Defi 2.0 investing - Haseeb Qureshi

[Applause] all right so i'm going to be talking today about d5 2.0 if you're not familiar with d5 2.0 don't worry this is just the talk for you so real briefly uh let me just introduce myself so i'm haseeb i'm a managing partner at dragonfly capital which is a global crypto venture fund previously i was a general partner at another crypto fund called metastable capital one of the oldest funds in the space uh i've made early investments into projects like avalanche near protocol compound maker dao bunch of crypto projects both in layer ones layer two's defy everything up and down the stack i used to be a software engineer in a former life and a startup founder so i have a lot of different perspectives on understanding all the stuff and understanding d5 and so d5 2.0 is something that is very near and dear to my heart and i'd like to help explain it because i think it tends to be one of the parts of d5 especially kind of the new generation of d5 that's very opaque to a lot of folks who are coming into crypto for the first time and just trying to wrap their head around the 1.0 stuff so uh in this talk uh just to kind of give you a bit of a road map so first i'm gonna start explaining kind of what the hell is going on in d5 in case you just emerged from a cave and this is all new to you um i'm gonna talk a little bit about what what is defined to 1.

0 what are its strengths and what are its shortcomings uh then we'll talk about what is defy 2.0 we'll go through some of the advantages and benefits and we'll profile a few of the uh kind of namesake projects in d5 2.0 and then finally we'll end by talking about what's coming up next and then i'll leave some time for q a so that's the game plan um so first of all defy it's big probably you guys have heard of uh many of the names now in d5 it's become a more and more mainstream concept there's been this explosion of startups and investors and founders working on all this stuff and uh the it's not just in the number of companies or the number of protocols it's also the total amount of value in defy so you know back when i first started investing in d5 this was like you know 2018 2019 back when d5 was just you know a bunch of proverbial dudes in a garage tinkering with ideas that kind of seemed crazy at the time and it was more of a cool idea almost like an experiment that maybe something could become of all of this and now d5 is obviously enormous it has grown from something like uh 10x in a single year and the year before that it grew another 10x so it's massive you know back in uh january 2020 d5 had less than a billion dollars in total value now it's over 200 billion um and same thing you've seen the same similar growth for loans in d5 which is one of the dominant areas where d5 has been very successful uh trading volumes largely you know this big pink bar you see here is uniswap um largely driven by uniswap trading volume in crypto has grown dramatically so trading volume in d5 has grown dramatically to the point where routinely d5 trades more volume than coinbase on any given day um almost always d5 trades more volume than coinbase and if you look in total trading volumes we've hit over a trillion dollars traded in defy over the last year so incredible numbers and numbers that you know i can tell you the early cohorts of folks who are working on d5 had no idea that it was going to grow this fast or become this big but here we are today so let's talk about a few of the d5 1.0 names that i imagine are familiar to some of you um so the first one is compound so what the hell is compound compound is a protocol that allows you to get a loan from a smart contract so you can think of it as just a decentralized lending market there's no paperwork you have to call up a bank you just get your loan entirely on the blockchain so pretty amazing uh second protocol uniswap unit swap is responsible for most the trading volume today in d5 unit swap allows you to swap any blockchain asset for any other blockchain asset it's a decentralized exchange with no order book it's a little complicated to explain exactly how it works but it's super plug and play super easy to use and it's all it all happens very cheaply or relatively cheaply on the blockchain and then third is makerdale so maker dao is you can sort of think of it as like two mental models for it one is like a decentralized central bank the second is as a decentralized credit facility it allows you to borrow against your assets in u.s dollars which you know is very cool it mints sort of unbacked us dollars with no us dollars anywhere in the system so it's super cool it allows anyone in principle if they have assets in crypto they can get access to us dollars without having a bank account in america or a bank account anywhere for that matter um so the the big thing about each of these three protocols so one they were all very early so you know compound unit swap maker down they were all working on this stuff way before d5 summer um you know compound i think got their start in 2017 i want to say uniswap was like 2017 2016 they were first working on this i think i think it was 2016 when um i think it was um the guy from gnosis who suggested no vitalik suggested this idea to the guy from gnosis and then hayden saw this and decided oh i should do this as a protocol maker dao is like i'm pretty sure to make it out of 2016 when maker dave first raised money and like came up with the idea that they were going to do this and eventually in 2017 they actually launched the first version of it so these are all super old protocols and not only the old they're very safe they're very simple and they just work okay so part of the philosophy behind defy 1.

