DeFi Market Structure and Mechanics - TOKEN2049 Singapore 2024

my name is uh P Duran I'm with Deli digital on the research side so primarily cover AI infrastructure um but today you know we're diving into defi with a few experts here um so yeah maybe I'll just hand it over to the panelist Tian we we'll start with you for just quick intros um you know kind of your background in defi um and then we'll dive into the question list how does that sound sure so um hi everyone I'm Tien CEO co-founder of Pendle um we're a y organization protocol I'm Lewis founder of oror capital um we run a liquid token hedge fund uh mostly prop Capital but started raising external Capital this year focusing on defi yield hello guys uh my name is julan Buu I uh founder and CEO of state capital uh it's a investment family office we also do uh quantitative strategies and we also have a venture Studios where we built a different type of Def projects okay um so yeah this is the uh defi Market structure panel um and and so maybe just first question is you know over the last 12 months what have been some of the largest changes in def5 Market structure and then you know kind of your forecast for what the next 12 months months might bring um so just kind of the biggest updates in terms of what's happened in Defi and how things might change going forward I can start um so I think one of the key changes that we've witnessed is the evolution of incentives obviously you know def5 Protocols are not profitable from the get-go have no Revenue no users and I think what we've seen in 2021 was a lot of these um bootstrapping effectivity and users came from rewards High aprs tokens but projects started to realize that's obviously extremely cancerous um you know tokens get dumped immediately mercenary Capital come and that game sort of evolved into points this year um because obviously they they want to cast additional layer of I would say opaqueness to the rewards that users are receiving and hence less mercenary Capital coming in yeah um and I think we'll continue to see an evolution of that as you know projects start to continue to Tinker about you know how can they continue bootstrap users using incentives but at the same time optimize for more genuine activity genuine or captive Capital rather than the mercenary ones okay yeah I guess like from protocol development perspective I don't think a lot has changed so compared to maybe like 2020 2021 period most most of the protocols that users utilize are still largely the same um yeah I I agree with um uh what oroboros mentioned right largely around the incentive structure and I also am of the opinion that because now it's so easy to launch l1s and l2s and everyone's really fighting for a small pool of applications um the only way is to really just utilize incentives to attract development and it becomes like it it it kind of results in a very um yeah short-term kind of uh implementation right like as a developer there are so many options you just have to go to the chain that offers the best Grant and then you become like a Serial um Grant applic mercen yeah yeah yeah yeah and and you just keep rotating that and again I'm not sure how things will change in 12 months because I think just because the barrier of launching an L1 and L2 ecosystem is so low right now from a technical standpoint um I I I foresee like this Behavior will likely sustain yeah and then I I guess a follow-up question is just around Moes right in the event that some of these incentives are being used poorly or you know there's just an endless bunch of new projects um I would say in web 2 it seems like there are barriers to entry there's Network effects in web 3 you know you have vampire attacks things are forkable um there's just a much lower barriers to entry for new projects to come after your market share so how do you think about from an investment perspective or a builder investment perspective like one building Moes or two finding protocols that have sustainable fundamentals over time yeah I think it's U it's a matter of the uh Team uh Power executions it's uh I think in in a space of defy you have to be even more competitive than in traditional Finance so don't be scared of uh um like like you said type of vampire attack or uh finding the right model to uh get liquidity from other protocols and I think coming back to the first question uh what we've seen in the past when the market is high you can incent ize people to deposit inside your your protocol like to liquidity treasury incentives but when the market is low then you have to find new uh mechanism uh to attract people to deposit inside your project or to cross and Bridge inside the the your uh um protocol uh blockchain and this is where you can up with points and points is it's kind of an interesting mechanism it's like miles in traditional Finance where you don't really need um money but you need to um uh pretend or make sure that those points will be worth something at some point and this is how you attract users without spending money uh where in um bull market you you can actually incentivize people to deposit because you have the power executions so coming back to the second questions I think how you overpass the liquidity fragmentations and people being able to use your product is by having a strong team and strong power executions yeah it it does feel like given some of these other factors are absent just like continual execution and betting on the right team is one of the few sustainable modes in dii so I I have something to add like from a builder perspective right um especially like when it comes to defi a lot of at least like the parts that are managing money um like the contracts that are that that users interface with those are open source but I think there is also another layer of technical mode that teams can consider uh implementing right um so for example front and back end they don't necessarily have to be open- sourced um I I think the like this is potentially something that can differentiate the user experience um so this is one and then secondly I think uh brand and reputation also matters Unis swab is I think one of the better example in this regard right like there