How Modularity Drives the Next Wave of DeFi Lending | Michael Bentley - Euler Labs | Laurence Day

oh my God we dressed for the occasion today I think you should stand up and give everyone a clear clear view do like it let's make sure to get it on camera too got some more if anyone want we've been talk we'll talk more about the swag after the session um GM everyone how's everyone doing okay everyone's asleep GM everyone how's everyone doing thank you thank you hey everyone I'm Lee I am a general partner and co-founder at variant fund we are an early stage Venture fund focused on crypto investing and today we're here to talk about how modularity drives the next wave of defi lending Michael Lawrence it's so great to be up here on stage with you guys today you two are two of the biggest builders in defi right now um and you guys also have something in common which is that both of you guys have been involved in pretty notable defi hacks in the past I would love to start off with just brief intros why are you still here why are you still building in defi Michael let's start with you uh I think there's probably something wrong with both of us to be honest uh should say that right away so yeah I'm Michael uh co-founder and CEO of Oiler labs uh what is Oiler it was a lending protocol that we started building back in 2020 during the the height of defi summer uh we started building well I started building that before that I used to be an academic I used to be a research scientist studying Game Theory at Oxford and uh yeah I just found everything that's happening in defi somewhere really really fascinating and so jump jump ship uh by 2023 Oiler was one of the most popular lending protocols on ethereum it was the third largest uh and it was absolutely flying and uh the the year sounded looked really promising ahead and everything and then uh a really Niche part of the oiler code base was exploited uh with uh yeah great effect in uh in March of that year uh draining 200 million uh in assets from the from the protocol and after a lot of hard work from the team uh we were we managed to recover all of the assets uh 240 million when returned back to users in the end just a few months later absolutely harrowing experience uh but one that we've back much stronger from as a team we learned an awful lot about ourselves and particularly our resilience and everything and uh yeah really excited and uh you know uh energized to be here now and see all the excitement about the uh oila V2 that we announced uh just a week or so ago so OA V2 is uh also a lending protocol but it's a a modular lending protocol uh that we think is going to unlock uh a massive amount of uh flexibility for lenders and borrowers more than they've ever seen before uh and we think it will will form the base layer of uh a liquidity layer for the whole of defi in the years to come so excited to tell you a bit more about that as we go along yeah we'll definitely delve into Oiler V2 and what that means but first um turning it over to you Lawrence why don't you introduce yourself a little bit God tell if the M oh no the microphone is working I'm Lawrence I'm the co-founder and CEO of the Wildcat protocol it is a uh move towards the flip side of what oila V2 would be focusing on which is under collateralized credit on chain prior to um all of that I was at index Finance which I imagine we'll talk about at some point but I was also an academic um I finished a PhD in compiler Theory and then I worked as a risk reporting analyst and model risk Quant at standard charted for years and years I started doing ethereum development the day after the parody multi uh Sig freeze because uh that was my moment of realizing that I'd spent years programming around money but now I could program money directly which was an irresistible nerd snipe um why am I still around I I couldn't see myself doing anything else this kind of you know program building in an adversarial environment is to a certain type of person the the only thing you could imagine yourself doing yeah I I sense a bit of masochism in common between the two of you which I don't know is that a British trait it's it's Unique to the part of leads that we live in yeah that's right it's not not a great so you know we're used to it that makes a lot of sense so Michael I wanted to start with Oiler so you've had a lot of ups and downs over the last year um to say the least including the infamous $200 million hack of Oiler and then the subsequent recovery which seemed rather miraculous and you chronicled all of this in a blog post that was aptly titled war in peace and it was approximately the length of tolloy war in peace actual War and so given that that essay exists we won't get into the weeds of all of that and what unfolded there we'll let people read it for themselves but I would love to talk a bit more about the key lessons that you learned from that experience and how it informed Oiler V2 yeah sure I mean the one of the I mean something that I guess is obvious coming into the industry but that was really like brought home to us was that building smart contracts uh that secure millions of dollars is really really hard I mean oil of V1 uh we you know as a as a team we always had this philosophy uh to build you know things slowly and thoughtfully and carefully you know when when you when we first joined in 2020 there was a bit of a culture of like build fast and break things and that was always that was never something that we really we really believed in so we always thought yeah build build slowly and thoughtfully and then you know Oiler was exceptionally well audited uh we had the largest bug Bounty of any any lending protocol we we've done formal verification all the rest of it and yet still this Niche part of the the the the code base uh someone managed to find an exploit so it's really a lesson not just to us but to everyone else I guess about how difficult it is to secure smart contracts uh and I think the fact that we were so thoughtful about stuff did earn us a lot of Goodwill from the community and we were really really grateful for all that when the when the exploit happened and the recovery happened um but yeah a key key lesson we learned was uh was about the the