Fireside Chat on DeFi's Institutional Adoption with Sergey Nazarov and Kartik Talwar

Right, [Music] we are ready for our next fireside chat. And for this one, I'd like to bring on Sergey, the co-founder of Chainlink to uh talk about all things finance and institutional and Oracles. But please give him a big round of applause. [Applause]

Welcome back. Yeah, great to be here. You you folks make the nicest events. I mean, this view is awesome. You're really good at this.

Well, thank you. Um the the best part about all these things is that um at least for the audience, we um we usually bring on speakers that we have known for many years, which makes a lot of these conversations a lot simpler and and just easier. So, uh this ends up being a a chat that we're having that you get to watch in the audience. And uh the cool thing thing about I think this session would be that um we just did the last pragma four weeks ago, five weeks ago uh in Ken and um I actually want to continue some of the things that we talked about. We obviously won't we will give people the context but um I think it's like a good time to uh keep financial institutions and Ethereum on the the spotlight and uh we'll sort of renew and uh continue the ideas.

U the first thing I want to start off with is actually asking you from kind of your point of view how have you seen the or Oracle landscape evolve 10 years ago what we thought oracles were were entirely different and kind of very purpose-built but that's no longer the case. So how do you kind of think about oracles as a category and and do you still use the same definition?

Yes, oracles have expanded as to what they cover. Um, oracles are basically cross-chain bridges, now data oracles, now we have identity oracles. Soon we're going to be releasing something called the chain link runtime environment, which is essentially an orchestration oracle, which manages smart contracts on multiple chains and multiple oracles together with a single piece of code defining uh this kind of more advanced smart contract that operates across chains and across or uh various types of oracles. So oracles re really have have expanded in my opinion you know far beyond data. Basically we have AI oracles where we run multiple AI models in different nodes and then we come to consensus on the results of the different AI models to do things like scrape PDFs or scrape um unstructured data like just written data and turn it into structured data that can then be used to trigger smart contracts.

Um at at this point, Chainlink is the only platform Oracle platform that does all of those things in one system. And that's why we've now moved on to create this uh essentially orchestration oracle and environment called CRA because when you build these more advanced smart contracts where you need uh a contract, you need a data oracle, you need a crosschain oracle for bridging, you need an identity oracle, maybe you need an AI oracle for risk management, the the complexity of the contract has uh massively exploded, right? So the the the thing that I think has changed by definition is is not as much oracles as it is uh smart contracts. So the definition of a smart contract circuit 2015 161 17 would have been a single contract on a single chain and everyone just wrote these really big smart contracts like uh like the DAO you know everyone's favorite smart contract from the good old days that got hacked and it was because a hu was a huge contract right it was this huge block of code and then everyone said well okay that's a really bad idea I'm never writing a huge contract again and then smart contracts actually became a collection of contracts that all worked with each other, right? So, you had like a a voting contract and a multi-IG contract and, you know, contracts that did very specific things and you you created what's known as a separation of concerns, right?

So, you mitigated the risks by by making the smart contract itself built of these different services and pieces which which brought the industry towards a serviceoriented architecture. Then circuit 201819 chain link introduced oracles and that gave you know rise to defy and defi's boom that's actually the point defi in my opinion is the point when a smart contracts definition extended to include oracles right so that's fair because you needed external information you're not going to have a defi smart contract without a few contracts you know like a lending the l the lending market contract the collateral the liquidation conditions the multi-IG Then you have the oracle, right? And if you don't have the oracle, you don't have a DeFi system or protocol. So that's when that definition extended. Then circa 202122, you start to see real crosschain, right?

Because there's liquidity on L2s, there's liquidity on other chains. So now you need uh another type of oracle, right? Now you need a crosschain oracle known as a bridge. Um now you have two oracles and you have you know five or 10 contracts on one chain and a bunch of oracle related bridging contracts on other chains. Now circuit 23 24 25 especially if you look at institutional the institutional world the definition of a smart contract has expanded even more right so now it's even more contracts on one chain contracts on other chains representing you know your bridging responsibilities now you have data contracts connected to a data oracle you have identity contracts connected to an identity oracle you have risk management contracts in more and more cases connected to an AI and an AI oracle Oracle.

