With the recent plunge in prices, and the unraveling of certain protocols, there's a fresh round of questioning about what the point of crypto actually is. There's tons of money pouring into it, but still not a lot of tangible examples of it making things better, except among people who have gotten very rich on it. On this episode of the podcast, we speak with Arthur Breitman, the creator of the Tezos blockchain about blockchain design, and what it's actually good for. The episode was recorded in early April at the Milken Conference in Beverly Hills. The transcript has been lightly edited for clarity.
Points of interest in the pod:
Can blockchains actually save the world? — 3:36
How Breitman got into crypto — 7:44
The magic box — 12:23
How do you measure decentralization? — 23:34
The problem with a multi-chain world — 34:55
Unstable coins — 41:03
Joe Weisenthal: (00:10)
Hello, and welcome to another episode of the Odd Lots podcast. I'm Joe Weisenthal.
Tracy Alloway: (00:15)
And I'm Tracy Alloway.
Joe: (00:17)
So Tracy, we are here at the Milken Institute conference in Beverly Hills.
Tracy: (00:23)
I know. It's very fancy.
Joe: (00:25)
How's it been going for you so far?
Tracy: (00:27)
It's been fun. I talked a lot about credit. I watched your panel about crypto and I have to say there were definitely more people at the crypto panel than there were the credit panel.
Joe: (00:38)
And also every panel that I've seen that's not about crypto, someone asks a crypto question.
Tracy: (00:48)
Someone from the audience submitted a question about how crypto would impact inflows into investment grade [bonds]. And it, you know, interesting thought...
Joe: (00:58)
Well, crypto's one of those things too, where everyone gets an opinion on it. So everyone wants to know everyone's take, even if it's not their specialty.
Tracy: (01:07)
This is very true.
Joe: (01:08)
But I'm very excited because today we are gonna be speaking to someone who knows a lot about crypto. Not just a tourist. And I think a topic that we haven't really talked about that much is blockchain architecture and design, and you know, all of these panels, like everyone's talking to they're super like fanciful. It's gonna create this world of like openness and transparency. And it's gonna cut out all the fees and it's gonna be so great, which is really weird because when I think of blockchain, I think of high fees and long settlement times, and basically like all kinds of interesting tradeoffs that in many ways are worse than the existing financial system. So I feel like people are jumping over some like basic architectural points in this whole conversation.
Tracy: (01:55)
The other thing that gets me, especially at conferences like this, is everyone wants to talk about crypto. Everyone seems to have an opinion about it, but I have my doubts about whether people actually understand the underlying technology and whether or not they're differentiating between different blockchains and things like that. I think people might understand the basics, like proof of stake versus proof of work, layer two versus layer one or whatever. But like within those technology buckets, I don't think people are really thinking about the nuance.
Joe: (02:24)
Right, it's always about how gonna be really good and it's gonna be really great for financial inclusion and democracy and all that stuff. So anyway, let's have a real conversation about how these change work and what you can do with them. I'm really excited. Our guest knows a lot about this stuff. We are gonna be speaking with Arthur Breitman. He is the co-founder and creator of the Tezos blockchain project. One of the sort of original projects that thought about smart contracting and malleability in a way that, you know, much more advanced than, or seemingly more advanced than what you can do on, say, Bitcoin. So we're gonna talk about what you could do with the chain. So Arthur, thank you so much for coming on Odd Lots.
Arthur Breitman: (03:06)
Thank you for having me.
Joe: (03:07)
We've been interacting on Twitter for years and years, so I'm excited that you're here.
Arthur: (03:13)
Yeah, me too.
Joe: (03:14)
So what do you think when you hear all these people at these conferences and you're here too, and they're like, ‘Oh, it's gonna like be really cheap. It's just gonna lower the fees for everything. It's gonna be super fast, transparency, financial inclusion. It's gonna save democracy. It's gonna save the planet.’ What goes through your mind?
Arthur: (03:36)
I mean, I think some of it is true and some of it is false. In terms of reducing the fees compared to traditional financial infrastructure, I actually think this is one of the real ones — that we can actually do this.
We can lower the entry costs for creating financial products for financial systems. Make it a little more innovative and develop these things much faster than they have in the past. I do believe that you can remove a lot of intermediaries and a lot of the friction in the financial system comes from the fact that you need to rely on trusted intermediaries. As an intermediary, you have oftentimes a big opportunity for fraud. And monitoring that fraud has a lot of costs and administration. You have to deal with only a few people who are gonna be trusted.
If you can get rid of that, you can actually generally reduce costs. So I do believe this part of the story. I don't know that it's going to save the planet and do all these grandiose things. A story I have about this is I once gave a talk at a some sort of TED-like event and it was about the promises of blockchain. And I gave an example for some of the over-hyped ideas. And I said, “Well imagine, for example, plastic bags are killing turtles, but the problem is we don't know where they came from. If we could identify who leaves a plastic bag, you know, it solves the problem. So let's put the plastic bag on a blockchain and we'll save the turtles.” And I gave that as the example of like some of the nonsense you could see, and someone pitched me that exact idea the next month.
