just some of the folks who we were chatting with. And I think...

One thing that I think maybe comes quite second nature to us just having been in the space for a while is that you do get a sense of, oh, the topic of the week is this, or this is AVAX week, Luna. And then, oh, next week we might move to NFTs. So I think a lot of the markets that have done the best for us are the ones that we have deliberately listed at a certain time. And I think combining that with how easy it is to initialize market where we don't need market makers, for instance, that's just made things so much seamless to react quickly.

For sure. So other than, since you guys talked about yourself being a derivatives trading venue, other than what you've done with DAM and also the limited order book that you have right now, any other products down the pipeline that you're interested in building?

Yeah, I think, you know, while we started with more linear Delta one products for purpose, just sell the biggest market in crypto today, crypto derivatives today.

Definitely interested in going down the pathway of looking at other types of derivatives too. So I think derivatives that offer convexity is on our list too. I think anything that fits within the perpetual framework of solidifying all of the liquidity into one instrument really appeals to and can use the infrastructure of what we've built.

So yeah, I think things like power perps really appeal to us. And from there, there's so many directions we can go into. Square root perps, log perps,

Yeah, things like that, that take advantage of the infrastructure. And then down the line as well, there's also more intangible things that could be traded via something like us. Things like, I think, yeah, events or anything that really has a continuous data structure can be traded. So that's obviously going into the more abstract front. But yeah, that's possible if there's demand for it.

I mean, at the core of it, it's always just been if there's an oracle and Mark can initialize. I think that what makes it sort of fun and accessible to do things like NFT perps, which is sort of a hot topic, maybe some four months ago, but has since dropped off. But that as a derivative is always, you know, it's still an exciting concept. It's something I've always found interesting. I think there are some sort of natural challenges around something like that, given that in theory, spot should always trade at a premium, just given that if you hold the underlying, it

entitles you to the community on the benefits future airdrops which i guess you'd be priced in but for some of these call them like bluer chip or harder to access type of assets like a bayc uh people might find

it more accessible to trade a derivative of that. So essentially having a portion of an ownership and have a view on which position it might go. And for people who might hold the asset also to use perps as a hedging tool. But again, the mechanics of things like NFT perps then sort of wanders into this murky territory when you start thinking about, oh, how do I collateralize my NFT? Is this purely a cash level instrument?

But the rabbit hole is so deep, but also just wonderfully fun to explore. Yeah, you definitely read my mind. I was going to ask this question about NFT terms. I was debating if I should, because that could actually lead to a lot of other discussions. Because I've been thinking about if we could use the infrastructure that NFTX provides for some of the large liquidity projects.

some of the, I guess, like large PFP that have deep liquidity, I think actually it's possible to deliver a per product. However, that also it's relatively rough because for NFTX, basically what it assumes is that like, you know, you have one category, it's like a D1 token representing a bunch of like different and different punks, for example, or different bake. Then essentially all

all of those are actually not having the same price but all but because they have they are fit in the same category of different features like they are valued at the same price so there's like still a lot of questions around that but i guess like generally speaking for the nfts that have like a lot of liquidity you can try to do that but obviously yeah like your point how do you

um like for some people obviously it's it's not the best because um for i guess for purpose fine but then like for nft eggs like if you deposit a you know more slightly more expensive nft from the same category and then someone else like you know swap out your your punk you're not going to be happy so i guess like there are a lot of questions around that

Yeah, I guess the analogue there is that a lot of the naming in for this, and I know I'm going down to a very particular nuance, is we call it floor pervs for that very reason, is that it's very hard to assess collateral for, say, like a rarer punk or like an alien punk, for instance, or a gold BAYC, which is worth significantly more. So to hold a value that at the same as a floor is obviously very disadvantageous. And you'd obviously not want to...

you know put that up as collateral for instance so that's obviously a huge challenge so i guess when you think about nft perps there's probably two parts to it there's i guess one the collateralization piece and then there's i guess the exposure i think if you abstract away the collateralization which sort of i think dave white talks about this uh in and proposes some interesting concepts in one of his papers one of his research papers but if you check the way yeah so i think he

into like ways you could deal with the rarity issue because some of these rarity, uh, the, the variety ranges, the variance can be quite extreme. So you definitely want to be cognizant of that. But if you abstract that away and focus on just the, I guess, price exposure of, uh,

the PURP itself, so the NFT PURP itself, that could be quite interesting just to track the floor price. You have an index, so the NFTX, we're still waiting for something like that on the likes of Solana, for instance, but on Ethereum that already exists. NFTX, I think there's one other platform that escapes my mind at this moment.