So once again, the definition of what a smart contract is expands and I I think where we'll where we'll end up is that you you'll have like a very good onchain transparency, onchain clearing and settlement world that can work efficiently and then you'll have a lot of off-chain processing, risk management processing, identity processing, data processing, crosschain connectivity processing, and that'll all be oracles. And so I think you end up at a world where 50% of the smart contract is onchain and 50% is is the oracles and that's what a smart contract is in a way. This kind of tracks the the agent thesis on on the web two side too and um effectively you're just giving something authority to manage information. Um where does CCIP fit in this? CCIP is AC is a protocol for running highly reliable cross-chain oracles that are also compliant with institutional requirements.

So, so all of these systems and oracles can get really complex or they can be really simple, right? So, I can say, "Hey, I make a data oracle." Great. What does that mean? Oh, I run a server on my laptop and it pulls from one API and it gives you the data to run your lending market.

I I made a data oracle. Yeah. Okay. Great work. You made a data oracle.

Good luck finding anybody with two brain cells to up together that's going to put any money behind, you know, this this lending market with this data oracle or anybody who like understands what what what's going on. Unless you lie to them, which sometimes does happen too. People make an oracle and they say it's decentralized and it's it's actually just like, you know, three servers in the same basement, but or like three different basements all owned by, you know, the three guys who run the thing. So, so basically, you can make it really complex, right? Like for our data oracles now, there's literally hundreds of data sources.

Those get rebalanced. It's it's really complicated. It's the same thing with bridging. With bridging, you could make a single node and you could call it a bridge. And you could say, "Hey, this is going to be this going to be a blast.

I'm going to bridge you from here to there." Uh, two problems with that. You know, problem number one is it's not really secure and reliable. It's kind of the story of every bridge hack. Some guy somewhere had one or two nodes.

He said it's decentralized on the website. people trusted you know the bridge they lost the money on the bridge not so good right CCIP never lost money kind of has many many nodes you know actually has three sets of nodes 16 16 and I think seven on the third set and um so you have security which is complex and and costly to make then you need crosschain to work in a way where institutional capital can flow in to Ethereum so retail capital can grow Ethereum, you know, from half a trillion to 1 2 3 4 trillion. I can see that happening. Can retail capital grow Ethereum very far past that over the next 3 to 5 years? I'm not so sure.

I think what you really need is the institutional capital to flow in. And the institutional capital is going to flow in from all the other chains that all the institutions launch. Like even even today or yesterday, I can't even tell what day is anymore. There's some stuff going on, but like

Thursday.

It's Thursday. Okay. Then I think yesterday or or or or maybe tomorrow. I don't know. Look, it doesn't really change much actually.

Um Stripe launched said they're going to launch a chain and I think Circle launched a chain and you know, like you're just going to see that, right? You're just going to see a bunch of capital go on a bunch of chains and then the question is how does that capital flow into Ethereum? And the answer is a reliable cross-chain oracle which is the the protocol the CCIP protocol that's what it makes possible and and then you you really have to meet the conditions of the people sending the money. So the the the act of sending value in in the simple world in the simple blockchain world is you know I send you my three meme coins for your five meme coins. We all stake our meme coins and the yield goes skyh high because the price went up and it's all wonderful, right?

That's not really how this is going to continue here in this industry, right? Like there was a period where that is the way it works. It's going to work in that way like in some places, but the way it's going to work now is net new capital, institutional capital coming in and they have conditions like I need to verify identity. I need to get valuation data to verify the value of the thing that I'm sending. All these things now need to happen.

to my earlier point about the complexity of a smart contract.

So, does that imply that effectively anybody who's trying to just build a a DeFi protocol um can't really use CCIP because you have to kind of you only kind of come in there when you have compliance as a as a part of this process. No, no, it's it's built for both. So, so a si a complex thing can usually do simple things. A simple thing usually can't do really complex things. So CCIP can do CCIP can handle very simple transfers at a high level of security and you can actually change the level of security you want for different transfers but very importantly it can handle complex transfers.

