And we are live. Hi everybody. Welcome back to Blockchain will save the world. I am here with Yuval Rooz, the CEO of Digital Asset. It's a big show today. We're talking about the newly minted, freshly announced Canton Network. One of what I think is one of the most ambitious blockchain and DLT projects in the world, and Yuval is going to help break it down for us, talk through the origin story, and help explain exactly what's going to happen in the next few months and weeks ahead. Yuval, how are you doing today?
I'm great, and thank you for having me, Anthony.
That's my absolute pleasure. Coming to us live from New York. It looks like it's sunny and warm there, so I hope you're enjoying a New York summer.
It is. There's no smoke from Canada today, so we're all good.
Blame Canada for that one. Glad things are well. Let's jump into a quick background for those who don't know you, for those who don't know Digital Asset. Digital Asset's been around for a long time. A number of very significant projects in your history. For those who don't know you, give us a little quick synopsis, if that's okay.
Sure. So, just real quick, my background is more front office technology. Started my career at Citadel as a developer, and then moved to DRW Trading, where I ran an ETF market making desk, and then started a venture arm of DRW, which is called DRW VC, which makes a lot of investments, actually, in the space. And then in 2012, '13, roughly, DRW started a unit called Cumberland Mining, which is their crypto trading arm of the company, and while I was doing the investments, we started looking into investments in the space. And, you know, after a few years, realized that there is a big opportunity for many industries, but starting with financial services to benefit from the technology, or maybe more accurately from the ideas behind the technology. And that led to me and another person who actually started Cumberland Mining to leave DRW and to start Digital Asset.
Very good. So, that's the origin story. And Digital Asset have had a number of different projects over the years. Any, I know you're not allowed to have favorites, but any that you'd call out as examples of where Digital Asset technology is being used today that you're particularly proud?
Sure. I think, you know, to be honest, I think that one of the first projects that brought a lot of attention to us is the ASX and even as of recently. So, I'll start there. ASX is the Australian Securities Exchange. We started a project with them on replacing all of clearing and settlement in Australia, meaning that every publicly traded stock that will trade in Australia will be managed on blockchain. You know, unfortunately, the project right now is at a pause until ASX decide how they want to move forward. Happy to talk about kind of my perspective on why it happened, but that's one. I'm still actually very proud of what we achieved there because one of the things that people were very concerned with respect to the technology is would you be able to actually meet the throughput and volumes of a publicly traded market. And I am, you know, I'm very happy to say that we were able to show that you could actually run one of the top 10 equity markets in the world using the technology. And there's, you know, other reasons why we are where we are, but, you know, at least I'm very happy that from a technology perspective we achieved what we wanted to achieve. Other projects that are worth noting in the public domain, we're doing a very exciting project with the Hong Kong Stock Exchange, which should become live any day now, which is really streamlining and reducing a lot of breaks that happen when you trade between China and the rest of the world. China is a T0 settlement, the rest of the world isn't. And that mismatch in settlement creates a lot of breaks, and we're using the technology to streamline that bridge, which is called Stock Connect. We went live with Deutsche Börse on tokenization of structured products, so short-term call options, and that took the issuance process from 3 to 6 days, which is kind of a problem when you're trying to trade a 1-month call option, to effectively real time. Other interesting project with Broadridge, we have, I would say that that's the closest C5 project to DeFi. It's really automate the whole repo market in the US. Just to kind of put numbers to people, the repo market globally is about $10 trillion every single day. Half of it is in the US, so about 4 and a half to 5 trillion dollars trades in the US every day. And there's a lot of opportunity to use smart contracts to automate that process to have it more efficient. So, with Broadridge we launched a repo product, which includes three types of repo, so intra-dealer repo when you want to collateralize your own internal or finance your own internal entities, bilateral repo, so between two counterparties, but then also recently, intraday repo. So, when you don't want to just do overnight financing, but you just want to do very short-dated, not even dated, short timeframe, financing, you could actually do a repo transaction for just a few hours, and it was just public that DRW and SocGen just did a 3-hour repo, a few weeks ago. And I think that you're going to see that continues to grow because given interest rates where they are, being able to have much more precise financing windows is going to be a very big opportunity. Moving on, tokenization, what we're doing with Goldman, HSBC around digitizing asset classes. Happy to talk about why I don't like the word tokenization. I just generally speaking find it very misleading. So I like to talk about digitizing asset using smart contracts, I think is a much more accurate reflection of what we do. And again, they've done a few bond issuances with those individuals, but I think you will see before this year end, a very big issuance by one of our customers on one of these platforms. And you know, so those are the things that are, you know, in the public domain. There's a few others we've done around securities financing with Aquiline, but there's a bunch of others around streamlining post-trade for physical net gas, syndicated loans, which is a massive, it's $3.5 trillion market in the US. Mortgages, so issuance of mortgages, there's a bunch of very interesting. And then we also have customers in the insurance space and sports wagering, out of all industries. But our main focus is financial services at the moment.
