State of Ecosystem Risk

Transcript

uh I want you guys to please make him feel very very welcome uh he's going to talk about the state of economic risk put your hands together for Nick Canon hello uh my name is Nick as mentioned uh I'm going to talk about the state of economic risk I work at gaunlet uh we provide Financial modeling Solutions built sort of using trafi techniques to help defi protocols understand and mitigate those risks the applied research team that sort of you know leads a lot of This research um is thinking about Novo Primitives and mechanism design and that informs a lot of the modeling and simulations that we do uh we work with protocols and Dows like a and compound and large delegates

and active in sort of the governance ecosystem as mentioned in l2s like arbitrum so straight out away um economic and ecosystem exploits by risk category as you can sort of see um the notional amount loss to economic exploits is more than security and governance exploits combined hacks and exploits commonly classified as security Oracle governance or economic also known as Market risk which what is often surprising is the disproportionate attention given to sort of the left side of this Slide the security and governance the the classic hacks the governance rug pools and

similar whereas we can see notionally on the right uh just in size and quantity economic exploits um are quite frequent and and large today we'll run through sort of one big happy ecosystem sort of where it all came from um permissionless ecosystems um where they're sort of evolving PVE not player vers environment maybe protocol vers ecosystem and what's next all started sort of one big happy ecosystem ethereum and all um this is a snapshot of you know the loans ending in 2018 uh not my current gig in D5 the previous one we sort of put this report

together as you can see you could sort of fit D5 lending in one little doughnut um and it was a stretch at that compound Dharma and dydx really wasn't PVP back then um Dharma had a P2P lending model model which is having its sort of comeback right now compound has the dominant model that we've all sort of seen fored endlessly Etc what they're really just trying to find is not you know compete with each other but really fine if there was users product Market fit Etc if you map sort of D scurve um today in this classic you know technological eruption disruption cycle you might categorize it something like this first there was eruption um D summer kicked off with compound and it's token uh

frenzy where forks and tvl growth were anchored by Sushi's vampire Tech uh 500 plus Forks of compound and Unis swap alone a Synergy protocol deployment sort of across networks a Unis swap in 2021 went to polygon and then other l2s as well then this very much spurred consolidation and then sort of maturity not you know maturity for the end of D or the end of crypto but just sort of that the stage we're at in this sort of life cycle the winning Dows go vertical aing curve designed stable coins maker has a lending protocol uh Unis swap does nfts synthetics does perpetuals lending coming into 20 2023 sort of look like this uh mentioning the forks the

bottom two you know the biggest Plus on this graph the charcoal and the sort of light blue are compound in a very commonly forked uh had over 90% of market share at the beginning of 2021 but coming into 2023 um while that had receded a little bit they still have only about 80% um you know mid early 2022 um they had you know lost a little a little bit uh sort of a cross chain where they had yet to deploy um but as we see sort of they've been able to Abate uh a lot of these forks and uh attacks on their market share we see a similar pattern for dex's um most notably Unis Swap and all its Forks uh udis swap didn't deploy as

mentioned uh to polygon until 2021 um but you know starting at 70% uh in early 2021 of market share and then down to his low is like 30% for sort of amms of the V2 and V fre flavor that are using their contracts uh towards the end of or coming into 2023 and even now we've seen them gain that market share back this big charcoal block still in the middle is pancake swap on BSC Unis swap just in March had deployed to binance um and if they do sort of or if we see a similar pattern um of that when they deployed on polygon we should sort of expect them to regain that 2/3 market share or at least they

hope so why are we talking about you know consolidation s-curves and protocols I thought this was about ecosystems well you know there's sort of things to learn the market structure of these protocols actually informs how these ecosystems evolve and like where protocol devs the next ones building their next deployments or the New Primitives actually think about deploying it doesn't have to be on Main net anymore two things two or three things inform this um Market leading protocols sort of have shown and proven resilient to disruption and longtail fragmentation protocol consolidation without m&a it's really tough to do protocol m&a and Dows and there's been very few attempts at that reveals traits of win or take most outcomes it's tough to expect uh trafi to sort of look like or defi to look