I I'll give you I'll give you an example. We have a feature there called programmable token transfers where you can take a token you can append data to the token and the data can be about what the token does when it reaches the destination chain. So you can send the token, but you can append data and instructions to the token so that when you hit the target chain, you can do things like deposit, right? So send the token to from chain A to chain B, but on chain B, use the deposit function on this lending market. Yeah.

So CCIP can do the simple thing of sending the token, but CCIP can also send your token and execute the function on the target chain. Okay.

And because it has the capacity to add data along, it can it can send the token, it can execute the function on the target chain, it can also add identity data to the token. So that when your token reaches the destination chain and if that deposit function requires identity information in order to allow you to use it, now you you have the

pass that information. Yeah.

Right. You have the token. You have the identity information to allow the function to to do what it's supposed to do, which is accept your deposit. you have the ability to call the function and execute the deposit and you're good and that's it. And um by the way, you needed an identity oracle to get the identity information.

Um so all of all of those more complex things are possible. The simple things are are definitely very possible, but that clarifies. So so it's not really like an institutional specific product. It's just that they are definitely the most obvious customer, but it still works for anybody else and it is open at the end of the day.

Yeah. Yeah. It's a totally open product and and lots of DeFi uses it. Um A is completely uses CCIP. It's it's basically their cross trainin bridging system.

The the people that do a lot of technical diligence and security diligence pretty much almost always prefer CCIP unless there's some like commercial deal that somebody gave them or something. It's it's because the security of CCIP is built on the same superior security foundations that the chain link data oracles and other oracles are built on. Chainlink is in a position where it secures now at rest I think over 90 billion. And with the growth of DeFi, I can see that going into the hundreds of billions. So the the the difference between chain link and like the startup that told you they have a bridge but it's actually a single server run by two guys who need to raise money is that those two guys have very little to lose whereas chain link has a lot of you know responsibility to make sure that the the system meets certain requirements has certain levels of security and reliability and you know that's why chain link right now powers over 83% of all of Ethereum DeFi and over 64 or 65% of all DeFi globally.

It's not a coincidence. It's because the people who actually understand how things work when they look at all the failure modes of the data oracles and the bridge oracles and all these other things. Chain link just doesn't and and you can see this in third party audits like there's something called layer 2 something beat.

Yeah. Okay. I think so. I think them they did they did a review of the bridges and they found a bunch of critical vulnerabilities on a bunch of the critical on on the bridges and they found none on CCIP.

So I mean that that's just the model of chain link right it's like if

what's the other 17% that's not chain link

on on the on the data oracles

on the Ethereum side. Yeah. A lot of it is people still making internal oracles because they don't fully understand why they should or shouldn't make oracles. The vast majority of it is is actually uh Maker.

Oh, just the USD.

Just M. Yeah. Is Maker who we who we partially power but don't fully power. Um so that chart probably needs to be bigger, but um yeah, that's that's the rest of it.

Okay. Um, I feel like this is still related and I've been asking this question to uh to basically most of our speakers today. Um, what do you think about these layer 1 stable coin chains and um how do they fit into the the DeFi landscape according to uh to you?

I think more money, more cash on chain that can purchase tokenized things is always good. I I don't care if it's a central bank digital currency, which people don't like, or if it's a payments chain currency, or if it's a stable coin launched on an L2, or if it's a stable coin launched on an L1, or if it's a stable coin launched on Ethereum. I mean, I prefer to be a stable coin launched on Ethereum. Big fan of Ethereum. You know, we launched Chain Link on Ethereum.

We launched our reserves a few weeks ago on Ethereum. So,

you power 83% of all the data on it.

Yeah, exactly. So, so big big fan of Ethereum Ethereum community, but if you have a reliable crosschain system that can get the central bank digital currency into Ethereum and you have a reliable crosschain system that can get the payment coin into Ethereum and likewise for all the other forms of onchain cash, that's just more liquidity. So, so the way that people are used to viewing liquidity in Ethereum is is the liquidity on Ethereum? Is it on the L1? No.

Okay. Okay. Is it on an L2 because we have some kind of special relationship with L2s? No. If it's outside of the Ethereum, you know, uh, kind of sphere, not so good.