Tracy: (05:11)
That's amazing. Just going to fees though, I mean, a very cynical person would say that the fees exist because yes, there are underlying costs like fraud and KYC and AML and stuff like that. But if you were being very pessimistic about the whole thing, you would also say that financial intermediaries, traditional financial intermediaries, they like to make money. A lot of those fees are opportunistic. I guess my question is like, do those fees actually exist because of what the intermediaries are doing and the cost, or because of the way the system works? And how do we make sure that in crypto, we don't end up with centralized intermediaries that are gonna charge big fees anyway.
Arthur: (05:54)
I think everyone who's in business likes to make more profit and likes to charge more fees and people are gonna charge whatever amount of fees they can get away with in general. So at some point there’s competitive pressure, right? And in finance, it's true that sometimes a lot of the competitive pressure is not present.
You'll have people pay, you know, 1% commission on a trade because they're doing it through a private bank and and they trust the private bank to do good execution and so on and so forth. And that’s nonsense, it doesn't cost that to execute a trade, not at all. But the reason that they're doing it is because, well, they feel that they need the trust of this very prestigious intermediary.
If you remove that notion of trust, I think you get something that's more competitive because the thing is going to work regardless of who you route it through. It limits the ability of intermediaries to charge fees. Now, some people will try, of course, especially I would say there's gonna be a cottage industry — there already is in crypto — of people who want to deal with more traditional investors and who reassure them that “no, no, no, it's all going to be fine.”
I mean, we see it, you know, you could buy Bitcoin and even hold it with a custodian, if you're worried about self custody risk. But nonetheless you'll find funds which charge you, you know, management fees for holding this. Some people are always gonna want that, but it's still putting some pressure down on costs.
Joe: (07:21)
Can you talk a little bit about your background and the founding of Tezos because you were a trader at Goldman Sachs, right? What did you do? What was your position in traditional finance originally and what did you see early on, that you thought there was like an opportunity to port some of these ideas or the opportunity you saw in creating your own chain?
Arthur: (07:44)
Well, my background is in computer science and statistics and my first job, I was working on as a quant. So I started out in statistical arbitrage and in 2009 I moved to Goldman to do high frequency training and worked on ETF market-making. Natural gas market-making. Also worked at Morgan Stanly and a few other places on these topics. I've also done some robotics after that, but that was quite different. The thing that attracted me to cryptocurrencies is it was at the intersection of a lot of centers of interest that I had. I have been interested in the theory of money and banking for a long time. So, you know, how does money arise? What makes good money, what makes bad money? And I remember, you know, way before Bitcoin, you know, convincing myself that the best money possible is backed by nothing.
If you use gold, for example, there's a problem where the fact that if, you know, if demand for gold increases, you get more mining. And so, you know, that's a waste, you know, because gold is used as a monetary instrument, you know, in an ideal world, we would have mined all the gold and increased mine all the gold … Because it doesn't serve a purpose to have more gold once you have a given supply. And I guess we could argue that having UST in the monetary supply, it can have some value, but I didn't see it that way. So when cryptocurrencies come around, I noticed this bad criticism around it. So people would say like, oh, it's backed by nothing. And I'm like, no, that's great. That is what you want with a currency. I was interested in the cryptographic aspect because I also had, you know, have a mathematical background. I was interested in cryptography. The individual sovereignty aspect was also some things that appealed to me from a political standpoint. So it really attracted early on.
Tracy: (09:24)
So you were one of the first ones to move from proof of work to proof of stake, is that right?
Arthur: (09:31)
We never used proof of work.
Tracy: (09:32)
Right. Sorry.
Arthur: (09:33)
Proof of stake from day one.
Tracy: (09:34)
Right, I mean you were one of the first to like actually do proof of stake. What was the opportunity there? And why did you immediately latch onto that versus proof of work?
Arthur: (09:46)
I got interested in doing Tezos because of my interest in proof of stake. Around 2013, there were a lot of innovations in this space around proof of stake, smart contracts, privacy. It was an explosion of research and innovation around the space. And one thing I noticed early in the Bitcoin culture is that there was not a whole lot of appetite in trying to integrate any of these upgrades into Bitcoin.
Joe: (10:11)
There's sort of a point of pride for them — that the code is so ossified or that it's so hard to change.
Arthur: (10:20)
There's a big… Part of it is pride, but I also think part of it is sour grapes. You know, for a while, the narrative was very different. For a while people in Bitcoin circles were saying, you know what ‘all these altcoins, they're gonna find the best, you know, ideas. And we can always integrate those ideas into Bitcoin.’ And we moved from that to “all those ideas are bad. We can't benefit from then in Bitcoin” which I think is a little sad because a lot of these ideas are good and important and have product market fit.
Tracy: (10:50)
This is one of the like ironies of Bitcoin, I think . It's ‘technology is gonna save us in many ways, but also we don't want the technology to advance too much because we like it the way it is,’ or we can’t actually do it.