We've taken up already half of the show with the intro, which shows the breadth of work that you guys have been up to and genuinely thank you for the kind of some of the clarification points and also coming out of the block with ASX. It's a well-known project. It was a poster child project for a long time, but I think what it shows you is that digital transformation is still hard. Whatever technology you're using, if you are replacing, lifting, and shifting fundamental infrastructure, whether that's a core banking system, whether that's cross-border systems, or whether that's just cutting out something that's part of a core enterprise system, it still takes time, it's still hard, it doesn't always work, and it's not necessarily because the technology isn't fit for purpose, either. There are people factors.
You're our spokesman, Anthony.
Well, not for nothing, I do the odd bit of marketing work from time to time. But this is what the show is about. In fact, this is what I would like the industry to spend more time doing is demystifying, clarifying, using human relatable terminology to help understand. Like the financial services, capital markets, investment banking, these are opaque and oftentimes very technical domains in the same way that blockchain and Web 3 often is. The way we use some of the terminology, the way you casually throw repo around as if everybody should know what a repo is is one of those things that the rest of the world struggles to catch up with us when we explain why is digitization of assets a useful thing to do? Not because, you know, everybody assumes assets are already digital. I have it on my phone, right? My assets are digital, aren't they? Well, no, because there's a whole bunch of plumbing that goes on in the rest of the world that is a little bit broken, and we're trying to fix it.
Yeah, it's super interesting what you're saying. My co-founder has been recently using this analogy or story, which I just think fits perfectly to what you just said, which is, you know, when he talks to clients a lot of times, he says, you know, we've been living in the LT world for the last 60 years. And when he says that, people pause and say, 'What are you talking about? This is a new thing.' Says, 'No, it's, you know, if you really just take the idea of a blockchain is really just how do we take ledgers and come to an agreement between those ledgers?' So, yeah, today we're using technology to solve that, but in the last 60 years, you were just using people as your consensus mechanism, right? And that's why I keep on trying to say that really DLT in a lot of ways is not really trying to change the world, it's really trying to change the way we operate the world, right? And this idea that today when you have breaks or disagreements, you just send it to offshore teams, you know, whether it's in India, the Philippines, different locations in the world, and then you have just people creating some kind of consensus, and that consensus gets done via emails, phone calls, messages, whatever, but at the end of the day, you get to some consensus. You know, they keep on talking that proof of work is a very expensive consensus. I would say that the consensus that we run today, while not having a fully digital world, is a much more expensive consensus, right? That's the reason why offshore got created is that running these type of tasks in developed countries was too expensive. It just kind of made businesses really not financially feasible. And when I think about DLT, all we're trying to do is say, 'We still need to get to consensus because at the end of the day, we need to exchange things when things settle.' And we need to have an agreement on those things. So, we're just trying to short circuit that very expensive consensus mechanism to make it much more efficient, much faster, and as a result of that, improve margin, open business opportunities that today wouldn't be possible. So, thank you for saying that because I just truly agree with that sentiment.
It's the going, we're kind of going backwards to go forwards. We're going from kind of expensive people consensus to slightly cheaper people consensus to then digitization. The fun part, and this is the reference that I like to make to Singularity University, the Peter Diamandis version of this is that once you can get away from physical manual, you can get to digitization and standardization, then you can democratize access, and then everybody can participate in building software, building applications. Okay, not always in a regulated industry, but the opportunity to reduce cost even further and transform in a bunch of different directions.