That's because there there isn't enough high quality crosschain connection yet into Ethereum to get all of that liquidity in there. Like the world you really want to end up at is okay whatever some chain some bank somewhere made a chain or or a payment network made a chain and they put uh hundred billion dollars right so so they uh they added a third more of the stable coin supply than there is now what's important is how does that new liquidity interact with Ethereum issued assets that's the question right But so so you're saying that fragmentation is fine as long as there's a solution to get the liquidity across chains. But that seems like a never- ending sort of a scenario because like you are you're still fragmented. You're you're just simplifying the the steps involved. We're not fragmented.

We're decentralized. Don't forget this. It's very important.

My apologies. I've I've only been in this space.

Only been here since the beginning.

Yes. We're de a we're decentralized. Don't forget, remember, we're decentralized. Number two, the internet has a bunch of fragmented, disconnected stuff until it all gets connected and it becomes quote unquote the internet, right? You launch something on AWS, I launch something on GCP.

What do you care where I launch what I care? Would you launch it? I think that the the the flaw in that argument is that IP packets are the same regardless. and and and the replication cost matters, right? So, so it is a lot easier to have a copy of a file stored in 50 places, but not the same argument you get to apply for a transaction on chain.

Um, sorry, no, no worries. If only we had a really reliable crosschain system that could guarantee to us that when we move something from over here, it ends up over there without getting lost. If only there was a secure and reliable way to do this in a very costefficient, ubiquitous way where if a form of payment appeared on one chain, it could seamlessly and at low cost with high security and reliability suddenly be able to purchase an Ethereum asset as if it was on Ethereum itself.

Is the the core research team working on this at the foundation? Is that uh

I mean the the core research team Well, I don't know. The core

don't worry.

No, no, there is a thing called 7 7683. Yeah. Oh, yeah. You have to still simplify this this hopping from the layer one and the layer two.

Yeah,

that's good. That's that makes things better for everybody.

Um

that's a good thing. Yeah,

that's a great thing. Correct. Um I guess the thing I'm kind of curious about is is it a fair statement to make that even if you have a 100 chains with a billion dollars, that is the same way to think about liquidity as a one chain with hundred billion dollars.

It's the cost, right? So like I put liquidity in a block on chain A and then you have liquidity in a block in chain B and then the asset that we both want to purchase is in a block on chain C. If it costs me the price of the asset to take my liquidity from block from block A and send it over into the into block in the block on chain C, that sucks. if it's if it's if it's at if it's near zero or as cheap as possible and it costs me almost nothing and it's simple for me to take my liquidity from chain A and utilize it to do something in chain C. What we've done is we've just interconnected all these blocks,

right? So I guess as I'm thinking about this thing, I feel like the thing I'm trying to understand is the fulfillment of a transfer makes sense in this example, right? I just have to go from A to C and I can figure out how to get there cheap, fast, whatever I'm optimizing for. But you are still presenting a user with like, hey, do you have USDY and USDX and USDT and the fragmentation is on the user level on the the cognitive overhead? And and I'm curious about is that a problem that actually hinders the growth versus looking at it from just the engineering side.

I I don't think that's I don't think that's a huge issue. We released something on production recently with Mastercard where we enabled Mastercard card payments in the off-chain Mastercard network which has like over three billion card holders to and we were able to convert the Mastercard payment over into a Ethereum blockchain transaction that was represented as a stable coin and then exchange that stable coin for uh a number of tokens on unis swap.

That's awesome.

Yeah, it is. It is awesome. And and and and and the point of all that is

when when when was this? And also like couple months as a test or like for

No, it's on production.

So who is opting in to pay

a master cards?

Like it shows up on their profile on their account.

It I mean there's it's an iframe. You could put it into your app and now you can

I see. Okay. So not for like a general card holder in their bank account. It's like if somebody's trying to accept Mastercard.

Yeah. If somebody wants to accept Mastercard payment directly on their website to sell a token, we through through the chain link system have been able through a number of oracles working together have been able to verify that the payment happened. We convert the the payment with together with some partners is converted into a stable coin. The stable coin by an oracle is then directed towards a DEX pool. It's exchanged as exchanged as a stable coin for the token in the DEX pool.

