"Growing DeFi with More Data and Collateral Types" - Sergey Nazarov

nazarro from chain link and Sergey is going to be talking about growing defi with a lot more data and collateral types and uh there's a lot of things that will be covered in this one so uh what I'll do is I'll welcome Sergey here and let him take it from here so Sergey welcome and uh I'll let you kick off with your presentation great thank you cark for having me at the event um yep hi everybody so we're going to be talking today about how def5 is able to grow because it has access to more data and various collateral types and some of the kind of economics and market dynamics that are driving its growth I think one of the first useful things for us to consider is exactly how the architecture or the underlying systems powering defi are evolving and I think what we're actually seeing is a shift towards a more service oriented uh kind of microservices type architecture for defi daps which many people call composability and the ability to join a lot of these applications together with with either other defi protocols as services or data sources provided through oracles like chain link as Services right and I I think maybe the first basis on which we should we should think about this is what happened in the web world or even in the enterprise software world when you moved from more monolithic singular kind of closed systems to Services libraries and all these kind of systems that could be composed together into various increasing useful configurations right and I I think the shift was kind of dramatic I think this is this this shift from we have these closed um monolithic singular kind of applications that can't interact with other systems at all or easily to a service oriented architecture where people built reusable services and they utilized those reusable services in various configurations in the web world that really led to a massive acceleration at the speed at the speed and quality for with which people are able to build things B basically you started to see you know things like uber various iPhone apps all kinds of various use cases being built because people were able to compose reusable services in a secure way they were able to improve those Services if they really like them and put them back out to people and it actually created this rapid rapid iteration where you had something like uber basically combining a GPS API combining a SMS API to message users and a payments API to pay users right and and that rapid iteration and the ability to combine reusable services in a service oriented architecture is is what I think led to a lot of the iteration and acceleration in in the web world and why a lot of the use cases we have today are so well polished and why many of them might even exist because of the resource constrained environment those were built in right by by not so large teams that then got something right and were able to expand it I I think we're seeing a kind of similar evolution in defi where you had more monolithic smart contracts that were more singular in their focus and they were more closed and kind of the code was much greater in its size and its complexity and that created essentially two security issues so I think the security Dynamics around something like the Dow and writing big contracts that kind of can have failure points because they're so big because they're so monolithic and singular and and packed so much to them was partly for a security reason initially traded away right for smaller contracts that can be composed in more secure ways and now what you have is you have protocols that are written in these more secure ways where you have many separate contracts interacting properly but people have been able to take the pieces of their defi protocols like the synthetics protocol the a protocol the urine protocol and they've been able to combine those pieces in into more useful things than any of them would have built individually right so you you you you have an ecosystem now of composable essentially onchain services in the form of contracts that can be composed in a very secure way because the the the scope and the code and and and the control of any one contract is relatively small but they all do something useful and and this this is where our space is really starting to mimic the the speed of evolution that you would see in the in the web World actually the thing the thing that I've SE seen over over the years of building smart contracts is that you could think about the evolution of smart contracts also in this sense of how do I compose different pieces together to quickly build something really secure and useful right if initially you started at Bitcoin you had One Singular system that basically enabled Bitcoin and multisig and that was what the world was like and that's what you could do right then you move to protocol smart contracts in about 2014 where you had these protocols these blockchains which you had to do a lot of work to put any new piece into and the pieces didn't always interact at all right and that was really the problem that ethereum solved for a lot of us it basically solved this problem where I have this monolithic system it has these pieces that can interact and now I have a way to write these individual small pieces of code like the erc20 standard and some Dow voting schemes and certain multi-c private key signing schemes and others and all of those then became their own separate kind of services and more and more security audited contracts that then began to get composed into the realm of token sales and DOW voting and certain private key signing things right and that was a huge leap forward for the space because people could build those individual pieces and compose them together in the configuration of I have a token I have a voting Dynamic and I have a multi-c dynamic and I have these kind of two or three Dynamics getting composed properly I I think that the next evolution of of smart contracts is actually a combination of the composability and the ability to write singular contracts with very specific goals on systems like ethereum and then for there to be other services and other capabilities represented on chain for use by those smart contracts effectively expanding the Universe of what they