But I think the keyword of what you said is the opportunity because it's not by definition, and this is by the way why I don't like the word tokenization, because it implies if something exists in a token, you effectively democratize the asset. That's just not true. And that's why I like to use the word digitization, and what you just said is I think that if an asset was born digital, was managed digital front to back, and now I'm going to use the word you used, you create the opportunity to democratization. It doesn't necessarily mean that that is definite what's going to happen, but I think you have created the opportunity to do that. There's many reasons why potentially a private equity firm, even if you said the marginal cost of managing another investor, would not want to open it to 10,000 investors because they like running a private equity fund with only 50 investors. That's just kind of their business they like to operate. But I think that if you gave them the tools to fully digitize how a private equity is being managed, that the marginal cost of adding another investor or potentially another thousand investors did not cost them more, I think you definitely created the opportunity to democratize private equity investing, right? And I think that that's the key around tokenization is how do we make assets, you know, highly more efficient or significantly more efficient because that creates these opportunities. It doesn't mean that that will be the outcome. It doesn't mean that there will be more liquidity, but you have definitely created the opportunity. And that's why I like to give this example. Robinhood fractionalized shares without tokenizing anything. They fractionalized shares just because the way of trading US equities is highly digital today. It's highly efficient that the marginal cost of letting you trade 100 shares versus 50 shares versus 10 shares is no longer significant. And therefore, allowing you to trade half a share is possible.
There we go. Great background context. And with that as the backdrop, let's talk about Canton. Because I mean it when I say I think this is one of the most ambitious blockchain led projects in the world. What's the problems you're trying to address, the problems you're trying to solve I think are significant both in the non-blockchain way of running financial services systems and the blockchain way of identifying and managing financial services systems. So, in your words, where did Canton come from? What was the inspiration behind it? And what is it that you're trying to change with this particular initiative?
Perfect. So, I might say a few things that are controversial for people in the space, but I'll try to make them as coherent as possible. And I think that if you start with Bitcoin, the idea of Bitcoin was that in order to create a censorship-resistant asset that has no issuer and have a limited supply, you needed to create a pseudo-anonymous privacy model. And that was very novel with Bitcoin. And I still think is very novel and I think it's something to really admire. But the blockchain and the privacy model for Bitcoin was not a blockchain to create applications and not to create, you know, other use cases. It was really just to serve this one use case, which is Bitcoin. Censorship resistance asset with limited supply, right? Like that was the use case. I think that what the problem, in my opinion, and this will give the context to how we thought about Canton is that then came Ethereum and Ethereum said, 'Well, let's take exactly the same concept.' Yes, there's account structure versus UTXO. But high-level really did not change the idea that there's no privacy. All the miners get to see everything. They validate the transaction. And you also have one ledger to rule them all, meaning it's the world computer. And I think that there's a few challenges with that. Well, first of all, the idea that one ledger will be able to run the world, in my opinion, is just against the laws of physics. It just doesn't work. And I know that a lot of people will say, 'Well, now it's the security layer for a lot of L2s.' I still think that when you start actually looking at what does it mean to be the computer of the world, you can't just put it on one ledger. It's just not, in my opinion, physics will not support that. And that was one thing. The other thing, this idea that you need to have third-party infrastructure provider that get to see everything, they get to validate your transaction, from my perspective, is another concept that just doesn't really make sense. And we're just seeing kind of the outcomes of that. You know, there's this nice term I like, people know how to use words to make something look like what it's not. So, miner extracted value. Sounds very sophisticated. Sounds complicated. But, if I came to you and I said, 'You know what? NASDAQ has a prop trading desk.' And that proprietary trading desk can look at all the market activity that comes into their matching engine. And before they submit these packets into the matching engine, they can put their orders ahead of time. I mean, someone would sit in jail for that. Like if someone found something like that, you would be in jail, right? NASDAQ would cease to exist as a company. That would not be acceptable. Not just to regulators. It would not be acceptable to anyone. If NASDAQ could see your trading on Robinhood or whatever, can see your transaction, can say, 'Hold on. I'll just put my transaction ahead.' And that's just one outcome of miners seeing your information. But, just generally speaking, it means you don't have privacy. And when you look at the world, whether you like it or not, and this is where I'm trying to separate personal belief from just the reality of the world. It just won't work. Regulated institutions, proprietary institutions will just not operate where everybody can see all of the information that is related to their activity. It's very similar argument that I say to people that don't believe in Bitcoin or gold. I say, 'Listen, as a trader, I just believe that there's enough people out there that do believe, and that's why I believe that there's going to be value to that asset. Doesn't matter if me personally believe in it or not.'