And then that oracle coordinates that the that the the the token from the Dex pool that you just bought goes back to the account which you designated as your wallet account when you did the masterard transaction. And so that's like I have a masterard. All of this can be abstracted away is the point, right? You you can show up and and the people who paid in the masterard, they don't know what stablecoin was used on the back end. They neither know nor care.

and a person can show up with whatever stable coin they have and I think there will be more and more of these uh stable coin clearing houses and and so you you will have more and more ways to abstract. I feel like this is basically now a new forex market where you just have pairs of different stable coins with each other and they're off by like a thousandth of something and you're effectively managing that liquidity because it this is what it kind of feels like somebody just settling something with a tenth of a cent difference and trying to make this work for uh for fulfillment.

Yeah. Yeah. Yeah, we we we did another transaction like this where A&Z Bank, which is the the first largest bank in New Zealand, second largest in Australia, they had a client account on an Ethereum private chain with a stable coin they made for the Australian dollar that got converted into a central bank digital currency called the E-Hong dollar, which which is also on on EVM. The E-Hong Kong dollar then went over CCIP and it was exchanged for a tokenized fund on on main on public Ethereum and then it was sent and and then the tokenized fund was sent back to to the to the bank client and and so in the beginning of that flow you had a conversion from um you know an Australian dollar stable coin to a to the to actually to a central bank digital currency which is even more interesting than another stable coin because the fund was denominated in the Hong Kong dollar, right? So you you had a client who had one currency, they wanted to buy a fund denominated in another currency and that that was all automated, right?

So all all of Yeah. And the clearing houses will automate that. There will be a bunch of stuff that automates that. At at the end of the day, what I think will happen is is everyone will just show up in their like Fidelity account. This was, by the way, Fidelity International this thing was.

And so you you go into your Fidelity account and you had all these things you could buy and now you have tokenized stuff and and the question is is the tokenized stuff on Ethereum and I think the answer to that right now looks like like a resounding very strong yes as because you're the liquidity is there or what what's the like the direct is this the first mover advantage argument?

Yeah, sort of. there's a kind of escape velocity where there are so many tools and and it's kind of like what makes JavaScript successful, right? There's so many tools, there's a critical mass of resources. Liquidity is definitely a big part of it. I think liquidity will become, you know, continue to be a part of it, but grow with crosschain connectivity.

Uh I think yeah the the the the current popularity of ERC20 ERC 3643 uh tokenization is very large and it's it's growing rapidly and um I think there's also a lot of tools there's a lot of you know DEX related tools lending market tools that can be forked so the the Ethereum community and and chain link as a part of that community I think I've done a great job and building an ecosystem of composable building blocks that now the institutional world is finally starting to appreciate and that that collection of composable building blocks doesn't exist anywhere else. A lot of the other chains are trying to build them. Some of the many of them are not building it. They're not getting them built by the community, right? Like Gnosis safe is a good example.

Chain link is another good example which is like an Ethereum community kind of member that built all this infrastructure. Gnosis built very popular and high quality multi-IG. Some of the other chains, they don't have a team like Gnosis to build their multi-IG. So their chain team has to build the chain and they have to build the multi-yc and they have to build like a bunch of other stuff. That advantage is is is going away as their communities

and that's like yeah basically slow velocity there. Um

it's a slower velocity, right? So if if if the Ethereum community is is able to to have the the the reliability and the high quality building blocks, I think if the rel if the Ethereum community can suck in as much liquidity as it can from all the institutional user bases, if there can be highquality reliable infrastructure like chain link that continues to provide reliable data, compliance, identity, you'll have this kind of critical mass of building blocks and that critical mass of building blocks will be very appealing to developers because historically when you look at the systems in web two that one they were the ones with all these building blocks, right? JavaScript, Linux, other good examples. I feel like one thing that at least in my opinion is the symptom of everybody kind of doing their own chain is the the not invented here kind of effect where they're like, "Oh, we should do this ourselves. We want to own everything ourselves."