Arthur: (11:03)
I understand the perspective, I think from the Bitcoin perspective, there's a big risk. If you let people change the ledger, the changes that you end up having are not merely technological improvements, but actually change the nature of Bitcoin and change its economic properties.
So I think what all Bitcoiners feel is that if they open the door to technological changes, they're also going to open the door to social control of the economic properties of Bitcoin, which is something that they absolutely do not want. So in some sense, they felt like they had to shut the door almost completely on changes in order to preserve the economic properties.
Joe: (11:39)
Let's talk a little bit about what you see happening on chain. We had a recent episode of the podcast with another person who once did ETF market-making or ETF trading, Sam Bankman-Fried. And he was asked to describe yield farming, and we thought we were gonna get this sort of very complicated answer, but it turns out in his characterization yield farming is they'll pay you to put money in a box. And if you put money in a box before other people put money in the box, then you have a bigger part of the box. And it kind of, it sounded to a lot of people not that different from a Ponzi scheme. What did you think about that? Is that a fair characterization of yield firming?
Arthur: (12:23)
I think it's a fair characterization of the way it's practiced if you weight by volume or prominence. Then yes, it’s a fair characterization. I wouldn't throw the baby with the bathwater. I think there's something there. You know, the more generous way of looking at it is, let's say you want to, you want to launch your token. You want to distribute it in some way. You want to get some liquidity. So you use part of your token supply as a way to incentivize people for providing liquidity for it. And you could see a company doing this. You could do this with equities. You could say, look, I have some equity. My business will be more appealing to investors if there’s liquidity around this equity? So I could, you know, I could set aside some equity and incentivize my investors to provide equity.
Joe: (13:06)
The issue with that, it seems to me, and I think we have seen this in crypto, is that once the equity is distributed, if your model inherently is about paying for customers in the form of equity, then once all the equity is distributed, do you really have good customers? Or do you just have a base of people who are gonna now go find the next company giving out equity for liquidity? And when I look at crypto, you know, I look at all these like DeFi protocols that were huge in the summer of 2021. And now their charts are way down. And I kind of get the impression that the game is, yes, you find the company or the protocol giving out governance tokens, as they’re called, equity. And then at some point the equity runs out and rather than sticking around, you just go find the next thing.
Arthur: (13:55)
Yeah. I don't disagree with that. The way I phrase it sometimes is that you see a lot of innovation in DeFi from a technological perspective, or even from a financial engineering perspective, but the point of DeFi cannot just be the trade DeFi tokens. And right now that is mostly the point of DeFi. So you need to, I think you need to anchor it to some actually, you know, real use case and where real demand comes from.
Tracy: (14:23)
So historically Tezos has avoided DeFi, right? You haven't been that involved in it in the way that say Ethereum has, but I think you started something last year called liquidity baking. Is that right?
Arthur: (14:35)
Yeah, absolutely.
Tracy: (14:36)
Can you explain that? Like, how is that different to the other bad ‘yield-farming-all-the-way-down’ type DeFi?
Arthur: (14:44)
Of course. So, first of all, I would challenge the fact that has this skewed DeFi, the way I say, is not sentient. It's a project and there are DeFi protocols running on Tezos. So the foundation helps in the general growth of the ecosystem. We distribute grants and we've given some grants to some DeFi projects. So there's no animosity or anything against DeFi. I think it has had little traction on Tezos and on other chains for a variety of reasons. But yes, liquidity baking is interesting. It was a modification of the protocol itself. So at the protocol level, an amendment was passed and ratified and voted on by the validators of the network that would take a small fraction of the, a very small fraction of the tokens minted on every block and then use that, deposit that into a pool to incentivize liquidity provision between Bitcoin and TEZ.
So the idea is that you have a little more inflation in your protocol. I think it comes down to about 0.3% per year. So, you know, it's not very, very meaningful for, you know, something that has a daily standard deviation of like about 5%, you know, like a few percent. So 0.3% per year, but it's to incentivize people to be willing to provide liquidity between Bitcoin and TEZ. And the point is that, first of all, these are not giant yields. The point is not to try to attract people with giant yields, or the point is not to try to bootstrap something new. It's just, it's a cost. You pay very small cost in terms of inflation, and you get something in return, which is liquidity provision.
Joe: (16:21)
I want to go back to the box for a second, because in theory, the box could become a bank, right? So in theory, we could all put money into the box, we get some equity, and then in theory, it could start, I don't know, making loans or doing something with economic purpose, something that actually creates value such that DeFi is not just trading of DeFi tokens.
Tracy: (16:43)
This is like the SPAC argument, right? You're sort of putting money in a box that might become something.
Joe: (16:48)
But is anything becoming there?