And that's why the extrapolation of the everything is a Ponzi system belief which I love using it as an example regularly enough because it forces people to think. You know, the American government is a Ponzi system. The dollar is a Ponzi system. Gold is a Ponzi system. Bitcoin is a Ponzi system. Your local farming cooperative is a Ponzi system. Literally everything is a Ponzi if you take it to literal definition. But suffice it to say the flip of that is its network effect. If you get enough parties behind a system, if you create enough value within that system, it will be sustaining.
Exactly. I totally agree. So, if you take that, we said, well to us just those two things are fundamentally not going to work. Now, I would do want to give the positives about the Ethereums and the likes of the world, all the smart contract platforms. And that will kind of lead into private chains or private blockchains that were created very early on is I don't think that you can compete with user experience. And what I mean by that is this idea that you have a wallet or effectively a browser into a network. And I'm doing network on purpose. And you could access all applications on that network to me is a user experience that you will never be able to compete with. And at the end of the day, we know that user experience wins. All right, so that's one thing. And I think this idea of composability that you get on these smart contract chains, you know, I like composability more. Some people at Digital Asset will call it application interoperability. I think is another way to call it. Not to be mistaken with interoperability across chains. I don't want to talk about that just yet, but it's more the ability of two applications that have been created independently to be able to work with one another seamlessly. I think is a very very powerful concept. And I think we took those two bad things in our opinion and two good things and we said, 'Well, how can you create a network that gives you the best of both worlds?' And I think that that negative is really what created private chains. Like, how do I have application ownership and control? How can I control who can connect to me? And that's what led the birth of Fabric, Corda, and some of these private chains. But I think that what they failed to look at is this idea that at the end of the day, you could have a really good product, but if you have bad user experience around it, it will never succeed. Right? And we've seen so many examples of that in history. And we tried to think about, 'Well, how can we change that?' Like, that was really the concept. So, Canton was born and really the name Canton comes from it was developed by our team in Switzerland, but there's also another nice meaning to it that, you know, Switzerland is one country, but it's actually built out of these cantons that have sovereignty, right? They have sovereignty, they have ability to create their own rules, they have their own tax regimes, but they all operate under this one country called Switzerland. So, we thought it was a really nice gesture to the team in Switzerland that really worked hard developing it, but also kind of taking some of the ideas behind Canton. So, really what Canton tried to do is to say what is blockchain trying to do? Blockchain is trying to say if there is a piece of information how do you get verification that the people that are interested in that information all share the source of truth? And I think that what Canton said is, 'Well, that depends on the application, not on the network, right?' So, again, if you take Bitcoin, Bitcoin is an application. It's a use case. It's an asset class. And that use case says, 'Well, all parties to the ledger need to know about this asset class and about all account balances in a pseudo anonymous way.' But, Anthony, if you and I created an application to issue poker chips for our weekly poker game, well, it's just all that we care is that the eight players that play every week on a Sunday night would know about our account balances. The rest of the world doesn't need to know about it. And it's not even relevant if we care that they know about it, but just from an efficiency perspective, there's just no need for them to know about it. Right? It doesn't provide any additional benefit. So, just if you think from a data usage and network usage, it just creates a lot of traffic and a lot of information that really doesn't contribute to the mission that we're trying to do, which is making sure that we all agree on who has how many poker chips. And then you can take it even more to the extreme, and if you wanted to have some application around a doctor visit, and I'm taking it on purpose to something that we can all relate to, but we don't want anyone to know about it other than ourselves and our doctor. So, there you have a privacy level that you're saying, 'I just want it to be bilateral.' And if you take just three examples that we gave very quickly here, well, what ends up happening is that today in public blockchains, the chain