The Future of Decentralized Liquidity Aggregation

Transcript

hello everyone i hope uh it's not too early for you after all those parties we had yesterday um so i'm munir founder of paraswap i would like to talk to you today about the future of um liquidity decentralized irrigation sorry and paraswap we are a liquidity aggregator on the ethereum blockchain and other sidechains as well i would like to start by defining layers in defy in web3 in general so we have the first layer if we ignore the infrastructure layer that is the blockchain itself we have the what you call the assets

layer which are like eth and years 20 tokens and to use those tokens for buy-in selling for lending and borrowing and so on we have what you call the capital aggregators and those are either decentralized exchanges and ending protocols and so on and to put that in the hands of end users whether those are like people or the protocols themselves or adapts and wallets and trading bots and so on we have what you call the service aggregators and those are either uis like an argent serious zapper and so on or it can be other services like verisoft for instance there is an api that will allow

other dapps and wallets or maybe even other protocols like an away for instance to access or manipulate those tokens that's the service aggregators layer what is an aggregator and here we're talking more about the aggregation theory and it has three characteristics well first aggregators have direct relationships with users and that's something that came especially from web 2.0 companies or like technologies for instance we don't need anyone to access google or facebook or any web 2.0 eyes we don't need anyone to access any web 3.0 company we just go to

the website so it's served through the internet um there is zero marginal cost there is zero marginal cost yeah better zero marginal cost for serving users the content is served on the internet so it doesn't cost more to serve a user in portugal than someone in australia is exactly the same cost and that's also something that is related to internet and web 2.0 companies uh which is not the case for instance in e-commerce like in amazon uh has a marginal cost for serving our end users which which is something we don't have in web in general whether a 2.0 or 3.0 um and the last thing is what we call

this demand-driven multi-sided network with decreasing acquisition costs a bit too long to say to talk about network effects which means that at a certain critical mass uh user acquisition becomes the cost of user acquisition becomes decreasing an example will be a social network like facebook at some points where we reach a critical mass in a number of users where users have their friends and their friends bring other friends and so on the growth becomes organic and natural so it does cost less for the platform or the aggregator to acquire more users and

that's something also we can see in defy like for instance say take uniswap in the beginning you just have needed liquidity providers to have traders and at some point and after reaching a critical mass the growth became organic because every party serves the liquidity providers they have the fees that are their main incentive and the traders who need to swap or exchange value they have sufficient liquidity and sufficient market depth we can classify aggregators in different levels and here we try to map some of the defy protocols and see how they fit or if they fit in the first place

so the first level is what you call supply acquisition those are terms brought by a guy called ben thompson that he's making a lot of research on the aggregation theory so that means that the platform or the aggregator owns the supply side an example would be netflix netflix owns the movies and produces the movies so it comes at a very high cost but the thing is by producing quality movies and they bring sufficient mass of users which gonna finance netflix which gonna allow them to bring new movies or also to produce produce higher quality movies and it becomes a

virtuous cycle an example i was thinking about yesterday is olympus with the om where we are introducing a new defy 2.0 whatever buzzword recently where it's no longer about doing liquidity mining programs but it's more about issuing bonds and selling the token at a discount in the future in exchange for kpis or performance kpis which means here the tao owns the supply and uh but it comes at a cost because we are sending the tokens at a discount so i think i'm not sure hundred percent but it seems to me that olympus oh modex whatever defy 2.0 looks a little bit like this supply

acquisition level one aggregators then on a level two um we have what you call supply transaction costs which are aggregators that don't own the supply but that supply comes at a cost and an example will be uber where they don't hold the driver but there is a high cost of an onboarding and maybe whatever liquidity mining programs where the protocols don't own the supply because it is still coming from liquidity providers and these guys are incentivized by those liquidity mining programs those farming programs and and this can be significantly costly for the protocol for better offers depends on how it's managed of course

and the level three is what you call zero supply costs which means there is no ownership of no supply no demand and no users um and it's here we're talking about social networks typically where the growth is purely organic and the aggregators themselves just need to provide the best possible products and the critical mass will be reached as the platform grows and here is the typical web 3.0 apps or d5 devs like uni swap away balancer in others um in aggregation theory we talk also about super aggregators who are like level three but have they have a third party which we call advertisers and those are the customers in this case

because those are the guys who are paying the platform so that platform will bring supply and demand and build this network effects but who's going to pay for that aggregator well in case of facebook and google in all the gaffa well those are the advertisers who are the main source of revenue i tried also to find uh is there a web 3.0 or a d5 protocol that has this property and the only thing i thought about is flashlights for for d5 pro cost like avi for instance and it's not the main revenue so it doesn't really look the same but it's interesting because if we have sufficient network effects in ave uh with uh lenders and borrowers we can make