Arthur: (16:51)
You basically have something that's nonsense, you get a high valuation, and then you turn that into actually making something that works. We've seen some people attempt that. I don’t know if anyone has really succeeded at pulling off the Crypto SPAC so far. You have something like Chainlink for example. It started out not very coherent and then started, you know, hiring serious academic and like doing really serious work in cryptography for this. So, you know, it remains to be seen whether or not that's pulled off. But for me it's more about, it's not about how serious you are. It's more about, you know, at the end of the day, you know, it's about value creation. Like, what value are you creating?
And it can't be completely internal. Another way to think about it is, sometimes I'll visit a small town and I'm thinking, what are the exports of this town? So you have economic activity, you have restaurants, but the restaurant is for the people who work here, so they have to work on something. So at the end of the day, you know, what do you export? What does DeFi export?
You know, what are you doing? There's a cynical view of financial markets that I really reject, which is the idea that all the financial market is a giant casino, you know, everything is zero sum. It's a very popular view. But like financial markets are very important because they help get liquidity. Secondary market gets liquidity and it helps capital formation for companies and companies actually build things. So there's generally value in financial markets. It helps companies that do things and build things. And unless you are actually doing this in some ways, with DeFi, it's going to remain immature.
Tracy: (18:50)
Since you said this, since you're not a fan of DeFi for the sake of Defi, can I broaden this out massively and ask what is the use case of blockchain? And we ask this all the time and you know, yesterday Joe was on a panel with Algorand and they announced a big partnership with FIFA and Joe asked the FIFA president, ‘well, what are you guys gonna be doing in this new partnership?’ And he basically said, we don't really know, we have a bunch of ideas. We're just really into crypto and we want to be a modern company. And we hear so many of these partnerships and I feel like it's still kind of unclear what it is that you can do with a blockchain like Tezos versus traditional technology or financial architecture.
Arthur: (19:36)
Yeah, absolutely. So I think for me the most basic use case and maybe the most important one is a censorship-resistant store of value, like being able to store wealth in a way that isn't anyone's liability. Because if you think about it, if you have a bank account, that's still the bank's liability. If you have a brokerage account, you probably don't even own the shares in name. And even if you did own the shares, it's still the company's liability, right? There's very, very few assets which are not someone else's liability. You have real estate, but it's not easily portable. You could own precious metals, but again as bearer assets they’re kind of complicated. So it's unique in this respect, it’s a bearer asset that you can transact with across the world.
So that second use case is making very easy cross border payments, very cheap cross border payments without intermediaries. And I think, you know, if you live in a reasonably free country, it might not seem like like a big deal, but as soon as repression starts start turning on, you are pretty glad that you have a censorship resistant stores of value. That viewpoint often gets derided because people will point at people holding Bitcoin all these other cryptocurrencies and say, well, look at them, holding Bitcoin on exchanges. They're not concerned with that. Or maybe they're not, but maybe they think that other people are going to be concerned with that. And they help set a price for the store value. So I don't, you know, it's basically either using the store value or speculating on the fact that people are gonna use the store value, and I think that's completely fine.
Tracy: (21:03)
So here's my other very broad question, which is, do people actually care about decentralization in crypto? And then secondly, how do you measure decentralization? And again, I asked the question because this happened again at Milken, but Cathie Wood from ARK was talking about the new thing they're working on is some sort of systematized way or systematic way to actually measure decentralization in crypto. So how do you actually do it, and does it matter?
Arthur: (21:31)
So if you're looking today at the behavior of, you know, price setters in markets, whoever it is, the marginal buyer for cryptocurrency does not care about decentralization, right? That's been pretty clear. But I believe that for these systems to present some value, they need to be decentralized. Otherwise they'll be outcompeted. You know, if you have some use case for your blockchain, you have to ask yourself, you know, what do I do that Amazon cannot do better as a service. And I think the answer to that often time is going to rely on decentralization. So there's been a decoupling in the sense that people don't care, but at the end of the day, when the rubber hits the road, decentralization does matter. But I think we're gonna need to see more cases of blockchain being sufficiently decentralized and running into trouble for people to start caring.
Tracy: (22:20)
How do you measure decentralization?
Arthur: (22:22)
There's many ways to think about it. You can have decentralization where there's no single party is in charge. But then the fact centralization is like, not only is that true, but on top of that, you know, no party is too prominent or has too much of an impact on the network. And then you could be looking at it in terms of the most narrow sense would be the consensus algorithm, you know, who can attack the integrity of the chain, but it could also be, you know, who has influence in this network — who has the clout to suggest a hard forks, for example. And it could be who is building infrastructure that's critical for this. So you have many, many degrees of that.
Tracy: (23:05)
So just on this topic, I mean, there do seem to be benefits and drawbacks to decentralization. And we've seen, for instance, with Ethereum, Ethereum has a figurehead in the form of Vitalik Buterin and they have been able to do some stuff relatively quickly that probably a truly decentralized protocol would have difficulty doing. And I think Tezos has a unique governing structure where you are quite decentralized from what I understand. How do you feel about that? Has that been a drawback or is that a big benefit of the protocol and how do you weigh those two things?