could do and what I'm actually seeing is is a is a rapid kind of composition of the offchain resources we're able to make accessible to contracts as well as what those contracts end up doing with that price data or that weather data that creates its own kind of contract and service that other protocols can use as collateral as a resource to to manage risk as B basically has another service right so now you have Services being shung together to create better services that then other people can build on top of and not only build uis and and various interfaces to utilize the value that the service provides but actually more services that then interact with with the ones that have already been composed basically and and this is really the very exciting and hopeful Dynamic that we're starting to see evolve in Defi and it's it's kind of the dynamic that I think is driving is a lot of this growth because you have a few really worldclass teams like synthetics a UR Nexus mutual and and a number of others that are building these kind of building blocks but doing it in a way that it's not a closed monolithic singular system just for their use it's meant to include value and interaction from other other systems and they've they've amazingly enough been been able to do that in a way that for example synthetic sa urine all interact in a meaningful way right and I think that's one of the large driving factors of of defi is that you have people finally building good focused products that are then reusable by other teams so that's that's one of the key um key dynamics that I'm that I'm seeing now in in this Dynamic there is also a useful set of components that are not inherently on chain they are not um a token contract to generate a token Ledger and manage it they are not a dow voting scheme you know they are not a multi- process there are there are a lot of resources out there that are about things like data that are about things like computations that blockchains can't or don't want to do for privacy or scalability reasons that are about payment access to payment systems that use cases want access to access to various data that basically use cases need whether that's election results price data weather data for insurance whatever whatever it might be and this is really the category of kind of services and capabilities that we put onchain in singular individual units that then get composed together with these other D5 protocols which then create a lot of useful value around the data that we provide and Ergo make another service that's even more useful for others to consume and interact with right so that's that's kind of the exciting stage I feel that the ecosystem and is at and that both the dii protocols that make certain kind of contracts that act as service for other contracts and we as the provider of certain offchain resources um are are enabling and driving forward now the other exciting Dynamic that I'm seeing is the fact that we are now driving forward the existence of multiple types of contracts if we started out by providing Market data from tens of different data providers whether they're running their own chain link node or selling data just directly into the link Network you basically had aggregations of data that were sold to a multitude of defi contracts which enabled them to quickly and securely launch quickly and securely launch more markets and we you know on a weekly basis now are able to to enable the launching of new markets new useful defi smart contracts but this year we've basically seen an expansion into other categories of contracts and what we notice is that as we put new categories of data new categories of offchain resources essentially new services that can be consumed on chain through a contract in the form of an oracle contract in the form of a chain link kind of representation of an offchain resource we see additional use cases kind of coming into existence and getting better right so one of these categories is gaming um in that category you have Randomness you have sports scores and you have various various other things people basically want to bet on and and game around as well as the generation of nfts which interestingly enough then get popped back into def as collateral and so you're see what what we've basically seen is a pattern where as we Pro as as dii protocols provide more services and as we provide more services or or more individual contracts that can act as services for other contracts you you see a kind of an explosion of of use cases and you see a rapid Improvement of the existing offerings to the point where they become more attractive and useful for users right and in offchain in in in gaming you see the provision of Randomness offchain gaming data for sports events and things like that we we've also seen a lot of improvement in the insurance industry like I'm very very proud to say that we have crop insurance live using weather data fed in by chain links into a system like Aral which is able to determine the weather and automatically pay out contractual outcomes to Farmers in both developed economies and Emerging Markets where they wouldn't traditionally have insurance and a lot of that is possible because we're able to provide essentially an onchain service in the form of a chain link that provides data or in the form of a chain link Network that Aggregates data to a sufficiently high level of security to trigger a certain amount of value another category that recently um arrived using the chain link network is proofs of Reserve so proofs of Reserve uh proof of reserves basically shows you that an underlying asset is there and that it's value has been retained and proves that to defi protocols and what that does is it allows the use of that asset whether that's Bitcoin or real estate or gold or bank account or whatever that asset is in the use of defi protocols so not only does data generate more defi markets it generates new collateral by enabling things like nfts and insurance which can then become their own units of value that act as collateral within defi right so you you enable the ability for defi to function you enable the ability for more collateral to function within while you're also enabling useful new use cases like gaming and insurance and then there's a whole large world