team itself and how resilient we were and how hardworking we could be uh how we could support one another uh and ultimately that same team is back now we're building V2 uh in that same slow and thoughtful manner not trying to rush things uh with a real heavy emphasis on on security um one of the things that we're really trying to do with V2 is abstract all the best bits of V1 and Abstract all the best bits of a lending protocol generally and then repackage that and rebuild it up in a modular fashion uh and with that modularity comes um yeah a lot of flexibility for the end user but it also allows you to build this uh really robust complex product at the end which you know every component every individual component has been really well tested and is really easy to understand uh yeah so that's that's the one of the key lessons is is is yeah simple is better modular is better yeah so you really took this opportunity to rethink the design of Oiler from the ground app and build it in a more modular fashion um Lawrence turning it over to you obviously you've also been through a lot from your days at index Finance I would love to hear about um the lessons from that and how it led you now to Wildcat yep so I wouldn't say there was a direct path to Wildcat from it index Finance was a play at moving ETF funds um into a tokenized Mana using balancer LPS that was a $16.5 million hack and the the subsequent wrangling and filing of lawsuits around that really kind of made me realize where the the barriers were or the boundaries between the existing legal system because like the law has not caught up to defi yet that the we we tried to fit Square pegs and round holes when it comes to dealing with the aftermath of these things in the legal system which drove me to law school to try and become a better defi citizen I've always been a believer that defi is taking the best bits of financial systems and moving them into a place where we have no middlemen that are extracting fees I do think that you know moving on to the next thing once ETF stuff has now like there are other mechanisms for it that are perhaps custodial but I believe that credit mechanisms on chain are one of the remaining white whales that we really need to solve for for a digital economy on Rails I think we you know we all use over collateralized protocols the you know margin protocols but for growth um we do need a mechanism for cleanly introducing effectively lend which is leveraged by proxy um on credit and so I've been working on that for the last few months we've been live for three lessons that I would have learned I think I've always been quite paranoid about security and I think that's certain about smart contract security because as Michael says it is hard to work in this kind of like cheing complete environment like with when fuzzing is not really welld developed but you can gate things there are mechanisms to do that now and I've started embracing these which have some costs but you know you you breathe a bit easier right I I wanted to level set on the title of this panel which is about modularity in dii and actually I think it would be helpful to just level set on definitionally what does modularity mean it's a buzzword that is taking off right now but can you share your definitions of modularity yeah there's there's a lot going on at the moment with this where you see it all over the place this word it's it's definitely a buzz word of the time uh for me modularity means that uh uh modular code means uh something that's uh really short really really simple does does one thing and it does it exceptionally well uh and it's easily audited easily tested and uh that it can also be used like money Lego itself it can also be used and composed with other modules um so I think yeah when you're building uh when you're building a product for things you can either start really from the uh you know the top down you want to you find something that users want and you really just want to Target that exact thing uh and you end up building this really big monolithic code base that has a lot of State uh in there and it's very um yeah it's very hard to go from that when you build top down to then extend it and test it and make sure every single code path has been has been tested with modularity I think for me it's about building uh identifying and abstracting key components of a system wrapping them up into these modular components uh and then building back up uh you know back from the bottom bottom up towards a more complex product um but I also I also wrote recently about modularity that there's there's different flavors right um there's not just a you know modularity by itself isn't isn't necessarily useful if it's if it's chaotic uh and I think if you look at uh in the past if we go back to the web 2 World um when you look at the development of things like Windows and Apple and Mac uh Windows was a modular system people could it was an open system people could contribute their own uh applications to it but it was an absolute mess that's what I would refer to as chaotic modularity the modules were built but they weren't built Co you know in a way that they were designed to work and fit together very nicely and so you had drivers for your printer and drivers and it was just it was just awful and uh Mac on the other hand you know uh I think they also built using these modular design principles but every app felt like it was designed to work with every other app on the Mac and so I call that cohesive modularity and so with Oiler in the long run like what's our sort of long-term ambition for building in defi is to build in a in a cohesively modular way so that we have Oiler as a lending protocol and then in the future other protocols to come that can be composed with one another as money Legos that have been really uh welld designed to fit together um right now when people uh trade in defi uh any given trade can often end up touching you know four five six different protocols Each of which uh has its own fee there's like hook fees and UI fees and there's you know lending fees and swap fees and so on and your exposure to different risks and different governance systems and so on and in in the long term our ambition is to build