are so many Forks of Unis swap but generally speaking everywhere Unis swap deploys uh on on regard I mean across all the different ecosystems right they continue to be the dominant amm and I think a large part of it is because users have a lot of trust in the Unis swap brand even without attractive incentives yeah the interaction with the protocol is very organic and I think yeah as as as far as I I I'm aware um it's it's really largely because of the trust in the unisoft brand and and then in I guess Pendle obviously you you guys have found product Market fit um and and and have had some of those gains like have you gone about specifically uh implementing certain practices to build up these modes or you know uh variations in your go to market um that has led to some of that success I think um yeah we try to replicate the unisoft model right we try to establish a brand that users can trust and then of course like at the technical layer if things can be delegated to the back end we try to do that um so now Pendle is actually a rather heavy like back and heavy product um we are consciously trying to create the mode and protect it yeah and then of course like on top of that the liquidity is at least in pendle's pendle's case is is another form of mode because um for new products that emerge um firstly the time in Market is not long enough so it's very difficult for these protocols to attract big amount of liquidity we've been around for a couple years now yeah uh we've been through a couple of cycles and then in the most recent maturity event in June we processed $3 billion with no hiccups so I think um over time with with um yeah more of these events like we hopefully can can uh construct the trust in the in the brand yeah makes sense um okay and then I guess around liquidity fragmentation that some of you guys discussed earlier um just thinking about like one of the original Promises of defi was supposed to be around composability you know software developer Network effects money Legos being able to piece things together in kind of a compounding ecosystem but given you know the explosion of layer ones and the low barriers to entry um you know I'd say it has fragmented quite a bit and there's been a relatively poor user experience um as well as frag mented liquidity so how do you guys think about this you know how does this evolve over the next 12 months does salana take more share do based rollups come to fruition are there chain abstraction Solutions you know does more share go to solvers and intent networks yeah yeah I believe we'll see um a strong concentrations of um uh few protocols um I think we'll see um there was a lot of L2 and L1 over the past few years uh but because now you will see um protocol being able to interact across different blockchains without even the user knowing where the liquidity is being deposit I believe uh today it's kind of a halfa but you uh one inch is going to release something that is uh cross chains and also uh Bitcoin and Etc you can swap and deposit liquidity without knowing uh yourself where you're depositing the liquidity and um this will allow uh projects to only focus on a product and then you will see a lot of L1 and L2 dying because they will not have uh the narrative to incentivize people to stay in their protocol because it will not be a matter of staying on the protocol it would just be a matter of uh the the use case and the and the and the EP py or the yield and this will as a product for example pandel will have the ability to use any type of blockchains and just focus on on on a product level without having to strike deal with uh or partnership with blockchain does that mean that the underlying L1 gets more commoditized and the apps capture more value yeah I think we're moving into what we see in the in in the traditional Finance for example HTTP is being the main protoc call for going on the web SMTP for uh messages so I think you will have and potentially you also have and this is what we've seen before is like L1 where all competing with uh etherum and saying oh we are the best um cheapest transactions and and fastest one and now they all uh being turning into L2 and providing um uh functionality to etherum uh so see potentially ethereum salana or other strong blockchains but very few will uh remain and survive I believe yeah leis you have anything there or no I largely I largely agree I think that the experience of a user has massively um been staging but we have indeed seen several middle layers being built to improve that experience um and exactly like what Julian said these days you can swap one asset from one chain to a different asset on another chain we're starting to see these middle layers being built we're starting to see protocols that are aggregating yield um in one asset across different chains so I think these will sort of help improve the user experience for sure yeah makes sense um all right don't want to get too too political here uh but the US is obviously hitting into an election I would say crypto has kind of been a Whipping Boy in certain jurisdictions over the last um four years but hitting into the election we've seen a little bit of a a shift in tune as people have kind of found that uh you know this is a sizable constituency uh for populist politics um so curious to get the panel's thoughts on you know what a trump Victory could mean or a comma Victory could mean for institutional adoption just given you know how important the US market is um in terms of regulatory leadership and does that open the door for institutional adoption so I think like none of us obviously predominantly in the US but speaking to a lot of us Builders it seems like there's a lot of um there's a lot of uncertainty around the regulatory landscape and they're just waiting for an Administration to give that Clarity um Builders advised by the legal council get different inputs week in week out about where they can be where they can't be where they can be domiciled where they can't be offering um certainly an Administration that you know paves the way for clarity would I think open up a lot of innovation in the US um I think also a lot of builders that were initially more