of of collateral that isn't proven or understood by defi protocols because they can't know its stability or or or solvency and once you prove that then that world of collateral can make its way into defi products both diversifying away their risk and essentially growing the value secured by defi we also see a number of other um categories of data for example recently we've we've been going in the direction of even putting election data on chain so we're now able to put election data for all 50 states and the president presidential election on chain in a very reliable manner from uh the the data source being the Associated Press and that I think will also enable new use cases because once basically once a service is available people begin to utilize it and they begin to compose it in all kinds of interesting ways then the real question is how did they compose it into their protocol such that that protocol creates a new highly valuable service that other protocols then use the way that a and synthetics and urine interoperate right now to create large large amounts of value right and so I so this is really I think the world that we're going to see more and more of this type of graph is really going to grow in my opinion pretty rapidly because teams have now made it an Express goal to build individual contracts that other contracts can utilize as a service and we have made it our Express goal to enable the protocols the defi protocols that are taking this approach with the best possible data the best possible Randomness the best possible proof of collateral so that their service that they generate is actually highly reliable highly useful to both their immediate consumers of that service whether that's a derivatives service through synthetics or a Lending Service through a and also interest inly useful to other contracts on the back end which then take the the functionality and repackage it into their contract and their interface in new and exciting ways which at at the end of the day is is really the design pattern that you see in the acceleration of the web world and that's really the I think the big exciting point that when people look at composability or they look at what's happening in defire why is it moving so quickly why is it getting built so quickly or or is it real or not real I I think it's extremely real and I think it's accelerating because there are smart people building individual pieces that fit together very well and they're building it in a way that maintains the security of each individual piece allowing them to be used and secure large amounts of value without anyone having to reinvent them and and that's a very powerful um powerful Dynamic that historically has led to a lot of new use cases and a a rapid iteration and Improvement of existing use cases now in our case once again what what we do is we essentially provide a category of these onchain um contracts which essentially act as services that enable people to access in our case offchain systems and offchain systems are important because in the case of defi for example you need price data and it's a lot of defi is driven by Price data because it defines what's happening with the underlying assets that defi is essentially securing and so you need an accurate picture of reality you need a definitive TR truth about price data in order to know what's happening with assets right we generate this you know as a more detailed example in two ways one way is that data providers themselves run their own chain link node and they're able to at the origin sign their own data and send that signed data into systems like ethereum which then allows it to be aggregated in our aggregation contracts so that you have a measure of decentralization across nodes and data providers and you have a well aggregated a highly reliable source of truth that's either signed by the origin because the data provider is willing and able to run a new piece of their infrastructure because they see enough value in contracts or in some cases you have data providers that aren't yet ready to sign up for new infrastructure or or build an entirely or run an entirely new piece of of of their stack in relation to blockchains and the the the beauty of the chain link Network I think is that it's highly flexible so it can both enable people to sign their data and put it into aggregation contracts and send it on to the relevant contracts and if a data provider as many data providers you know if a data provider doesn't want to run a new piece of infrastructure they can simply sell their data with zero changes to their infrastructure immensely efficiently and quickly um many times the chain link Network can even just pull data from their apis without them having to do anything and this enables a large Universe of data uh in this case price data that's accurate and re and reflective of reality to make defi products that are as good in many cases better than CI products right which is which is where some of the value will flow to to def5 from like we said earlier we're seeing this same Oracle Network model get composed into various categories of inputs various categories of services basically so in this case we are composing together an oracle Network that gets rainfall data temperature data various weather data and forwards that unto other services like the rainfall evaluator contract the temperature evaluator and basically insurance contracts that evaluate outcomes and can actually be really used by other insurance protocols if they really wanted to right and and and the fascinating thing is that even if the initial implementation of chain link was was was very and is very good at providing aggregated highly proven highly secure price data chain link is really a flexible um kind of network of nodes and a framework to compose oracles around any topic around any piece of data that that really comes in the form of structured data so an API and in this case what you're seeing is the ability to to compose an oracle Network around rainfall and weather which I'm very proud to say enables crop insurance in places in the world where they would have probably still had to wait another decade to get crop insurance and therefore they can manage their risk properly and and and and benefit from Smart contracts in a very