a suite of apps which uh minimize fees for users minimize friction between them and work cohesively together in a modular way right so there's basically like different degrees of modularity and different ways that modularity could be implemented and there are different user experiences that um arise from that ranging from kind of the mess that you described with Microsoft products all the way to a much more cohesive and user experience that is still modular Lawrence do you have anything to add to that yeah I mean what we disagreed about this yesterday where you know things that you'd call modules I would call Primitives um in lending I mean my the word modular has been shadowed now completely by Celestia where we talk about data availability but I mean I more of the take that kind of I would still consider a protocol modular if you could deploy different functionalities as sets of protocol of contracts which supported functionality all under the same rollup so you know in my context it's a fixed duration markets versus a liquid Market versus one with a call option they're all part of the same thing but they are independent networks but beyond that nothing to add for yourself right so modularity Primitives yeah all kind of describing relatively the same Concepts I want to bring in the idea of risk management into this conversation about modularity and how modularity impacts um risk management so Oiler V2 and Wildcat are taking pretty different approaches when it comes to risk management can you explain each of those respective approaches and why you decided to go with that um yeah uh so I mean first first things first is to say that uh in in defi the idea I think a a s founding principle is that risk management is that you should have freedom of choice um and uh really it's up to you to uh determine what your risk reward preferences are like if we just ask each other now what our risk reward preferences are be all really different and they change data depending on our context like Leah I know you like skiing I like can't imagine yeah that just sounds awful the risk reward just doesn't sort of map up for me yeah every user can express their own preferences and agency yeah so yeah it's about having agency and um but then yeah if we look at the differences between Oiler and Wildcat um and sort dig in there a little bit I think fundamentally they they take different uh approaches because there are different types of lending application Oilers over collateralize lending and and in an over collateralized system uh as long as you can value the the collateral and ensure that the the collall is always worth more than the um assets being taken out in loans um then ultimately you can encode a whole variety of different mechanisms in a smart contract that can can do that without needing recourse to uh any legal Tex right so with over collateralized lending you can do everything in the smart contracts I would argue um and that's a little bit different to what happens with Wildcat I guess but then uh with over collateralized lending even within that there's there's multiple different models of how you can manage risk uh I wrote an article recently uh about paternalism versus the Invisible Hand in in Risk Management of lending protocols because I think even among you know in among lending users of lending protocols there's two different types of users that we see today some users like to uh deposit passively into a lending protocol and have the the risk managed for them their expectation is that their assets will be in there the parameters will ensure that you know uh changes in the economic environment will be handled by the protocol itself um and typically with that you know the risk reward might be the the reward might be a little bit lower but the risk is kind of managed on the other hand you have more like power users so I I suppose who want to uh take more risk and uh uh and and yeah go for higher rewards uh and with that they have to actively manage their positions and they're deploying deploying their assets into uh a spectrum of different isolated pools maybe and with that you end up with this sort of free market emergence of a risk management system where the the the Invisible Hand of sort of capitalist uh you know free markets ends up determining what uh you know what the risk reward is across the the Spectrum as a whole so yeah we've we've uh one of the the benefits of the modular architecture of Oiler is that it allows us to build uh that optionality into the protocol for users of different types some users will want the the the paternalistic model and they can have that and other users will want you know real immutable pools that can't be touched that where the you know the risk is uh they actively manage their own risk preferences and that's that's exactly what Oiler has built to to uh drive and that's why I think it will be this liquidity layer in the end for defi because of all the flexibility in the system to to really yeah manage your yeah manage your own risk reall preferences yeah I mean we we are focusing on a completely different type of risk then you know like um if if you're dealing with something that's under collateralized your risk is that of defaults rather than liquidation so you know there always when you're dealing with something like that you've got two as two sides of things you've got in my context is a smart contract risk um and you've got a default risk so we will allow people to utilize Wildcat they can create insane markets if they wish if someone on the other side that wishes to lend them decides that's in their risk reward function absolutely fine go for it um whereas what we do one of the lessons that we learned was instead the risk that we can mitigate which would be the smart contract side of things we very heavily clamped down on such as we are specifically whitelisting call data access into functions which that's where we can control things so like the risk of the actual lending itself is very much to themselves in a similar way but when it comes to contract interaction very much locked down that does look like it takes away some of the modularity but the re I mean the reality is that the the debt tokens that are issued as a