sufficient like higher flash flashlights so the higher uh the bigger ave the bigger the amount of flash launch we can take the use case now the initial use case of flashlights is not really about whatever we know about flash launch it's about like uh not i mean protection against liquidations and stuff like that but it's used in other use cases and it's been monetized there is a fee that is taken from flashlights whether this will evolve in the future or not but that's the only parallel i saw in in d5 to talk about aggregation quickly in financial markets it exists in multiple forms you can think revolut etoro and others but also in uh what i would like to highlight here is those what you call aggregators alright we can call them like technical

aggregators which are sors and those are look a lot like what we do in parashop they will execute trade on multiple exchanges and try to maximize the price and reduce slippage and maximize the market debts the thing is here is that to use those those software uh it is we cannot use them directly so it defies the first property of aggregators which is this direct relationship with users you have to be a broker and i want to use a store i have to go to a broker and as a user i don't even know if there is a sort those are things that are highly technical in the financial markets or like traditional finance um it exists also in c5 in multiple forms so it's like users so you can call coinbase or sell cells use

as an aggregator as well but also those are those aggregations that look like trade phi uh like s fox tagome and others uh who allow this what you call best execution or sors the thing is anyone can access those tools just need to do like some onboarding but it's not close to a certain category of brokers anyone can have access to those troops but in d5 i mean we can have also the same thing but it's just like a web browser away from accessing of course in paraswap in any other aggregator we can just like go to a website or plug in the api and we can do some very complex transactions and and maximizing

the price in a very optimal way without needing any intermediary so i would say that a parasaur io zapper exchange or tapper zapper ui or xerion are aggregators but the question is how do we compare to d5 with c5 what are the differences c5 still has the advantage of giving more options which are for instance i can exchange btc with ripple with whatever which is not possible today in d5 like real btc not a synthetic version of it but the thing is with with c5 aggregators or like what she technically today called aggregators is that

there is no deterministic way of making a full transaction in as we do it in d5 for instance i have btc a cell btc to usbt the thing is if this is executed in finance and coinbase if the trade fails on finance there is no way or not deterministic way to revert that transaction and get first amount that i have spent uh and especially if well in very high volatile moments like a cell btc i receive less usbt i'm not okay with that i hunt i want to revert the market moved on the other direction i mean i have less btc than i started with in the first place which is not the case in defy but in the same blockchain so i have wbtc i'm saying usbc

the evm guarantees guarantees me that i will get whatever amounts i'm expecting or i have a cryptographic guarantee that this transaction will reverse but in the same thing would arguably apply to layer twos uh as long as there is a smart contract platform that runs on chain that executes the trades and that's the case for all evm chains uh even non-ibm chains even other blockchains like solar for instance but maybe not the case in some hybrid layer tools like a stock for instance where everything happens off-chain and only the sediments happens on chain which are like set of proofs that are pushed push unchained so a stark wave system looks a lot like c5 and that's

something we can see in uidx for instance so now the question is what about cross chain um while crossing they bring more capabilities so things start to look like cifi or even trade phi but now we are breaking composability and by breaking composability we are breaking atomicity so that problem we had with the c5 aggregators will now be happening also in cross-chain platforms um so we are having the same challenges in c like c5 and shot phi and the same techniques of risk management obviously these guys have thought about those issues will apply here and i think they will apply in a better way because the innovation

is very interesting but i think also this become much more complex the challenge is to deliver the same security level um without sacrificing decentralization and i think this is happening also already um so now the question knowing all of that the question is will what we currently call aggregators are they really aggregators and if they are will they stay aggregators my thinking i don't think so because uh if you take for instance paraswap calls like dex aggregators aggregators but in my opinion those are infrastructure products the only the aggregator's side is the ui like powersoft.io or

uniswap.exchange this is an aggregator with those three part properties we saw before but the product itself or the protocol is a pure infrastructure product and i think there will be a separation between uis that are like check all the marks for aggregators and those are uis of any protocol or any protocols or also the uis of arjun zerion zapper which will be representing defy to end users at the end at the end of the game and maybe uh tax aggregators will become much more complex with multiple components one that's making aggregation on a single blockchain and others that make an aggregation on different blockchains and

those in my opinion will be pure infrastructure products so that's all i had thanks a lot for your attention and happy to answer any questions