Arthur: (23:42)
Oh, well, I think in general, decentralization is a cost, right? And there's benefits associated with that cost, right? So I think it's worth paying, but it feels almost like an insurance premium. You pay the insurance month after month and you say, why am I paying this insurance? And then you have a fire and you're glad you paid for the insurance. So it's very easy to see the cost of decentralization and the benefits are not always as obvious immediately. In terms of Tezos, we have a governance procedure that allows for upgrades to the chain. So anyone can propose an upgrade on a Tezos chain and it's sent to a vote. And if the vote passes and it's a very conservative vote — it happens over three months. But if the vote passes, the upgrade is adopted.
There's a side effect though, because in order to have a mechanism that completely automated upgrades through voting, a lot of engineering work the very beginning, went into making easy upgrades. And just as a side effect of that, that's, a benefit, you know, unrelated to the decentralization is just the fact that it's really easy to do hot swaps of the protocol has let us upgrade faster. We've had about nine upgrades in the past three years and we have the 10th upgrade being voted on right now.
Joe: (24:49)
I want to go back to Tracy's question about use cases. Because okay, store of value, I get it. Cross border payments, or payments, I get it. But on the other hand, I kind of think Bitcoin has solved those problems. To some extent. And so when I think about like a smart contracting platform and when I'm here at Milken and everyone is talking about, we're going to be trading real estate on the blockchain or virtual worlds and, you know, virtual land, where like, you know, your cartoon ape can live or whatever. I want to push further on like what can we do with a blockchain that we can't do with a traditional database and what that's gonna look like and what applications are gonna take off and actually change society or as you put it, and I really like the phrase and I'm gonna steal it from now, but like what are the blockchain’s exports going to be?
Arthur: (25:44)
Yeah. And I think, you know, the most natural match is the store of value. And then we go down from there. But there’s still the value that’s prsent. So there were a fair number of security token offerings on the blockchain. And a lot of them have come from the real estate world and there were a lot of people putting real estate on blockchains and they weren't doing it, you know, it was not an innovation, now the bank saying like, oh, we gotta do something with blockchain. You know, they came from people who actually had the need and actually had the use case and they saw the value. And I don't think it's impossible for a centralized system to do this, but having a system that secure, that's automated, that's global. I think the global aspect, the fact that it just works and you have this entire infrastructure around it. You have wallets, you have custodians, think of it, some sort of layer that plugs in into a large ecosystem that supports it, that lowers your cost. So there's value here. It doesn't come straight from the decentralization. I think it comes from the network effect of having a generic platform where you can build all sorts of things.
Tracy: (26:58)
So I have a dumb question, and I think I've asked this before, which probably makes it extra dumb, but when you actually tokenize something like real estate or, you know, a JPEG that becomes an NFT, what exactly are you buying? Because my understanding is you're basically buying a database entry and then there has to be some external body, right? Like an OpenSea that points you in the direction of it.
Joe: (27:23)
Well, and also just to add on to it, in theory, there needs to be some legal recourse, right? Such that the owner of that token actually is entitled to the cash flows of the property, whether that's in the rent or in the sale of it, as opposed to whoever just happens to be there takes the money.
Arthur: (27:40)
There's a lot of argument around this space, and I've been guilty of that, that if you follow the argument, you will reach, you know, very strong conclusions. And then there's always some subtleties, so like, oh, this could be centralized and therefore there's no value. And so for example, the legal recourse is a really good example. Sometimes you'll hear, look, there's no point in having tokenized real estate on a chain because at the end of the day, you're not going to be able to assert your right -- you'll have to go to a court, you'll have to go to like whoever the issuer is, they can refuse to honor it. So all the security of the blockchain is for naught. And I think that's taking the argument way too far.
You know, there's a saying that possession is nine-tenths of the law and in some sense, smart contracts is automated possession. So sure. You could have legal challenges. You could have all of all sort of things, but it reverses the burden. Smart contracts reverse the burden of the lawsuit. And that's really interesting because you know, you'll get transferred the title of something. You can start doing things because, “hey, you know it works with lending protocol. It's integrated in everything.” And if people somehow think that, you know, you're not entitled to it, they have to sue you and they have to claim it back, as opposed to you saying like, ‘Hey, I, I need this’. So, and you see this in the traditional financial system, in the form of escrow, right. If you have a small party dealing with a large party and they say like, look, if you don't pay us, there's no way we're gonna be able to sue you. Right. So you use escrow for that. And I like to think of smart contracts as automated escrow.
Tracy: (29:07)
Escrow exists because the two parties don't trust each other. So the solution that you're creating is basically trying to fix that trust problem. But I guess my question is, it goes back to that. What are you buying aspect? So I buy tokenized real estate. I still have to trust the person who's selling it to me because there still has to be an entry somewhere. Or there still has to be a centralized party that's telling me that this database entry points to that thing over there.