0 was um so one is that the primitives you see in them are very simple so what are the primitives that we use in d5 1.0 so why is liquidity mining is there anyone here or how many people know what liquidity mining is okay so not everybody so liquidity mining is basically the idea that you know most startups or most companies when they want someone to use their their project or their or their product they will basically give incentives right so they'll you know some kind of referral code or whatever this kind of stuff right you you um you sign up on our service and you get 20 bucks for free well what what you do in crypto is instead of paying those uh incentives in cash instead you pay it in equity or in the native token of this protocol uh and you sort of cut out the middle man right like a normal startup if you're let's say your uber and you want to pay your drivers to get drivers in a new city because you're starting a new city and you've got to give them some incentive to start driving what you would normally do is you'd give them a bunch of money you'd say hey i'm going to give you each you know 50 bucks if you come and drive for me well what defi does it says look you know uber has to go raise money from vcs get the cash and then give the cash to the drivers well in in crypto land we can just give the drivers just you know tokens because the whole thing is tokens and it's all native on the blockchain we don't have to we don't need to go through that extra rigmarole to give people the incentives to help bootstrap the overall ecosystem but not only that now they're aligned with us because of course you know some people who are liquidity mine they might sell the token and go get cash but some of them will just hold onto the token in which case those people are now long-term aligned with me in a way that an uber driver who maybe is like oh i like uber but i'll also drive for lyft might not be so that's liquidity money and liquidity mining really kicked off with the advent of compound which was the first protocol to do liquidity mining slightly inadvertently but turned out to be the shot heard around the world and now tons and tons of projects in d5 do liquidity mining but even liquidity mining is almost a feature of d5 2.0 d5 1.0 you know you see make or sorry not negative compound uh doing uh liquidity mining you saw ave also do some liquidity mining but unit swap basically had very limited liquidity mining and then stopped and um even uh um makerdao does no liquidity mining so it's actually not all d5 1.0 protocols even do liquidity money the second primitive is bonding curves so this was one of the things that really made d5 1.

0 kick off is realizing that if you price things along the curve continuously you can do things really efficiently on chain that the things that preceded d5 1.0 which you might call d5 you know 0.1 or something which is you know things like dharma v0 which was a you know basically people would post i want to borrow this much at this apr and someone else say oh i don't like that i'm not going to take that and so it's very you know certain kind of people it's like a bulletin board right everyone has a bespoke thing um and what we realize and of course you know ether delta the first big daxon ethereum um etherdelta was an order book which is kind of the same thing right is people posting an order someone else takes the order but it's very one by one every individual person needs to get matched and bonding curves allow you to aggregate a lot of that information continuously much cheaper and much more efficiently on chain to allow everyone to be able to play even if no one was there on the other side to take the other side of the trade right you sort of trade against a pool of the protocol so uniswap use a bonding curve to beat ether delta compound use the bonding curve to control the interest rate otherwise how do you know how much you should be borrowing and lending ad and and many of the protocols in d5 1.0 use this concept of bonding curves and then finally you have insurance funds so insurance funds are this idea that look it's possible especially in some of these risky protocols that if the if things go haywire that we may need to backstop the protocol right so you know bit max very famously had a big insurance fund finance has the seifu fund um many primitives and d5 need the same thing because there there is some inherent risk it's not a riskless endeavor to be borrowing and lending against other people in a pool and so a lot of the motivation you can kind of see where a lot of the um the thinking around d5 1.0 comes from especially when something like an insurance fund is that d5 1.