says, 'What is the privacy model?' And you as an application need to figure out a way how to work with that privacy model. Well, this is where, in my opinion, Canton is very different than anything out there. Canton said, 'Well, actually, you tell me what is your privacy model, and I will then enforce that privacy model.' So, rather than I am just going to come up with a very homogeneous privacy model, I'm going to say, 'Well, the world is not homogeneous. It's actually very heterogeneous, and therefore, I need to be able to satisfy multiple privacy models, and you, the application, will define it.' And I think that that exists also in the internet. So, if you think about the internet, the internet, when you create a website, you could decide not to register it as part of DNS. Even if it's on DNS, you could say that you are like Craigslist. We're just a bulletin board. Everything here is public. Or CNN. Your privacy model is you want everybody to have access to all the information. Or Wikipedia. Or all of these websites that their privacy model is there is no privacy. We want to be fully public. On the other hand, you have something like chase.com, and chase.com is to serve a bank. And a bank, whether we like it or not, is highly regulated, and they cannot offer their services to anyone who's not KYC'd AML. And therefore, you can land on chase.com, you will see some informational material about Chase. You'll see maybe some information about where markets are today, and things like that. But that's it. If you want to start accessing the actual service, which is the bank, you have to put in username and password, meaning you have to be allowed by the application provider to have access. And that's what we wanted to create with Canton, this ability to serve all these types of applications. But then taking it one step further, I think that we talked about application interoperability. So, the reason why it's very easy in our opinion on public chains today to do composability and interoperability is because the privacy model is identical. And all of the assets, you know, ERC-20, big breakthrough for Ethereum, it just said, 'Let's just normalize all these bearer instruments to be kind of like very simple and very, you know, so it's very easy to effectively do all the DeFi and all the decentralized exchanges just to exchange it because they're all kind of identical bearer instruments that represent some other asset class.' We said in Canton, 'Well, how can we actually compose two applications that don't have the same privacy model?' So, imagine if you implemented USDC, so a stablecoin that is pseudo-anonymous, meaning everybody can see account
balances, but we can also implement the application of, like I said before, of a doctor. So, now imagine that when I go to the doctor, I want to settle my transaction with the doctor using USDC. Well, now I have two applications that don't match from a privacy model, and really what Canton allows us to do is you and I, Anthony, let's imagine you're my doctor, will actually be able to see on Canton a composed transaction of my doctor visit at your clinic and USDC as one transaction, and anybody else who's just a USDC holder would just see that money moved from one wallet to another and would have no idea what is the reason of that transaction. So, you could actually honor the privacy model from both sides. And I can keep on going, but you know, high level to us, that's really the big change is really flipping it around rather than saying, 'Here's a privacy model, figure out how you work with it.' is actually saying, 'You, the application, define to us what is your privacy model, and the chain will enforce it for you.'
Got you. And so to summarize, I mean, some of the challenges that we have with the existing financial system and ledger management, we kind of established that at the beginning of the show. In the kind of early evolution of some of the chains like Ethereum and Bitcoin, you have standardization, but they're designed to be universal ledgers. They're designed to be universal infrastructure that everybody has to follow the same rules for the most part. You then get into application design, application infrastructure, application management, and you're saying, 'Okay, well, we'd like different applications will behave differently in regulated unregulated industries, in information that should be made available for everybody to see, right? You want corporations to be transparent, but people want to preserve their privacy. Or you know, there's even in B2B transactions, you're going to have some intellectual property or some elements of transaction details that you don't want people having the ability to correlate or having the ability to see.' And so, we need a mix of both ledger technology that people can trust to be scalable, automated, standardized to some extent, but we also need to have a degree of composability when it comes to the privacy and other elements of how the infrastructure works. And so, put those things together and we get Canton.