Arthur: (29:40)
Absolutely. But I still say you trust them with less than if you were just signing paper documents. And in some sense with the JPEGs and the arts, I think it's more trustless in some senses with it, because there's nothing tangible, right? It's not like somehow, they can say, no, you don't own that thing. Because what you're buying is, I think if you buy a piece of art for example, and you know, just to be sure, a lot of people buying JPEGs are just buying it because they see it go up. And they say like, I'll buy it and then I'll sell it tomorrow. There's a lot of gambling associated with it. But there's also genuine art community.
There's like genuinely digital artists who are minting digital art and selling it. And the argument I like the most here is, okay, so digital art is a saying, and that's not going go away. People want to collect it. People always want to collect art. So what's the alternative, you know, before that people would collect digital art and they would receive a paper certificate saying you own this piece. You're not gonna put it on some corporate database. You want something that's going to be here for the long run. You want something that gives you some meaningful form of ownership. So the best [thing] for that is a blockchain. So I do think there's really good match for art, digital art and blockchain.
Joe: (31:07)
Let's talk a little bit more about what the future looks like. And you know, Tezos is an Ethereum competitor and a lot of the things that you could do on Tezos you could do on Ethereum and presumably on Solana and Avalanche and the Binance chain and Algorand etc. In your vision, is there one chain that wins out? Is the goal for Tezos to win? Are we going towards one chain that's the winner? Or is it, as they say, a multi chain world?
Arthur: (31:41)
So again, to my point, there is absolutely arguments where you can say, well, you know, there's network effect in having one chain. You have more security, you have more assets, more composability. So one chain wins all. But there's limits to these absolutist arguments. One way to think about it is if you have salts in a solution and then you let it crystallize, if you crystallize it very quickly, you get something like like glass and if you crystallize very slowly, you get a few big crystals.
So the lowest energy state is just like one big single crystal, but you're not gonna get to the lowest energy state. Because at some point you'll have things that crystallize. So what it means is that, yes, we have a multi-chain world, but the applications on the chain will stay because they have their own network effect.
So I don't think we're gonna have chain that specialize on a use case in a technical sense. There's no way you can meaningfully say, oh, this chain is the best technologically speaking for sports. And this change is the best technologically speaking for finance. The designs are largely going to converge. However, if you have a lot of really, really popular applications, which are talking to each other and form their own ecosystem on a chain, they're not necessarily all going to migrate away from one chain to another, just because they can get more security in doing so.
Tracy: (33:00)
Can you talk a little bit more about that interoperability point? Why is it so technologically difficult, you know, give us a really like, easy to understand explanation of why it's so technologically difficult to make the chains work together?
Arthur: (33:16)
The main thing that a blockchain does is maintain consensus, right? That's a hard problem that they that they solve. And if you want to have two chains communicate with each other, they need to be in consensus with each other. I mean, each chain needs to know what state the other oen is in. But to have them be in consensus with each other, you would need to have all the validators of one chain be all the validators on the other chain. And by the time you've done that, you've essentially merged the two chains. So fundamentally, bridges that port state from one chain to another are hard. You are either increasing the computational cost on your network or you're losing security properties.
Joe: (33:54)
And some of the big hacks. And I don't think we've talked about it on the show before, but some of the big hacks that people hear about in crypto lately have been on these bridges, right? So can you talk a little bit about what's going on, why is security inherently difficult and what have exploiters discovered about the weakness of those bridges?
Arthur: (34:18)
So, first of all, the bridges are big honey pots, because in general, the way a bridge work is you're gonna put all your assets on a chain in one little pots, and then you're gonna mint a representation of those assets on the other chain. So you have a big pot of money that's sitting there. So first of all, you are a really good target for hackers.
Joe: (34:37)
So I lock up say $10 million worth of value on Ethereum. And then I mint $10 million on Tezos that represents a claim to that $10 million on Ethereum. So that’s a lot of money.
Arthur: (34:55)
Yeah, but that's what we’re saying, it's like, you know, what motivates security attacks is money, right? And the second thing is, so it depends, not every bridge is created equal and then some bridges have a way more way secure design. If you look at, you know, the entire Cosmos ecosystem is built around the idea of having these bridges — light client bridges between chains. Now that's more secure than other approaches.
But a typical approach is you're gonna have a set of signers. So a set of people who are gonna monitor both chains are gonna see what happens. And then they will sign messages saying like, “yep, I got deposit on this chain.” And you notify the other chain that deposit happens and vice versa, but those signatories have the power to take away the funds because they could pretend that the deposits happened that never happened or that a withdrawal happened that never happened.
They need to, so they need to be in consent. They need to follow the two chains, but there's a lot of attacks you could do. First of all, you know, if you have, I don't know Polygon, for example. So Polygon is an L-1 chain that markets itself as an L-2 chain on Ethereum. They have 5 billion.
Joe: (35:30)
Sorry, explain that quickly.
Arthur (35:35):
Yeah, Polygon is an L-1 network. And the proof is that they are actually investing in L-2 solutions. You know, if you're an L-2, you don't invest in L-2 solutions for scaling yourself. But a lot of their marketing initially was like ‘no, no, no’ building on Polygon is just like building on Ethereum.