0 was all about survival in d5 1.0 there was no you know no one cared about d5 1.0 uh no one was watching in those early days it was like look if you fail if you mess up if something goes wrong if there's a smart contract hack you are on your own no one is going to save you and so it's survival of the fittest and so in d5 1.0 you can see everything is optimized around survival everything is like look we want to make sure that maker now is still around 100 years we want to make sure that compound's still around 100 years um and as a result of that there's less emphasis on capital efficiency and more emphasis on robustness now so that was the first thing about d5 1.0 which is very easy to notice when you look at all the protocols the second thing about d5 1.

0 is that when they started engaging in liquidity mining in the early days of defy but you realize that there's a lot of mercenary liquidity mining out there where basically people just come in they mine the token and they just dump it and they don't stick around they don't they don't actually become participants in the protocol and this caused you know in the early days of d5 there was this pollyanna-ish assumption that if you if you had people who came in in liquidity mind your protocol they were gonna become users and it's it's like customer acquisition right you you give some token incentives they take the token incentives now they hold your token and now they're you're a big happy family and in reality what they found was that no no actually what happens that alamator comes in they farm all your tokens they dump everything on retail and like the token just you know it's basically just pure selling pressure and um and that made people realize like hey if we are going to be doing liquidity mining we need to be really thoughtful but what are we actually getting for the liquidity that is being offered it's wrong people began to realize to think about liquidity mining as a way to acquire customers liquidity mining is instead a way to acquire liquidity and not even to acquire liquidity but rather to rent liquidity and this is a realization that would come fully fledged in the arrival of d52.0 um the third thing is that protocols in d5 1.0 are minimally levered i mean there's very little leverage in d5 1.0 um you can get leverage in something like maker dow or in something like compound by borrowing recursively but the protocols themselves are not levered right so um they and they also don't have a lot of interconnections or interdependencies so compound doesn't depend on you swap unisop doesn't depend on maker dow because remember they were all born out of this you know state of nature and so they all kind of assumed like look i don't know if make or die is going to get hacked i don't know if compound is going to go underwater so i can't rely on them i can't like you know string them along together otherwise we're going to end up in some you know great financial crisis type scenario where one thing goes down the whole financial system comes down that's what d5 was trying to avoid to begin with and so the philosophy in d5 1.0 is everything should be self-sufficient no bailouts no interconnections everything if it fails it fails on its own and of course d5 1.

0 really only use the simple primitives and compose them in very simple ways so now that brings us to d5 2.0 so what is d5 2.0 so the first thing is that d phi 2.0 is not a specific thing there is not sort of one thing you can point to that makes something definite 2.0 device 2.

0 is more of a family of uh design choices and ideas and um you know patterns that now have come to characterize more of the next generation of d5 protocols so what's different about d5 2.0 so the first thing is that you see that there's a lot of improvements in capital efficiency so remember d5 1.0 everybody was just about survival d5 2.0 is about okay how can we make things more capital efficient and how can we make things more composable because d5 1.0 kind of didn't like composability in some sense composability indeed 1.

0 is like okay it's nice if you can use compound but i really i kind of don't want uniswap using compound because then we're coupled with each other and then if we fail everything could explode and we kind of don't want that and d5 2.0 is like hey yeah come on bring on the composibility like let's like let's let's merge this stuff together let's remix things um there are a ton of innovations that are emblematic of d5 2.0 so one of them is pcv which stands for protocol control value a lot more flexibility in staking new incentive and governance designs like vote escrow tokens which we're going to talk about concentrated liquidity insurance coverage smart contracts and permanent loss design or impermanent loss protection structured products lots of new concepts that are pretty esoteric in the company it's not like lending or trading it's like you're going kind of farther and farther down the uh the tree of complexity in finance and then lastly is it's higher risk almost everything in d5 2.0 has moved up the risk curve and what that means is that uh not only are those risks in composability there are a lot of protocols that now co-depend on each other in d5 2.0 um those are also risks in leverage there are more leveraged protocols in d5 2.