But definitely possible to put implementation for this. I think you definitely have to think about people gaming the index, for example. You need to have a certain level of confidence. People can't just list like a floor SMB, for instance, and then

potentially spoof the index price. So there's some NFT specific guardrails you might have to put in place. But I think you do make a good point. A lot of this does come down to liquidity. If it's just too gappy, then you might get some stale pricing. So that's also something just to keep an eye out for. And also I think it's getting a bit

clearer now who I guess the market leaders on Solana NFT is unfortunately, you know, a mutual friend of ours. But prior to this, you know, liquidity was very fragmented across many, many platforms and it made it very hard to pull prices

together because there'd be huge discrepancies across all these different platforms who are all clawing for their piece of the market share. So it's definitely a lot easier right now. But prior to that, where there were perhaps, you know, five, six players, it was a far trickier exercise to try and say aggregate an accurate index for a certain NFT asset.

I mean, I think that's not a problem longer term anyway, because I think regardless on, I think generally speaking on Chang, there will be like once someone becomes a relayer, everybody else will become a relayer. So it's like everybody will try to try to list someone else's liquidity.

That's basically what Fractal has been doing with Magic Eden and vice versa and some of the other ones too. I think that's actually not a concerning point in terms of liquidity fragmentation. I think as of today, the way I read, I know this is a bit off topic, but I think for extremely rare NFTs, it should be just P2P.

I think that's probably the best for the holder versus I think for the ones that are generally more common in terms of features. Like that could basically apply, you know, the logic as in like you can fit the categorization framework and then try to evaluate that based on, you know, different features of that thing and then provide a different types of perps. I think in terms of floor perps, it's,

It's not going to work that well for the NFT collections that have a large different gaps between, you know, the lowest, the lowest, basically the four prize versus like the highest. Like if that's the case, I think you'll have to have like different types of categorization for that.

But yeah, this is an interesting discussion. We can definitely talk about it more offline. I do want to get back to what Dendi brought up earlier in terms of the, you know, some, like you want to offer all the products around like perpetual swaps. So does that mean like you think options is something that you don't think would work on Chang? I think there's an interesting discussion for this discussion.

I think options as they exist today with a bunch of different strikes and a bunch of different expiries,

are kind of the opposite of the reasons that DeFi flourished because of simple UX and consolidated liquidity. So we think that with especially people like the open team really pushing the boundaries of what options really means and seeing the invention of these DeFi options vaults that are obviously very early in their design.

But what they essentially do is really consolidate all the liquidity for options under one blanket.

And I think prior to these, I guess, innovations that allow DeFi options to really thrive, we didn't see too much activity on options order books. And I think the real reason is that DeFi users in an on-chain environment care more about execution compared to specificity.

And while I think options as they exist today really thrive in how specific they can be, DeFi options really won out when it became simple for users to have these types of exposures, whether it's long or short in option. So I think for options to really thrive on chain, it would have to come down to

a very simple UX that kind of just provides users with exposure to different options Greeks in different forms compared to the structure that we see in an off-chain context today. So I actually think they will likely take off, but perhaps not in the form that we see today.

I'm struggling with that, actually, because I've been thinking about a lot of these fixed rate products and then options and stuff. I think ultimately, it's a question about whether DeFi is something with an ethos of anti-liquidity fragmentation. If we think that's the underlying principle, then...

It's possible that only the large whales or some of the institutional traders, they would use some product like options or fixed rate. And then for most of the people, they don't care. They would just use PERP or anything that does not lead to a liquid fragmentation. That was like what I've been thinking lately. So that's why I was asking you the question.