And that's kind of how a lot of the

in the institutional world or in the web

in institutional world. I'm sure we see that

in in the web three world. It's I want to flip my token for VCON.

Yeah. Specifically in the institutional world. Um you've been talking to financial institutions for probably about a decade too. uh as Shamlick and my kind of question is um what has changed in their kind of uh viewpoint of how they think about decisions and generally speaking if you have any advice for somebody now thinking about how to bring anywhere from their liquidity to some of their process on Ethereum or just in blockchains in general? What should be the common mistakes that they should not make and things that you've seen that always kind of end up firing back?

I mean the first mistake they should all avoid religiously is using anything other than chain link. That is the first one. If you're using a

nonbiased view I see.

Just to be very clear my position is if you're using anything other than chain link to do something that chain link does you're making a huge mistake. You should immediately immediately stop what you're doing. Go read about chain link and and save yourself a lot of pain and anguish. That's that's my first best piece of advice. You know very impartial view here.

That's the first one. the the reality is that they've gone from being afraid to do it to afraid of missing the boat, which is great for us. So before because the government and the regulators, people people that don't deal with banks in the middle office and the back office, I understand why it's hard for them to understand what the hell is going on there. Like before I started dealing with these people, it was hard for me to understand how does a bank work? What is a back office?

What is a middle office? What is a compliance officer? Why does this matter? Why does that matter? I mean it's you can't just you know there's no it's not like you can watch a TV show make shows about bank compliance.

I hope they don't because like it would be a very not interesting show. So basically once you understand what these what these people are driven by they're driven by the profit motive but what they're also very driven by is um avoiding regulatory uh backlash right having to pay billions in fees which many of them do every year having to have their brand damaged and their clients leave them because a regulator stood stood out against them. These are very very very serious risks that they are hyper paranoid about. And so when the regulator is sitting there, the US regulator or others like last year I spent a ton of time in Singapore, Hong Kong, Dubai, these places working with the regulators, working with the institutions there because that was the only place you could really work with anyone because the regulators would allow you to do it. Now the US has done this, you know, fantastic, magical, just amazing 180 from we we are completely against this, this is illegal.

If you do this, we will come after you to we really love this. We want to be the global leaders. Whoever does this first, we're going to we're going to incentivize you to do it. Please do it as quick as you can. And so now they are they are all looking at each other and they understand that someone's going to win this race.

like this race is is it's happening. That's it. So, so there's going to be winners, there's going to be losers. And now the fear is not about, oh, you know, I it makes sense and the settlement time is better and collateral management is better, but I'm kind of afraid, you know, I don't know if I can get this to the board. The executive committee won't say yes to this.

Then Bitcoin, there was a drug dealer involved at some point. We don't know about that. You know, like all this Now they're like, "Yeah, there was a drug dealer involved, but my competitor down the road seems to be doing this a lot, and if they beat me on this, it's gonna be real bad for me, so I got to worry about that." And then they go to the executive committee and the committee is like, "That drug dealer was so long ago." You know, like the competitor though, the competitor though today seems to be doing something, so maybe today we should do it.

They're really worried about this and that's great because they they they they're rightfully worried that that like this is a really big opportunity and that and and there are very big institutional competitors of theirs that are moving very quickly now. We know because we work with with I would say but are you implying cutting corners strategy or just pick the thing that's available out of the box and go with that?

I think it should just be Shane. I mean that's really no but but all that aside um which that actually is my real advice but if I wasn't working at chain that'd be my advice if I knew what I knew but because I know how the other systems work and really many of them are such and it's unbelievable how insecure they are and they're they don't they don't even have an idea of what some of these features that they need are just you know just to be this is me speaking as if I didn't work at chain link this be my opinion but but what what no what they should What they should do is create an architecture that allows the minimal slice of their existing systems which they they have to have involved in these transactions for legal and what are what is known as control location reasons. They take the minimum slice of those systems which is what we have we at chain link help them understand like what is the minimum part of your existing stack that you need. They take that they make it compatible with the blockchain events. So you turn an accounting event from your backend as into an accounting event on a chain.

So you make it compatible, right? So an event on a chain is is understood as an accounting event in your system. Something like that. That's an

just regular bookkeeping.