0 and when i say leverage i don't mean leveraged in the sense that you can get leverage through this protocol but i mean that this protocol is itself implicitly levered and so if the market goes down you know 2x uh market goes down 50 this protocol might actually lose more than 50 of the overall value and there's some you know kind of reflexivity in d5 2.0 that wasn't really there in d5 1.0 and there are generally just fewer checks and balances around this risk threshold now of course d5 2.0 all emerged after defy summer defy summer for those you don't know was the the massive growth and and catalyst that happened in d5 in the summer of 2020 after covet um that's when d5 really started taking off and growing dramatically and of course since then we've basically been in a non-stop bull market there's been you know very very high yields on chain there's lots and lots of money and very few things fail when everything is going up or everything has gone relatively up compared to where it started now i mean you could argue maybe now we're slightly bearish sentiment but at the end of the day very few things in d5 so far have failed and the last time that we had a really severe drawdown was in march 2020 when covet hit that's the last time that we saw a really cataclysmic event that hit all of d5 and d5 1.0 survived it but it's quite likely that a lot of stuff in d5 2.

0 could not survive a march 2020s type you know sort of three sigma event um and that's by design these protocols are intentionally levered um and of course when you're in good times more and more leverage is natural it kind of happens uh as a natural evolutionary byproduct of a free market so what are some of the primitives in d5 2.0 you know d5 1.0 we talked about bonding curves and liquidity mining d5 2.0 has a few tools that it likes to use in a lot of the protocols so one is protocol control value and protocol control value is the idea that your protocol has its own balance sheet and using its balance sheet it does so d5 1.0 protocols they have balance sheets right you know uniswap has a bunch of uni tokens compound is a bunch of comp tokens but they don't do anything they just it just sits there and it just you know maybe they do grants or something right but they don't like go around and muscle the market with the huge amount of money they have on the balance sheet but this is exactly what tribe does uh or fey uh uh which is a stable coin this is exactly what fracks does right fracks uses its balance sheet to mint on backfracks and go do stuff in the market and this is new this is more like what you'd see out of a central bank where they use their balance sheet to go do stuff whereas makerdao is like a central bank but it never does any it's a central bank that just you know it's like you know i don't know the hong kong dollar or something it just sits there it's like i'm just i'm just a dollar i'm that's it i have no other aspirations in life um and the the the pcv crowd is very different so the second thing is liquidity as a service and the queries of service the idea of like look when you guys did liquidity mining liquidity mining was basically renting liquidity from retail when you when you do liquidity money basically what you're doing you put up a bounty you say hey i'm going to pay out this many tokens to whoever gives me liquidity and people give liquidity to the protocol and essentially when you stop giving up that bounty the liquidity goes away that's what we see right we see this very cavalier mercenary capital that is the same thing as renting capital renting liquidity it's it's equivalent to a rental so you're renting liquidity from retail now renting liquidity from retail is fine but you could also i mean you could do it you know real grown-ups do which is rent liquidity from you know professional liquidity providers and that's what we've seen is sort of the rise of dows that are effectively professional liquidity providers which is tokamak olympus pro now the third category is what's called vote escrow governance so vote escrow governance you've seen this ve thing yeah ve prefix this ve prefix what it means is vote escrow and what vote escrow basically means is that if you want to participate in governance you need to lock your tokens and the longer you lock them the more governance power you get it's basically a way of saying like look you can't just like buy up a bunch of google stock and then go do like a corporate takeover of google which you can't anyway because the founders control so much but imagine you could the the idea in vote escrow governance is that actually what we're going to do is we're going to avoid the situation where people come in in a very cavalier way and they come in and they try to take over the protocol by buying a bunch of governance tokens instead the longer you commit yourself to illiquidity in this protocol the longer you sort of tie your hands and say yes i'm going to be a part of curve or i'm going to be a part of convex you get more power in the protocol based on how long you tie your hands and lock and throw away the key so that's vote escrow governance and the last thing of course is strong memes so you guys are probably aware of this especially with olympus dow wonderland mim or abracadabra these are very these are communities really tied together by a lot of memetic power and the emphasis on memes the emphasis on these kind of um very online and active kind of trolley communities is very different from d5 1.