Yeah, I think that's a really good thesis. Largely agree. I think most of these products are still dominated by retail. It's in the on-chain products. So yeah, definitely a good point. Have you guys been offering cross-collateralization or is it something on the pipeline?

Yep, it's on the pipeline. It's going to be our next big ship after limit orders. So yeah, definitely look out for cross collateral because I think that will help unlock a lot of the latent capital in the insulina and also wider option ecosystem too. For sure.

And David, you chatted about NFT pervs, like we just did. Anything else? I guess like any other asset class that you've been thinking about, alternative asset class, like the non-fungible token stuff or non-native token stuff?

Non-native tokens, I guess. I mean, the big can of worms is obviously tokenized securities. I mean, could you offer commodities like other indexes through a DAM? I think there is a possibility. I mean, this is sort of a big legal puddle we sort of jump into. But a big thinking, I think that the thesis for...

for this dam is removing the market maker, removing the middleman. So I think internally we term this market maker extractable value where someone is clipping a spread and taking it outside the system. So could you apply that to traditional markets? I guess, I mean, personally, I have a thesis that market makers make too much money out there. People may or may not disagree with that argument, but by abstracting that, you're really rewarding the users. And I mean, can that be applied to traditional markets

Absolutely. So that's, I guess, maybe a broader goal. And I think maybe projecting two to three years on the track for us is thinking more about how do we make that real world bridge? I think a lot of these DeFi protocols that we see, interact with, know of, all play within maybe the same 4,000, 5,000 customers, maybe a little bit more nowadays. But

in terms of just reach outside of a strictly DeFi native ecosystem, very few, if any, projects have made that jump. I think that is a big milestone for us to be able to go beyond and interact outside of just the crypto landscape. So that's something we've been thinking about. But again, I feel like as we are now, there are so many other product priorities to attend to before that.

Yeah, definitely longer shot, but worthwhile if you are at a much larger size at that point. What have been some of the biggest challenges for you guys running a decentralized exchange?

I think from a product standpoint, obviously, from I think he knows any exchange operator, but the scariest part is always during large drawdowns. So I think January 24th was probably the biggest test for our platform so far.

It was really like a war room scenario where we were making sure that all the trades were going through fine. The Oracle was actually stuck. So we had thankfully built a lot of guardrails around Oracle failures during type of network congestion.

so that users don't get wrongfully liquidated. But yeah, it was quite a stressful situation because I think, yeah, Oracle is stuck. We're figuring out how to get them back live. So yeah, I think that was probably, and these type of market scenarios, probably the most challenging parts of running it at DEX.

And yeah, I think, I mean, also the funnest, I think we ended up coming out of it in a pretty good spot with a lot of improvements that we wanted to do to the matching engine and everything. But yeah, I think these are always like the hardest and biggest tests of the platform. So yeah.

Yeah, product definitely kind of speaks to Lindy effect, like these sort of battle tests. These little like stressful markets are great for battle testing things. I feel like that was a big challenge for us. But I think when we stepped through quite well as a team, so I think honestly, not a bad thing that happened for us. Maybe can't say the same for a lot of other people out there. So I don't mean to gloat by any means. On the other side of the decentralized exchange, it's thinking about the sort of the regulatory framework. It's been quite a

big discussion of the last year and a half in particular around how these platforms are run, the stance on tokens of these platforms as well. I think just even just interesting things to know for things like the DYDX token, so platform on L2 right now, or Starport, sorry.

There's been a lot of thinking around how they position the token. I think in their documents, they explicitly state that our token will never have expectation of profits, which is sort of one way to place you outside of the Howey test, which is the test for an investment contract. I feel like the other way, which we talked about earlier in the conversation, was around decentralizing front ends. And that sort of speaks to the fourth limb of Howey, which talks about

getting derived from the effort of others. So if a sufficient amount of other people are also running front ends for you, then does that then remove the whole sense

centralization of are you just deriving your profits or everything from the drift front end instead of no 10 people so when that when that owner shifts then you sort of fall within a more decentralized categorization so i think for us it's always an ongoing balance for for that and just trying to strive to hit those goals sooner where possible i guess the trade-off is always with decentralization it is costly and sometimes things can get slowed down so

I guess as a core team or like sort of the initial core contributors, it's a very careful decision you have to make around how quickly or when to decentralize. But that's, I guess, a pretty active and dynamic conversation that we have internally. So I'll pause here as I'm sort of hesitant to go too deep down the rabbit hole and let myself get lost in that. So I'll pause for now.