result are as much a money Lego as anything else that makes sense so it sounds like the the commonality is really allowing users to express their own preferences and choose their own level of risk tolerance and to participate accordingly yep yeah definitely yeah Michael um I wanted to talk about something that you mentioned in your recent post about Oiler V2 which is about this vision of a flourishing defi ecosystem that includes real world assets non-fungible tokens IUS for uncollateralized Lending peer-to-peer lending and Oracle free lending that all sounds great but how is that going to become a reality well uh I would say is already a you know becoming a reality right as we speak right there's lots of great projects launching that are tokenizing uh tokenizing things from The Real World whether it's you know equities or treasuries or bonds they're all these things are slowly but surely they're they're becoming tradable in uh within Defi and uh I think that's really exciting you know for the past few years defi has been a kind of isolated environment you know crypto as a whole it's been disconnected from traditional Finance um and that's that's slowly slowly but surely changing thanks to U you know protocols like oo and others that are enabling those assets to to come on and mix in with other other um you know crypto native assets I suppose um yeah the we're seeing you know yields from treasuries now slowly flowing in on on chain um which I I think is help going to help the the sector grow it's also great for traditional Finance where there's been a like a Darth of innovation for for decades now uh I remember when I I first left University for just a few months I ended up working at the Royal Bank of Scotland back in 2008 when the crash happened and uh just like I was just couldn't believe how bad the computer systems were the risk management systems everything was just decades old and you know and now with these with like defi has just been moving so so quickly defi rails and the code bases are really really Advanced and it's um I I think there's a lot of opportunity there a lot of growth opportunities for real world assets to come onto Defi and not just so that you can earn treasury yield by holding a tokenized real world asset but so that you can trade these things we have these Free Fair open markets that are very very transparent uh and much more efficient um than the Clos systems in trafi so yeah again in terms of Oiler and and and how it will enable this um a lot of lending protocols today just allow you to uh lend and borrow fungible tokens so ERC regular ERC 20s but with Oiler you can plug out those things and plug in nfts as collateral or uh use hooks like uniswap V4 to encourage use of um you know tokens with maybe secondary transfer restrictions and that kind of thing and so yeah I think the modularity of of Oiler V2 is really going to uh allow allow massive growth of that that's the sector that we're already seeing today right so it's like the the modular design of Oiler actually enables the system to scale more to new types of assets that aren't really able to participate in defi right now yeah absolutely it's just simply wouldn't have been possible in you know with some of the more Legacy things even oil V1 like you know RV compound other competitors you can't just throw those things into the system and the nice thing about oil of V2 is it's so so flexible uh and that's really really powerful for giving users the options to do anything you know but you mentioned like uncollateralized that in principle could be done although I'll try not to you know make yeah you'll be all right yeah that's really cool so as we start to wrap up here I was wondering if you guys could share some predictions or thoughts on the road ahead for defi over the next 12 to 24 months like what do you see unfolding from here in terms of both the opportunities as well as challenges h I think we we keep banging on about this the thing that we need is some sensible regulation um we are not getting it out of America anytime soon but if you look over to Europe we're starting to see some really good sensible like wellth thought out things from say the UK law commission there's consideration going into this in uh the EU as well like Micah 2 is coming out which is focusing on defi you know for better or worse but there is thought going into it I think the thing that we need need to be aware of as we head into you know kind of what we will now call the next cycle is just avoiding cdy platforms this opacity breeds collapse um I think we need to be very careful of it yeah I I could could agree more with that on the on the regulation side I think um yeah the elephant in the room is still you know defy smart contract smart contract risk but I think we're seeing you know it might not look like it to Outsiders but there's a there's actually a huge amount of progress when you work internally and you work with all the the top security firms and so on you see a huge amount of progress behind the scenes in in the the developments in formal verifications and like all the tooling for fuzzing and all the rest of it it's all getting incrementally better day by day by day um and so yeah I'm optimistic that we will see a much better year this year than we did last year and the year after and so on things will will uh yeah just keep improving so I think that's really exciting and then with that um that's been one of the things perhaps holding back more uh yeah institutional demand for for getting involved in Defi and so as that improves and with things like modularity helping pave the way for I think I think we'll see more real world assets and uh and things come and change more more that that sort of meme of the institutions are coming might actually finally it turns out they are actually this time is different it actually maybe is so yeah got it so the institutions are coming and hopefully smarter regulation is coming although don't hold our breaths on that right fingers crossed yeah and hopefully we can make that a reality Michael Lawrence thank you so much for being here on stage today this was awesome let's give it up for Michael and Lawrence thank you very much thanks everyone good [Applause] job