specific way the second fascinating thing is then the value that systems like arbo secure which as I understand is now past 250 million in Risk value and and and set to surpass over a billion next year that value can eventually be turned into units of of collateral that make their way into defi so not only is there an enablement of defi by by providing data but there as you enable more Val to be stored in various contracts on chain people find creative ways to turn that value into collateral for Defi and therefore both diversify defi risk which is extremely important and going to be even more important the way the global economy is going and also basically provide a defi provides a market for um for those contracts to sell their value into or to sell the insurance cash flows or whatever they create right so this is another fascinating Dynamic where you have protocols that weren't about making Financial products they were about making insurance or nfts but now the value they generate or the value they put on chain in the form of a contract even if even if it's not a token right if there's no if there's no token associated with that there's just a contract with insurance related value or nft related value you can now use that within the defi ecosystem to diversify away risk uh in in in very in very useful ways and I think Diversified kind of collections of assets the way the global economy is going is going to be quite quite an attractive proposition this is you know kind of just another example of this where you basically see something called chain link vrf which is our approvable onchain Randomness solution uh providing Randomness to a number of different um gaming contracts and that gives those gaming contracts assurances that they're operating properly gives their user assurance that they're actually fraud proof in that they use a source of Randomness that's that's verifiable and once again that accelerates gaming when you accelerate gaming you accelerate the value people place into gaming and when you accelerate the value people plac into gaming you have more value on chain to eventually be turned into collateral for def5 so that's that's another Dynamic the most recent Dynamic we've seen and that we're really excited about as well is the ability to create proofs of Reserve so when you create a proof of Reserve you basically prove that an asset is actually somewhere and that an asset is actually of a certain value or is solvent or is secure or is in the state that The Entity or the contract relying on it believes it to be in right so you in fact if you had a system like proof of reserve the 2008 financial crisis probably wouldn't have happened or even if it did it would have been softened by probably an order of magnitude right because you would have just been able to prove the status or the real world um kind of value underlying value of mortgages right so proof of Reserve even though it's starting out by proving the value of or the amount of Bitcoin locked up in a wrapped Bitcoin contract on ethereum providing automated Assurance to defi protocols that the wrapped Bitcoin that they're relying on is actually backed by real Bitcoin on the Bitcoin chain proof of Reserve is really extensible to a multitude of other forms of collateral um basically any piece of collateral that you can get definitive proof about from a data source whether the data source is a blockchain or whether it's offchain and this once again shows how Oracle networks when composed can also they can forward in data but they can also act to generate definitive truth or to generate certain proof or data about the world in ways that's that that are very useful for smart contracts right and and once again I think what this all leads to is better and better composable contracts that that are that are small security audited and interoperable with other contracts on purpose those contracts essentially act as libraries building blocks Services you know whatever you want to call them them that then get composed successfully with other teams contracts and soon enough you have an ecosystem where people can build truly amazing things without you know even having to write many of the contracts or worry about the data source inputs or the proof of an of of an underlying assets reliability or the randomness that's triggering a contract because that is is I think what's been holding up our space for some time is that people were forced to have to build a Randomness generator and a way to get data on chain uh very securely which is extremely complicated in its own set of computer science problems and then they were building a system where they didn't have proof of the underlying assets and whether those assets actually were solvent or in existence and so you know there wasn't the level of proof that they wanted even though the contract was providing proof about State changes and so now once you have a world where you have really smart focused team teams building the building blocks and teams like chain link providing them with various services that they need to be fit into those building blocks in order for them to work properly I think you you see a rapid acceleration and I I think that this is really the the dynamic that that that is going to drive us into into mainstream adoption and rapidly improve use cases to the degree that everyone wants to use them I I think the challenge for for defi is really at this point twofold it's converting the rest of crypto as a format for for for Value into defi contracts which I think is is going to happen and then the the the second Dynamic is how do you cross the chasm to the early majority right how do you cross the chasm to the users that want to utilize defi protocols but they never heard of it or they might not even know they're utilizing D5 protocols right so just the average user and I think what's going to happen is that once there's enough building blocks that are secure enough and able to be connected enough uh connected together in a in a in a usable enough way there will be all kinds of applications like Banks uh fintex Robin Hood any number of other applications that end up providing access to defi protocols because these people don't want to build any of this just like they don't want to build Bitcoin right their