I mean it's bookkeeping. It's compliance. It's identity. Nothing's going to go live without any of this by the way.

So you need to solve that problem. then you need to solve the problem about how to build the contracts and our um views on that are use you know use 3643 in the US use something called CMAT in Switzerland you use these regulator um approved and and kind of known standards that are already doing well so you have token standards 3643 is a good example here in the US um so use some of these standardized uh methods of making tokenized securities equities funds then work with chain link or you know make sure you have an oracle to implement all the necessary connections into those contracts. Make sure you have a way to make all the contracts work across chains. That that last one's really important because even if you launch a really good tokenized asset on your chain, whatever chain that is, it could be Ethereum, it could be your private instance of EVM, could be whatever, if you can't connect that token, that that issued asset

to the rest of the world.

Yeah. to the rest of the world's liquidity, you're gonna you're gonna be very unsuccessful.

You just get the same bank you are already.

Well, you get it's actually worse. It's worse.

You get you get less than your bank because because the percentage of your client base that's onchain to buy your thing is very small and your bank doesn't want you to take your existing client base and have it buy the thing they just made. They want uh net new clients to come from somewhere else and buy the thing you just made. So now now you're you're looking at not only how do I synchronize this with my existing systems to make it go live, how do I get all the data, identity, compliance stuff working in relation to all the things I need to get connected to the contract, but also how do I connect it to all the sources of liquidity? Um and and what Ethereum wants to be in my in my strong opinion is the place where you want to do public chain issuance. So you want to be the place where everyone goes issues their tokenized asset, the tokenized equity, the tokenized fund, the tokenized commodity and part of that is going to be the liquidity inside of Ethereum and part of that is going to be the liquidity that that token can access because it's live on Ethereum even if that liquidity is outside of Ethereum.

Right? That's where crosschain connectivity is important beyond 7683 beyond intense because intents unfortunately I I don't think will work for banks because banks need a lot of

need slightly better guarantees.

Well banks need very clear predictability about who will do a transaction for them like who will process the transaction

knowing the counterparty. Yeah.

Right. Who who Yeah. Right. And especially for money transmission which is which is what some of this stable coin activity might be related to at some point. So you you you really uh need a way where you're you're guaranteeing certain u once again compliance factors.

The the the problem is with explaining this to the Ethereum community is not that they're not interested and and not that they don't understand the value of this. I think it's just it's such a different category of problems from the usual DeFi problems of how do I loop this coin, you know, even more to get the yield even higher, which is fine. like sure let's loop the thing until until you know the the the mathematics can't handle it anymore whatever you want to do but um you you you need net new capital right that that's that's the reality

y so final question December 2026

what's your prediction for net new assets represented on chain on Ethereum from kind of where we are today

on Ethereum

on Ethereum

that's tough I don't know what the current mixes RWAS on Ethereum or off. I think it's still largely on Ethereum. I mean I mean I think our

chain link powered net assets issued December 2026.

I think we're at 93 billion now. I think DeFi can double by then. So that's probably 200 or more billion. Then you have the institutional stuff. You have all the RWAS.

That's not a lot right now. That's really really nothing. It's very early. you you're at you're in the 200 to 350 mark. I think that maybe

that's just institutional or you're saying the total is that

total RWA.

Okay. So you're we're getting close to like a trillion.

We're getting we're getting close we're getting close to a trillion. I mean we're getting to half a trillion for DeFi. I think at least we're getting to my opinion. I think RWAs are going to net net grow faster than DeFi because it's still so early. I think that'll be at a trillion as industries.

chain link power is hopefully the majority of that. So, I don't know. It's it's it's hard to predict. Sorry. I'm really bad at this.

I don't do this. I don't predict like numbers and what the Bitcoin price is going to be. I'm actually not bad at it because when we look back at what I've said publicly,

I think you just contradict yourself three times in a row.

I disclaim. This is me not wanting to answer the This is No, I already figured that changing the subject. That's all.

I already know that. So uh so well in that case use chain link and thank you Sergey for

thank you very much

amazing conversation.

Thank you. [Music]