apprehensive with then become Bolder um to innovate as well um so for sure I think you know if you have an Administration that's friendly towards um defi Innovation then certainly that would benefit the landscape um yeah for sure it's um I think if Trump um seems to be the the the most uh um like has the most atite for crypto I'm not sure personally if it's I mean for him a way to get more votes uh but it's a good signal for regulations uh regulatory uh institutions wise um and and it's interesting because um us is one of the most um strict towards um defi projects and most of the defi project out there they block us from being uh used but it's also where products are being the most used so I'm I'm I'm I'm confused how this is uh possible so if um us get friendly in crypto because they believe it's um it's a strong uh CR weapon uh in in the world as uh dominations I think yeah will probably get back on a on a bull market uh but um yeah it depends and it was also a good uh um uh uh vehicle for for example py Market prediction Market yeah all different things so in a structure product Market derivative Market of course this will be a very very strong signal yeah I agree um okay next one I guess uh you know I'd say uh the past two years have been an explosion on the AI side and you know it seems like every month there's a new model drop every day there's a new paper drop you read the Leopold Ashen Briner paper you know it makes a credible case for AGI by 2030 right um and whether that's the case or not like you know there are 100,000 h100 clusters being built it seems like you will have very capable agents on chain or inter mediating users interactions and so I am curious how that impacts like the defi landscape who might benefit who's you know potentially disintermediated how how does the intersection play out um in in y'all's opinion I think the um the uh the market of data and and being able to uh sell or aggregate um um and build uh models like large model that we seeing in in uh open Ai entopic and those type of Open Source uh model to um to type into the economy of data like which is the one of the biggest uh Market worldwide um for defy and blockchain in general it's it's a tremendous amount of um it's a huge Market probably potentially one of the biggest one but it's also a lot of noise so as a I'm talking in term of investment uh for us it's it's it's really risky uh because I would say like 99.9% are all type of um vaporware yeah vaporware it's even bigger than just AI because you get crypto plus AI so by all the magnitude it's uh it's it's huge amount of uh vaporware uh projects but 1% or less than 1% will capture trillions of US dollar market cap so it's a matter of understanding U having the right team that can um um tackle this Market or build interesting product because now today everyone can have access to strong U model apis type of things so it's a matter of finding the right thing that will position a product and be able to attract users uh but it's it's all about about the team I would say yeah so go ahead oh yeah I have a point to add on AI I think one of the stronger propositions of AI at least in the case of defi is the enhancement of security I think like far too many times like the attacks have become quite sophisticated now um and I think a lot of the attacks um at least from from experience and and looking at the last 12 months yes they've become more sophisticated but at the same time with artificial intelligence there are probably ways to minimize and mitigate the the risk exposure right like uh far too many times I think um uh an exploiter uh someone with malicious attempt would just replicate a website and then that that looks exactly the same and then try to fish the vulnerable individuals I think yeah if if there are AI um agents that are very dedicated to taking down some of these militia sites I think we can cut down quite a lot of the unnecessary exploits um uh not so sure actually I think it will be uh even faster for people to commit uh crimes but all the magnitude way bigger than being able to protect yourself because the the the amount of people that are willing to or they incentivized to attack others is actually bigger than the one that willing to defend those people being attacked so by Nature you will see more attacks than defense well I think this is probably a case of incentive alignment right yeah but yeah cuz today the amount of hacks that we are having without AI uh big and v news is one of the website that list all those hacks and this I think this will continue and get even bigger and faster amount of hacks but yeah the only way that you can protect yourself is by having personal agents that check but still I think the power of those hacks will increase and the amount will increase yeah I I think the the cyber security you know cat and mouse game just continues right like you know one branch like you know the attackers Speed Ahead The Defenders catch up like it it it really is like quite quite cyclical actually if I can add to that I I think um what's also very interesting the first question that you asked was the evolution of the the landscape and we're talking about points I think if you know we have ai agents optimizing defi yields back in 2021 then obviously they're going to Crush the game because it's very you know onedimensional it's very 2D but um these days I think as you know incentives evolve into points a lot of what we've realized um about extracting returns from defi comes from the more fundamental less empirical work of you know talking to assessing which which teams are going to successfully pull off a listing for example which teams are going to be um well looked upon by investors and I think these are like the sofware things that requires you to talk to exchanges VCS listing teams the teams themselves um these are things that obviously the AI agents can't do and it's obviously not information that's publicly available so I think it's the intersection of these two things come um together it be it'll be quite interesting to watch how you know this this obviously PS out yeah it's obviously evolving very quickly so I think we'll we'll see a lot of changes over the next