users say I want Bitcoin they give the users Bitcoin their users say I want yield from defi because it's 4% and the yield that I have in a bank account is is 0.1% % all the interfaces on the web will just say sure I'll give you whatever you want here we go here's the defi protocol you wanted um you get 4% now all you got to do is hit a button convert dollars to stable coins congratulations you're getting 4% by the way it was actually 8% but you know we keep 4% because we're whoever and so it's a great deal for everybody as as long as defi smart contracts continue to function properly and grow in value and collateral and returns and all these things for which there's a kind of virtuous cycle right the the movement to the early majority I think is is now is going to be underp by the building blocks and the D5 protocols and the services that people are now starting to compose and polish into highly secure highly reliable collections of services that essentially Define Defi and Define the value of what defi will provide not only to the crypto Community but to the mainstream which is which if the crypto Community is worth 360 billion right and defi only holds two to five 25 5 to 3% I mean it's a no-brainer that you would have your cryptocurrency in a defi protocol gaining some kind of interest rate right and the the rails to do that are just getting accelerated and accelerated in in a really really rapid Pace but in terms of the mainstream the amount of value that's out there to make its way both into the cryptocurrency format and into defi is is orders of magnitude larger than what we're dealing with today and in the environment that we're going towards where you have low interest rates and low transparency for assets and the low transparency becomes a concern because counterparty risk is what people care about when there are solvency issues and solvency issues occur when there are Global Market booms and busts namely busts I I think you're you're you're you're basically actually going to see Defi and smart contract insurance and to a degree blockchain fraud proof blockchain gaming driving a lot of the value that goes onto blockchains the way that tokens previously drove a lot of the value going on because now you'll be imagine if you can offer people a 4% interest interest rate through a crypto bank account and all they have to do is click a button that says yes like that's coming and people don't care if they get their 4% from a crypt backed Crypt based defi protocol powered thing or from a bank they could care less they don't know and so once all of that gets composed in the right way that's those rails are going to move this value into the crypto format but importantly I think it'll move the value into the crypto format not for people to gain tokens but for people to experience and benefit from Defi and so that's I think the really powerful thing is that tokens are one thing and that's great and people can have all kinds of ideas about tokens and and and and what they're worth or not worth and that's that's kind of their choice but if you're able to offer the Global Financial Market 4% % or 2% interest and yield in environments where they can hardly get 0.1 or 0.2% and you're able to make that secure and transparent and regulated and reliable um I mean that's what the Global Financial system is based on right that's that's what drives a lot of activity and so the the the next question then is how do we continue to enable this this is kind of the body of work we're engaged in where as more smart contracts appear people have more and more reasons to consume them and that means there's a larger market for data providers to sell their data and that's what we enable likewise as more people use contracts they pay more fees those fees partly go to pay for security of systems like chain link and data data providers and so on and eventually you arrive at a place where you have a very large um user group that's paying into daps that are paying for data and security and basically services in the form of onchain contracts and both of these Cycles are what we're really accelerating we're accelerating the creation of new smart contracts by providing the services and the data and all the capabilities that people need on chain that's generating more usage which is generating more demand for the data and the services and demand for the security of that data and those Services which once again as we put more of it on chain we notice that people build even more contracts so it's a very virtuous cycle and this is kind of the body of work that we're engaged in uh if if you find this body of work interesting we're we're thrilled to work with you we have two formats in which we very efficiently and effectively work with people now on a kind of remote open- Source basis one of them is the chain link grants program of which there are multiple categories about both integrating more data building developer tools to build contracts better providing security and providing blockchain Integrations um and that's the grant program if you have a Dev shop or a team that would like to work on these problems you know in an open source kind of remote work way we have a um a growing and and and and increasingly more popular grants program that enables these these cycles and these services to go on chain and then finally if if you find this problem to be compelling and something that you find interesting and exciting and you see the value in accelerating the next evolution in smart contracts into this into this useful state of contracts that are about providing value to the average person because of what the contract actually does and what we enable it to do then then we're we're we're thrilled to speak with you and and to find a good way to collaborate you know we're a remote team open source and really in our team in our team Dynamic kind of the best ideas tend to win and so we we we encourage you to to talk to us if you if you find this problem interesting and important awesome um thank you so much Sergey that was a great talk I am uh I think this is a really good way to uh really kind of paint the picture on how big the opportunity is um that we can use crypto to leverage and and kind of offer a lot more to the rest of the world so uh it was really good to kind