We're all part of scaling Ethereum, but their incentives are at odds with the Ethereum ecosystem. It's been a good growth hack to say we're complementing Ethereum, but everyone's competing. And so the Polygon contract is five out of eight multisig and it has like a few billions of value. Like if you're holding a signature and you…
Joe (35:45):
So there's eight people. And in theory to access those billions, you just need to get five of them. High stakes.
And it's high stakes and where are you gonna put the signatory? You're not very gonna put tit in your apartment. You need to be able to access it. So you need some sort of like probably some data center, but even if you have a data center., you have to think about who has access to it? When there's billions, if your attack is worth billions of budget that you have for actually tricking things, I mean, you have people being kidnapped, for example. People have been kidnapped for a lot less than that. So that's, you know, there's a lot of risk associated with running a bridge.
If you look at Axie Infinity, for example. So they're not on Ethereum, they're running on the Ronin bridge and this is a bridge and they were hacked for like $800 million or something like that. And it's easy to look at this and say like, “hahaha, you know, silly NFT games, they got hacked. They're probably sloppy with security.” And I have no idea what's, you know, who’s backing it, but I've seen a bunch of people say like, apparently the evidence points at, you know, them being North Korea, because, you know, 600 million for North Korea, that's meaningful money. But if your adversaries are nation states that puts the security, and you don't really have the security problem, like the way that blockchain are designed, like the security works a lot better, you know, everyone's responsible for their own safety. You don't have this giant honeypot of all the funds of the one bridge which are in the same place.
Tracy: (38:17)
This actually reminds me of something that I wanted to ask you. And it gets back to the multichain versus absolutist argument or tension. And I guess Tezos has been around for a long time. And, you know, you've been doing this for a while. You're well known in the industry, you've talked to a bunch of people, you have different partnerships, what is it like actually going out and selling the technology to a company or convincing some sort of entity to use it? What is that, I guess, endeavor like, and how do you compete against something like Ethereum or Solana or whatever?
Arthur: (38:54)
Yeah. So, you know, in general, I don't go around to companies who have no interest in using blockchains and tell them, like you should be using Tezos.
Tracy: (39:02)
I'm sure some blockchains do that. I mean, there's probably a few of them at this conference.
Arthur: (39:07)
Yes, it's quite possible, but, you know, by and large people are, you know, every large institution has at least, you know, every bank, since 2013 has like an innovation center that actually wants to do something important. So there's already, you know, like at least some existing interest and some ideas of doing something.
So it's more about like, you know, why should we use Tezos as opposed to any other blockchain? And then, you know, we rely on some of the good attributes of Tezos, which is very strong decentralization. The fact that we have a really good software stack for building secure application, it's I think a lot easier to verify the security of contracts written on Tezos and on contracts that use Solidity, good developer community. Sometimes it can just be handholding. Sometimes people say like, you know what, I've tried to build on this chain and I couldn't find any help. And being able to go to these people and provide technological expertise makes a big difference.
Joe: (40:01)
I want to get your take on another big thing that's going on in crypto right now that's attracting all kinds of controversy and Medium posts and tweets. And that is the chain Luna. So in my understanding, maybe you you'll explain it better, but there's a chain called Luna and they have a stablecoin called UST and you can get 20% on UST if you buy this coin and then like stake it and people are making tons of money and it's going up.
And then also, I think the Luna Foundation is buying a bunch of Bitcoin, like billions of dollars worth to build this war chest, to hold the peg of the stablecoin, because to hold a stable coin' peg, there's a lot of interest in stable coin architecture and design. And some people think like this is like gonna blow up. And someone once told me, “like all the stuff that you're worried about with Tether, you should be looking at Luna” and other people like super bullish on it. What is your read of this situation? And I see more talk about this story than a lot of other things right now.
Arthur: (41:03)
So my theory is “no convertibility, no parity.” So if you cannot convert your assets from from your dollar stablecoin into dollars you're not gonna have any form of parity. And it's such a powerful law. I mean, you even saw it in equities. The classical example is Royal Dutch Shell, where after a merger they had shares trading in Amsterdam and shares trading in London. And they’re the same equity.
They give shareholders exactly the same rights and one traded, you know for years at like a 15% discount to the other. They’re the same instrument, but you cannot convert one into the other. So if you don't have, you cannot just hope that people are gonna be rational, you need arbitrages to be able to come in and use the arbitrage. So with a lot of fractionally-backed stablecoins, the arbitrageurs can do that to some extent, but the system relies on constant growth in the number of people who want to hold a coin. And if you don't have this growth and if you have, you know, as soon as you use your contraction and people start asking for redemption, then the system collapses.