0 right there is no uniswap troll army but there is a wonderland troll army and this has become more and more emblematic of how these communities form and how they also get early adoption and um and die hards who end up being very committed to the protocol and keeping it keeping it running so now liquidity this liquidity part is interesting right because we've we've started to move further and further down the the spectrum from uh what started with pure retail being the way that we satisfied liquidity on chain to moving more and more to protocol rented liquidity which is what tokemak does tokamak is basically it's basically a market maker that's a protocol and so you can say hey tokemak i will pay you some money implicitly if you will mark can make my token and tokemac's like okay i will do that just like a regular market maker it's like paying you know jump or or you know some other citadel to market make your token and then you have olympus pro and what olympus pro does is basically allows the protocol to itself own the liquidity so in a way it allows the protocol to itself become its own market maker which is again a lot more similar to what a central bank does a central bank goes in and market makes its own currency right it basically goes into the market it buys it sells in order to make sure that it hits its price targets that's what essentially olympus pro does it allows other protocols to do the same thing to go in and own their own liquidity which in this case owning your liquidity means you are the market maker you are the one buying and selling and setting price and giving everyone liquidity in the market um and that is now increasingly a thing that's being taken over by protocols themselves which you know in a normal world if you think about it if you zoom out far enough and imagine this from the perspective of normal corporate governance it's kind of weird like imagine if you know google was like hey we are going to be the buyer and seller of all google stock and where everyone wants to buy or sell google stock they buy it from us it's kind of strange but that's in fact what olympus pro enables it enables anyone to sort of become this so um so let me walk you through a few examples of some of these d5 2.0 names some of these i'm sure you have heard of others of them you are maybe very deeply familiar with as i'm seeing from some of the reason you guys faces so uh olympus now i'm not going to belabor exactly how olympus dial works because if i tried i would break your brains but um i'll tell you what it portends to be which is that uh it says that it's a decentralized reserve currency protocol but it's not pegged to anything okay um the big the big thing about olympus now that i think connects it to most of the other things in d5 2.0 one is obviously the the the pcv so the protocol controlled value which basically means that olympus basically takes uh it basically buys up its own liquidity to make it so that olympus is basically the market maker for ohm um and uh that allows it to be the kind of central bank for ohm because it owns all the liquidity and therefore it doesn't have to worry about you know cavalier or mercenary capital coming in and jumping in and out of the liquidity pool instead it's all owned by by the protocol and so this is a this is a very new model compared to what came before and then of course the memes right so you guys have probably heard of this meme three three um again i'm not gonna explain what this means because it it will also break your brain but um it basically is a is effectively a rallying cry for the holders of olympus now of the home token to never sell and to keep keep doubling down tripling down on olympus now um and this is very common in a lot of these 2.0 names is that you know in d5 1.0 if you're in maker down and like maker that's not working nobody's like you must hold maker if you don't you're a bad person and you don't understand the vision of maker dow they're just like look i mean you know do what you're going to do but like the system's going to be secure whereas with d5 2.