like trafi in the end do will there be 50 to 60 sort of global exchanges with stable market share maybe probably not and then the market instability you know hinders our ourselves gunet and other service providers to drive value across multiple segments but this also goes for ecosystems in their dows how do they think about Blue Chips vers new entrance so aren't these permissionless ecosystems yes sort of um Talk want to talk about the Frameworks uh a lot of people throw risk Frameworks out there how to classify how to segment these different risks and the reason this obviously matters is like if you're trying to think about ecosystem goals and needs where do you allocate them how do you get visibility into what

is the biggest pain for pain point for your users and of course like the protocol is building on top of you I generally classify it you know without getting too philosophical about the debate but an ecosystem risk is anything that touches multiple protocols this is very commonly um known or seen through flash loans uh the mechanism that basically allows for you to borrow in an atomic transaction a certain collateral from an A a lending protocol and then use it elsewhere to borrow or trade or something similar so sure there's risk but so what um a few reasons this actually matters to

ecosystems one they're fragile uh as I'll show in like future slides market share concentration is high to extreme in really all categories uh categories being lending IND deex and CDP and synthetics and things like that switching costs are very low uh the ux might be bad but if users can bridge in or on to your L2 or your chain they can definitely bridge out and while the you know the barriers to entry the switching cost are a little bit higher for protocols we've seen cross chain deployments pretty easily uh it takes a little bit more thought than a few transactions but definitely doable and also liquidity breeds liquidity this is sort of tried andrue from traditional Finance but definitely here and exists now permissionless composability can of course accelerate growth but also

introduce systemic risk so what does this look like across these ecosystems sort of chose five and wanted to Benchmark against you know ethereum a little bit um what you can see is the traction uh so sort of on the x-axis you're going to see time obviously and then on the y- axis you're going to see concentration of the top four protocols that are Market risk bearing so what we're excluding here primarily is amm while obviously the liquidity is extremely important they aren't parameterized generally a fork V2 Unis swap contract isn't really adding you know measurable or anything more than negligible risk to a new ecosystem so the categories being sort of plotted

here are primarily lending CDP um synthetics and derivatives you can sort of see concentrations all over the place um it's not extremely stable um up and down uh arbitrum seeing a dip in sort of um heading in the right direction post token launch optimism as well uh but they would admit in the Retro they did in their governance form that they probably overspent for this um they paid a lot in allocations and there wasn't a lot of accountability bance is a little bit deceiving in the sense that like majority of liquidity on those chains are dexes and launchpads Market concentration um as mentioned before is definitely key so again on the y- AIS you have the top

four protocols by each category and hhi or herend Dale hman index is a common concentration metric so what this does in other sectors is basically compare how how to reason about stability and how you should think about at least for ecosystems you know what is driving what is driving your value and your users and and and that traction um everything below you know for the most part Dex is an aggregators only thing below 75% concentration everything else where Services is sort of the middleware layer driven by anchored by the instap CDP in maker uh synthetics of synthetics algo stable coins up in the top right uh anchor by fra as you can see these are very concentrated by traditional Market

standards anything above 2500 is considered concentrated above 5,000 is just extreme this makes it tough to plan and makes it tough to have a competitive Advantage but also tough to allocate resources one thing I want to touch on you know as pre previously shown is Oracle versus Mar Market manipulation Market manipulation really just involves trying to move the prices of an asset in traditional finance a lot of this is illegal uh there's descriptions like banging the clothes where you're you know manipulating an underlying asset um and Oracle exploits mainly involves uh bad pricing data um security audits are for the category leading Protocols are very high as you

can see Lio with eight uh this is a on the right with nine now granted these are for different parts of the code base and different versions but for category leadership these Protocols are understanding that like redundancy is actually needed and the the incremental value here is extremely extremely valuable unfortunately smaller protocols really can't afford this uh how do they think about the incremental doll dollar chain that they should spend should they spend it on ux should they spend it on marketing and this goes into how they think about deploying to new ecosystems when they get that traction will they get that incremental audit when it actually matters more back to the chart from the beginning um now that we sort of

understand this breakdown you you sort of want to think of like how should an ecosystem who has funds public wants to think about public goods wants to think about standing up subow where they should direct their intention is this sort of a snapshot in time we see all these audits have already happened but like is security and governance solved this this like a 20189 problem uh what is what does it happen now are all these economic risk and exploits new should we focus on that now well actually not plotted over time the blue being in economic and the orange uh for security this actually shows pretty consistent over 2 and a half years both in as mentioned before size and count PVE rather than you know player