Lawyer by training. To continue on that, though, because you mentioned compliance and regulatory challenges and whatnot. Obviously, different regions have different types of regulations. I guess my question is around how do you think about engaging different

like different users from all over the space in the world? I guess I'm sure the marketing and the operations against those people, like different people in different places are going to be very different. Great question. I think as a baseline, few regulators have taken as strong a stance as the US right now. So I think broad brushstrokes wise, our customers is everyone but the US. In terms of

I guess, approaching these users. This is where I guess thinking starts to divulge a little bit. I think when I first came into this space, I had a hypothesis that regions, you had to hyper-localize some regions. So the UI, for instance, for a region like Korea, where up

up instead of the traditional positive is green, negative is red. I believe it's flipped around positive is red and going down is blue. So there was sort of a thesis that, hey, do we need to hyperlocalize for everyone? But that hasn't proven out to be the case just yet. And I'm not sure if it's a function of us having not sort of gone to these regions or perhaps understanding maybe some of the cultural specificities from these regions around trusting these new platforms or having collateral in their native currency.

but that's something which I think right now we need to just double down on just ease of product right now just ease of like using it I'll let Cindy talk to the product nuances here but at the end of the day for us it's creating the same experience you'd find on chain but in a non-considered manner so that I guess underpins a lot of our thinking around approaching users so I think

Despite coming into the space with very much, oh, we need to hyper-localize for all these different regions, I feel like we might just double down on a particular product or a particular feature initially. But I'll pause here and let Cindy comment on the specifics there.

Yeah, I think the approach of going from top down, building that exchange first and then figuring out like our growth strategies and also different types of adaptation strategies to different markets is really...

where I think we'll win. I think global campaigns also given like, yes, there are regional areas to focus on, but at the end of the day, our audience is really all of crypto, all of DeFi, which can be all over the world. So yeah, I think we haven't really, I think besides obviously clear regulatory issues,

areas to avoid, really focus on super hyper localization because we want to focus on the bigger picture kind of product at first. So yeah, that's been our strategy, go from like top down and then really focus on kind of building local communities after the main product has been like at the final stage.

Yeah, for sure. I think for DeFi specifically, protocol and front-end separation has been the trend. I always felt like

I think there was a conversation between myself and the founder of audio. Um, Renil was saying that, um, they, as the infrastructure can do, it can make it permissionless, but it's up to the front end, the local front end to decide like what, you know, what they offer to their local users to be compliant and whatever. And they are fine with, um,

whatever those front ends are doing or UI were doing. I thought that was interesting trend that definitely separates a lot of the Web3 product versus the Web2 ones. Yeah, no, I think this is a great conversation. Any last minute comment about, I guess, your long-term view, anything that you feel like you haven't covered, you wanted to share with our audience before we finish?

I think it's a great conversation. Make sure you trade on Drift, let us know how it goes, join our Discord and all that jazz. Often maybe like a cryptic farewell, I suppose, but something that we've been thinking about internally is governance mining. So that as a parallel to liquidity mining, how do we...

supercharge, just getting people involved. So that's something just to, I'll leave out there as an open sort of question. And I guess I'll just leave it there. I think ultimately that is a direction we want to go in. So please check it out platform, use it, participate and above all, give us feedback. I think one thing we do pride ourselves on is just responding to most queries within the hour. So please test us on that. So yeah, that's all for me.

Awesome. Thank you so much, Cindy and David. It was a pleasure chatting with you. Thanks. It was fun. Thank you so much, Mabel. Next time we'll do this in Mandarin. David. Sure, sure. Alright, thanks guys. Thank you. Bye-bye. Bye.