0 it's very it's very like no no you're a bad person and we're going to yell at you on twitter if you sell um so that's olympus now now the second protocol it's interesting because it's a very similar family of protocols as olympus now abracadabra which uh it's its native token or sort of the governance token is spell uh the stable coin is mim mim and basically what is abracadabra the easiest way to understand abracadabra is it's just maker dao but with riskier collateral okay so maker dao has a very laborious process for onboarding more collateral right if you want to get something included into maker down you got to go through giant forum arguments and people do risk modeling and all this stuff and there's like very low uh ceilings on how much you can actually mint uh when you want to when you're onboarding new collateral and um abracadabra was like hey you know all those like weird levered tokens that you have you want to get levered on them come to come come to ours come to our shop we'll like let you borrow against all of it and so they they basically take a bunch of leverage collateral which is already in other protocols like yield generating collateral and other protocols and they let you borrow against it which is just leverage it's just that it's just going farther down the risk curve and so in a way again this is very normal in many markets is that as the market matures you see more and more businesses getting built farther down the risk curve taking on more risk and that's effectively what abracadabra is so last thing is fracs i don't have a bunch of time so i'll run this pretty quickly fracs is a partially collateralized partially algorithmic stablecoin um i'm not going to i'm not going to belabor the way that it works there's a i wrote a blog post about fracs that you can look up if you're very interested but the way that the thing about frax that's very d5 2.0 is that fracs unlike most of the stable coins fracs can mint unbacked fracks like a central bank you know a central bank like if you're the fed you do not look into your reserves and say hey do i have enough money to print usd no you just print usd because you're like look i have stuff to do i'm gonna go do it um and so fracs is the same thing it's more like what we think of as a traditional central bank it mints on backfracks to go pursue attractive market opportunities if it says like hey you know what i'm gonna go mint a bunch of fracs and put it into compound so i can earn a yield on it maker doesn't do that right maker doesn't mint unback die and say hey i see there's a lot of great yield farming opportunities and die i'm going to go make it myself as a central bank dye doesn't do that maker doesn't do that but fracs again moving farther up the risk curve being more aggressive much more uh sort of much more activist in the way that it uses its protocol treasury um dows are also evolving alongside all this d5 2.0 stuff so you're seeing you know the d5 1.0 daos are very very old they're you know maybe arguably more sclerotic they don't move as quickly um i i have written here somewhat somewhat optimistically that protocols are becoming more decentralized it's kind of less true actually a lot of these protocols are not as decentralized as the d5 1.0 stuff but the d5 1.

0 stuff is made with a different design philosophy right d5 1.0 is like look we build this stuff and it's going to stick around forever and it's never going to change unless we need to like do small little changes with the margin whereas d5 2.0 almost everything is iterating very quickly and as a result things are moving faster there's more multi-sigs than there are you know kind of decentralized whatevers and there's also a lot more treasury management and oversight that is more complex because these protocols do so many more things there's more to govern in these protocols which makes them more complex and the team structures are becoming also more complex right so you guys may have seen this big sushi drama where arca suggested basically like a whole org chart defined by the dao for sushi and uh and this kind of stuff is going to become more and more necessary because of all the complexity for most of these d5 2.0 protocols um so what's next um i think you're going to see more and more again you know these uh i didn't have time to talk about convex which is very deeply intertwined with uh with uh with a curve uh also fracs is very deeply intertimed with convex and with curve there are all these interconnections tribe very similarly there are a lot of interconnections between these protocols and i think you're going to see more and more of this kind of money lego concept finding its way into more and more d5 2.0 sort of remixed protocols um without d5 1.

0 we wouldn't have bonding we wouldn't have liquidity uh liquidity mining we wouldn't have you know a lot of these foundational primitives but they're going to get built on and just like finance becomes more and more complex over time we're seeing the same thing happening in d5 except you know the on on speed run uh speed compared to what we saw in traditional finance um there are a bunch more projects i didn't have time to talk about things like ribbon finance gelato tracer dao popsicle visor dopex inverse finance so do your own research also i should be clear none of this investment advice i'm not necessarily saying any of these things are good investments but if you want to understand where defy is going it's very important to get a sense of how the space is evolving and what people are working on so how much time do i have left do you have time for questions no i have no time for questions thank you for your zero questions uh i hope that was helpful wow give it up for us steve that was a lot thank you so much