Yeah, and I think you touched on something maybe that I didn't talk enough about, which is the scalability. And how do you solve for scalability with a user experience as well? So, Canton does not assume that there's one universal ledger. Again, from the name Canton, there's many Cantons, right? So, this idea you have to be able to shard the problem. You have to be able to split the problem. So, you know, I give another example, if you use WhatsApp, you know, all you care within WhatsApp is that you have consensus and proper ordering within a message. You don't really care if there was consensus across messages. So, you can think about every one of your messages as one of a Canton, a Canton, a Canton, a Canton. But so Canton itself can shard the problem set. We call it domains. We don't call it L2s. We call it domains. So, you can effectively have infinite amount of domains. But again, you want to have a user experience that when I connect to the network, I still feel like I'm connecting to one global network. Very similar to the internet. The internet is not one network. It's actually many networks that use internet protocols to communicate with one another and to create a seamless experience that when you log with a browser, you feel you're connecting to one network. So, very similar with Canton, a Canton client node, effectively your wallet, can connect to the Canton network as one entry point. You could feel like you're working with one network, but it is actually sharding the problem. And this is where I think the challenge, when you look at all the L2s, in order to compose a transaction across multiple L2s, you effectively have to go down to the mainnet to coordinate that transaction, which makes it very expensive and very slow. In Canton, you could actually coordinate a transaction directly between domains. So, that's where we believe also it's not just application composability, it's also how do you do L2 composability, which is a very big challenge. And again, from a user experience perspective, we don't expect the user to think about, well, how do I break the sharding? Because, you know, in Polygon and some of these other chains, you need to explicitly create another L2. You have to explicitly create it. And maybe for very simple use cases, that works. But when you start trying to think about, well, how do I build a syndicated loan that has so many different states with so many different privacy levels that change dynamically as the problem gets more and more complicated throughout the life cycle of an asset, you realize that the big O problem or the complex or science perspective becomes unmanageable by a human. So, again, for us, the idea of creating these different domains of privacy is something that has to be dynamically managed by the chain, not by the developer. So, all that the developer is modeling is what is the state at the moment and who are the stakeholders to this state. And it's the chain underneath that is going to manage all the sharding of that problem.
Got you. And I'm glad you started to come on to the kind of the participants, because I think this is the most interesting part of any decentralized network is how you create that scale, how you create network effects, how you encourage innovation. How do you I think a lot of early stage layer ones and even some of the layer twos are so complex or require such effort from a developer's perspective to learn domain-specific programming languages or to have a SDK or integrations that are just either too complicated or for enterprises or banks, even, are just too much like hard work. Enterprises today want to consume software as a service. You talked about the user experience challenge. Often times, being able to deploy new software should be as close to the existing stack as possible. If there's a requirement for specific domain languages, if there's requirements for unfamiliar, untested technology with no SLAs, you're going to have challenges in enterprise adoption. And as much as we would like decentralized technology to scale with the 100,000 angry degens on crypto Twitter, the reality is that this technology will scale best if it's put in the hands of those organizations that intermediate or that facilitate transactions of money and data, because that's what as human beings we do the most of. Right? So, if this is technology that's designed to help humans, put it where data and money flow, and that's kind of where you guys are.
Yeah, absolutely. And I think you touched I mean, I will tell you Canton and generally speaking our stack has a long way to go to make that user experience as seamless as you talk about. I think what we have been really trying to think is, you know, as a former developer, I would definitely not call myself a developer today. You know, when you think about the role of So, I started as a C++ developer. And when you think about it, well, that's very good for high-frequency trading because you have much closer interaction with the hardware, much more control of the machine. But 95% of applications, probably even higher in the world, do not require that kind of familiarity with hardware, and that's why Java became so popular is because it created all of this abstraction that you as a developer don't need to worry yourself with that complexity. We thought about when we created Daml, our smart contract language in Canton, as kind of a similar idea is well, if you're now moving to saying I'm going to manage a distributed system, well, the amount of variables that you need to control to get that right becomes significantly more complicated. So, the first thing that we try to do is how to abstract that complexity of managing privacy, all of those things that we talked about right now by the developer, and try to create a much more abstracted interaction for the developer. It does require, you know, rethinking about how do I model things because if you are a developer who have been building applications that were centralized for the last 20 years, it does require a change in paradigm of how to think about application architecture and a lot of things. And I think that we can still be doing a lot more to improve that, but that's kind of where we came from. So, we started by saying like, if you're just going to create something that allow the developer to control all of these variables, it will never scale because it will just become so complicated so quick that it would not be feasible to build really large-scale applications. We solved that problem. That was early on. We're now very focused exactly about the things you said, deployment of application, monitoring of applications, you know, availability, scalability. There's just so much still that we can do, but that was kind of our starting point, how we started, and there's a lot more to do.