two years um all right I mean there's there's been a lot of talk uh on the panel today about points you know I think we went from aird drops to 2024 kind of being the year of points it seems like there's starting to be a decent amount of push back um what do you guys think comes next how do you think about go to market you know as a as a protocol founder um or you know what would you look for in a project in order to have a successful go to market strategy in a web 3 environment there's definitely a change in a type of points now points maybe in the first half of the Year there was a lot of uncertainty right because ion layer in a way popularize the notion of points and then have all the lrts building on top of Babylon issuing their points as well and because of the uncertainty there was a lot of inefficiency in um the price discovery of the points and when we as Pendle launch some of these markets there was a hefty demand for points and as a result of the demand that a lot of users were actually paying a premium for acquiring the points uh which also resulted in a very interesting Dynamics where the other side of the users could acquire fixed yield attractive very attractive fixed yield maybe around the range of 40 to 60% this was around the March April period but as time passes there's a lot more data points because ethery Renzo ethena a couple of other projects have gradually materialize the points into actual air drop tokens and and then people have actually uh like a lot more data points that they can work with so I think right now um because of the historical data over the last couple of months the premium for points has come down quite drastically um and and I am of the opinion that the way people perceive points today would be how they perceive the project like in the past six months right maybe in the in the January of this year you can be a really shitty project and you launch a points program there could be a healthy interest for the points but today if you're a shitty project and no one pays attention to you uh the points not going to make a lot of difference but if you're a good project it can have a couple of benefits as well right like it is a very very efficient way of attracting initial liquidity and this is very different from like just test net points right because test net is all fake you can actually bought up the numbers and and I mean no stake there's no stake there but in the case of points it's actually real money involved and everyone's trying to participate in in in in a certain campaign in anticipation of a healthy reward so um I I think points are a very proven way to to help uh projects bootstrap initial traction and then it's also a very good way for Price Discovery like at least a reference point for uh the upcoming tge um so I think it will continue to be a very um essential fixture in a goto Market strategy for any kind of upcoming protocols but the appetite for it will will gradually gravitate towards the general sentiment towards the project I think it's interesting as well I mean you mentioned that points are starting to become almost like a form of prediction markets for TG listing valuation and we've seen that but I've also seen that the participants have gotten it massively wrong as well many times um so I I think it's quite interesting to watch if this Market becomes more and more efficient at predicting the TG listing valuations but you know obviously we do a lot of points you know in Def fund and we hope that it lasts so um we obviously hope that you know it continues to be a main day um yeah got it all right last last question we got 2 minutes left left here so I'm going to maybe a little bit spicy but what is the uh D5 protocol major D5 protocol that you think is the most overrated potentially your your short um and then a followon question is what would you like to see built over the next 12 months what are the you know core missing Primitives that you think the ecosystem needs to really Propel itself forward or you know just in general what are you excited about in pursuing as an investor or as a builder so so so I I'll start with I I I won't name a protocol I want to short um but but I'll start with what I'm excited about so we're very excited actually about um prediction markets and betting related products um I think that there are a lot of um innovation in um defi that can be transferred there especially when it comes to leverage I think a lot of betting markets today are what we call they are almost like fixed expiry products if I you know I would use the equivalent in finance and hence leverage can't be used but um we're looking for basically teams that can not only build such a product and I've already seen such products where you can access leverage in betting and prediction markets but not only that but understand why these problems have never been solved in web 2 and most of the time there's a reason why um and acknowledge that I think the problem with a lot of the the I think defi had a lot of teams that were from Finance Building def5 Products so they understand it perfectly but the beding world tends to be a bit more obscure to some of the builders and they think that these are problems that exist that needs to be solved and their solution fits it but often they don't have the context to which why it hasn't been solved yet so I think that's what I'm quite excited to see um you know over the next one to two years yeah um I think I'm not looking for products that or defi um protocols that will solve uh some existing problem in web 2 what I'm looking for is the a new type of ideas that were never built before using FH zero knowledge proof type of things I think we'll see new products that we never seen before like uh uh or stuff that will be be able to build that we don't actually understand today because those type of new technology and this is coming I believe yeah 100% you're good um yeah just a quick one I I am hopeful of BTC yeah um I think it's the single largest asset within the entire crypto segment and even if a small percentage of that gets converted to defi activities I think it is still a very large addressable Market yeah okay awesome well thank you guys