of see that perspective on kind of how you get to enable that as developers especially the ones watching this and uh and now you can also use chain link to sort of get this um there are a handful of questions that we've had from our chat but unfortunately we are out of time and we're running a bit over so what I'll do is I'll ask a quick one and then I'll kind of relay the remaining ones to you and the rest of the team um so the quick question that I have is just in general a lot of people are curious about what is currently the cost of U updating and uh and sort of adding information to uh chain link uh and sort of if it's too high uh how are you making it easy for people to sort of not bear that cost as we add more and more data sources yeah sure so there's there there's there's kind of two Dynamics here the first Dynamic is what level of security is somebody comfortable with and what we do right now is we perform onchain aggregation because that's the most secure method for doing this and that's why we're able to secure over four billion on production without any issues and that's why we have over you know you know billions more publicly announced and and and and kind of more up coming at some point right it's because we prioritize security because we work with people like Jes who used to be the chief scientist of RSA today we recently announced that our new VP of engineering is Ben Chan who was previously the CTO of bito and built you know the multi- solution that that allows the movement of about 15% of all Bitcoin right so the the the first priority is security and that's why onchain aggregation is the first step um onchain aggregation can be costly but realistically if you're securing billions of dollars in value you should seek to pay for security if if you're if you're able to get it and the the key point is whether you're able to get it and with chain link you're able to get it and believe me from the Oracle failures that have been out there and that we've seen you absolutely want it right the the the second answer to that is the scalability of either the onchain environment or the offchain environment so since 2019 we've had something called threshold signatures and offchain Reporting which you know we're currently working through now is in is in certain phases of Security review and is is is basically the scalable version of chain link that is the expanded scalable version that does the aggregation in a offchain manner but the important Nuance is that you actually want to maintain an onchain aggregation as a possible fallback at a at a lower rate of frequency and you also want an offchain aggregation that is highly scalable and so the highly scalable offchain aggregation is something we're working on now and and that will enable um probably an order of magnitude increase in the amount of networks we can laar in you know the cost of doing that and and any number of other things and that's basically born out of the need that even though we have many many networks and we're constantly launching them for users to launch new products we uh have a large amount of demand for a lot of different price pairs right so I I think I think the important Dynamic here is is firstly that you actually want to prioritize security more than anything else with a system like this and there's a reason that offchain aggregation provides that security uh I mean onchain aggregation provides that security offchain aggregation also provides that security but it's not something you slap together in the weekend and you kind of cross your fingers the the people that are doing that by like you know forking some chain and trying to make it into an oracle or like trying to cook up something themselves with them and their one web developer you know we've spent um I think over a million dollars at this point on on on security audits I think we we work with the academic research community on a regular basis these are not simple problems right so we've been kind of baking OCR and threshold signatures for some time and it's kind of getting to the stage where it's in those stages of security audit and review and it's getting properly assessed and figured out right um as as whether that's secured I think we're kind of at at that that stage that will will drop the the costs really significantly and also just so you understand the costs are not completely passed on to our users our users all pay a certain portion of a fee that is then pooled for one Oracle Network so any one user isn't paying all of these costs it's shared across those users and also it's it's it's partly you know paid for by a kind of Link reward that pays the node operators to properly operate the network as the user fees grow and they are they are pretty much rapidly growing at this point so I I do imagine we be at a point where with the scalability of OCR the scalability improvements of various chain and the growth in users will eventually be at a point where user fees completely pay for specific Oracle networks that are sustaining and feeding the security of those users defi protocols but the point is that they're pulling the we're pulling the security and the fees around a single Network and a single piece of data that's getting put on chain and and that clar that kind of security is once again the goal right because your your your goal is not just to build some system that can do something scalably but in an insecure way your goal is to build the most secure system you can and gradually scale it in a secure way while maintaining clear economics that work which they clearly do because the op Network operates properly both in environments where gas fees are high and low so the chain link Network operates even in environments where the gas Hees are high because the economics of the sheen link Network work properly which is once again to viewed not from a purely economic point of view but from a point of view of security great no uh that was a very insightful answer um and hope that the couple people that asked questions along these lines uh have a bit more clarity now so with that I want to thank you Sergey for uh giving us uh your time today and uh kind of going in depth on how big the opportunity in crypto is and how we can make a lot of data sources integrate with the rest of the