versus environment protocol versus ecosystem now of course there's tons of synergies and alignments protocols you know want novel Primitives they want people their protocol or ecosystems want novel Primitives they want them to be successful but those don't match one to one I don't think anyone would exactly expect users protocols and ecosystems to have you know perfect alignment but there's probably not enough attention to the weight that Protocols are throwing around they get a lot of ecos u a lot of incentives they have delegation power uh they have other options in the space and um in different time Horizons they're operating under for example uh the goals of the arbit arbitrum Dow uh you know recently spun up about 3 months old I I read a few hundred delegate platforms and these

sort of sum up to two main buckets growth and decentralization attracting Builders onboarding new users being efficient and sustainable uh decentralization for fairness transparency to Foster education and avoid conflicts this seemed kind of obvious um but how do you go sort of go about doing that back to like what is the biggest problem how do you understand the sustain sustainable competitive advantage that makes sense to deploy you know what is a multi-billion dollar treasury when I talk about stability um other Industries do this as well uh this is a analysis of US Auto industry and what you're trying to look at is the absolute a average value change over time usually in more stable Industries

uh this is like three and fiveyear trailing and what you can sort of tell anything below sort of 1% which is about where US Auto lands is considered stable or under 2% sorry US Auto is at about 1% as you might guess defi is not exactly like that so this is what defi looks like even when we shorten the time span to 1 month and 3 month this is the top 15 protocols now they're not exactly sorry the axises got a little muddy but um same on the time um and then you know this mid part is between 10 and 30% the majority of the fluctuations so what you're seeing um is

how this tracks over time the biggest protocols really aren't worried about market share uh per se as they would in US autos and sort of a more predictable sustainable industry The Tam of today is definitely not the Tam of next year at least they're not building you know with that in mind but this makes it tough for ecosystems and their communities and Dows to reason about like who is going to hold the top dog who's going to hold the top spot if I spend incentives or if I build public goods thinking around derivatives on arbitrum will they be here and will this be our Competitive Edge in a year's time so what's an ecosystem to do a few options here uh they can King make by attempting to increase the stability for existing Blue Chips right the blue chips

are got them their early traction um they want to keep them there obviously they have other options they want to lose them because Blue Chips uh attract sort of the other smaller protocols that want to sort of feed off of and and be have a symbiotic sort of relationship with those users and liquidity this is like a GMX example where the G glp token is used across six other protocols at least in size um from options and other derivatives that feels a little bit like dow r capture so that puts a lot of stain you know bad taste in a lot of their mouths as far as like the at the ecosystem level second option being to in incubate novel Primitives by sourcing and subsidizing new protocols and amplifying their exposure but then you

get back to sort of these Blue Chips like what would they think of that um you know as as shown the concentration is extremely high um does does a leading protocol want their Dow that they're deployed on natively to go pursue another protocol in their category maybe maybe fine one or two but how many more do they actually want the third is Foster competition Again by allocating resources and facilitating trans facilitating transparency you know how how should how should you think about this over time and that is obviously extremely tough given you know such an asent industry predictions for what's next few takeaways uh first economic exploits are expensive and frequent um

the ecosystem and protocol goals aren't always aligned they mainly are um you know up only feels good but when it's down you sort of conserving resources and trying to understand what's next how does the new applo new deployments work uh retroactive airdrops Etc and then no ecosystem has really won yet just to say that everyone has options and they're still considering them few predictions ecosystems will expand subsidies to balance risk and growth a lot of the new ones like Aptos andu are already doing this for sort of you know subsidizing smart contracts deployments to get people that are in the evm ecosystems to come over to the move languages conditional permissionless ecosystems whereas uh l2s like op have sort of admitted they've

you know spent a lot of money and not sure they attracted the right users and growth and Builders um so it'll always be permissionless or at least a lot of them will be um but if you come and you want those incentives and you want those resources from sort of the the chain um or the ecosystem there's going to be rules there's going to be accountability there's going to be milestones and benchmarks for those things and then continued apps going vertical um they'll continue to do this and wield sort of power over their ecosystem so instead of ecosystems King making protocols or apps it definitely might be the reverse where the Unis Swap and OB of the world you'll see how many l2s are in their form trying to get a deployment before they're even live some of which are here yeah thank

you