So that's, you know, how Basis, for example, was built, or wanted to be built. Luna is a little different. What they're doing essentially is they're saying, “Okay, so we have an idea of the price of Luna in US dollars. And so anyone who comes in can issue some against UST. And we're going, you know, to pay 20% because that's great.” You know it's a way to grow. The difficulty with that is that there's no limits to how much UST can be created. And the Luna holders are actually paying for the 20%. You know, the 20% is coming from somewhere now. They’re happy to do it because they're saying, well, you know, the growth in the demand in the stablecoin creates visibility for the ecosystem.
It creates hype and it creates market cap, but there's limitations to how that works. What they don't do, which you need to do is you need to be able to set the interest rate and in a way that's going to balance the supply and the supply should be determined by what's backing it and the demand, which is, you know, how many people actually want to hold this stablecoin. And they don't have a mechanism like this. And instead of building this mechanism. Because you could do that. You could do something where you say, “Look, we want to have no more than 30% of the market cap of Luna being, being into UST.” And if we have more than 30%, then we'll lower the interest rate, including, you know, potentially going negative. And instead of that, they're going and saying “No we're gonna buy Bitcoin.” Which is, you know, if you wanna have also some centralized ownership at and go and buy, you know, go and buy some dollars.”
Tracy: (43:49)
What's the rationale for doing, what do they say is the rationale?
Arthur: (43:51)
I think the rationale is headlines. The biggest holder of Bitcoin, how cool is that? Or potentially it’s the idea that, well, maybe Bitcoin will appreciate. And that helps us with our backing problem.
Tracy: (44:02)
Yeah. Since you mentioned arbitrage there I was wondering, given your ETF market-making background, do you see a lot of parallels between the world of crypto and ETF trading,
Arthur: (44:17)
ETF trading, not per se, right? Not directly, but just having a general financial background, I think is helpful when looking at a lot of DeFi protocols and understanding how they work or don't work.
Joe: (44:31)
I have a question. And again, it's sort of about these parallels. So one, people got hyped up about blockchains early 2021 when the Robinhood stuff was going on and people started learning about payment for order flow. And they're like, oh, you're giving a penny or a millionth of a penny on every trade to some high frequency trader jumping ahead of you and so forth, but crypto or blockchains themselves and Ethereum and other smart contracting platforms have this concept of MEV and minor extractable value. And if I wanna place a trade, you know, blocks only happen every once in a while. And in theory that trade goes out there and everyone can see it. And a miner can jump ahead, I think is how it works, essentially jump ahead of me and get a better price on that execution and then sell it back to me. And I get a worse execution. How big of a problem is this? Can you talk like, and how does it work on uh Tezos and what are the sort of things that people should understand about, you know, the power that miners have or how MEV comes out of the system?
Arthur: (45:37)
That's a big question,
Joe: (45:39)
I know it's like, these are things that we haven't really talked about before, and you're good at explaining these things. So I figured I would just ask.
Arthur: (45:44)
I’m happy dive into it. So the first thing to understand is that order flow is not fungible, right? Order flow that comes from retail trading is not the same thing as order flow that come from Goldman Sachs, or that comes from a fund manager. You know, if I come to you and I say, ‘Hey, you know, I want to buy this, uh, this equity from you, these shares from you. Sure, sure. You know.’ But if Goldman Sachs come to you and say, ‘Hey, don't you wanna buy this from us?’
Joe: (46:11)
So like, retail flow is much more desirable because we're all dumb.
Arthur: (46:15)
It's not even that you're dumb. It's just not informed. You're buying, you know, you're buying it because either maybe you're gambling or because you're saying like, I like this stock, I wanna buy it for retirement. And so on and so forth, this is not like you have just received some information from a wire for something, and you've done all of that. And now you know that the execution is going, the price can go down. So it's quite different. And once upon a time, you didn't have to mix all the order flow together. So if you were a big fund manager and you have this large S&P fund and you need to rebalance and you talk to another big fund manager and they need to rebalance, and then you strike a deal because you know that the other person is not trying to screw you. Right. They're trying to...
Joe: (46:53)
They're trying to solve a mutual problem.
Arthur: (46:55)
Yes. And also you're gonna do repeat business with them. That's quite important. And that's, you know, that happened upstairs in NYSE. And then at some point SEC looks at this and says, well, you know, we see all these big trades happening and they're very advantageous, but…
Tracy: (47:08)
This is RegNMS, right?
Arthur:
This is RegNMS, yeah. The retail guy doesn't get access to that. So we're gonna put everything on the electronic exchange. And brokers will have to route to whichever exchange has the best, you know, the NBBO, the best bid and ask. And a lot of individual investors when they start doing that, they had no idea how to do it. So now they start showing these big orders, or even, you know, they chop them up, they don't really do it in a way that hides the amount of volume that they want to do. And they get eaten by high frequency traders who appear as a cottage industry as soon as RegNMS is passed. And so a lot of these, big sell-side people get very, very upset at it. And if you read ‘Flash Boys,’ it's a very one-sided book that talks only about the sell-side doesn't treat a single person on the high-frequency trading side...
Tracy: (47:56)
I remember there was this crazy bit in ‘Flash Boys’ where he was talking about, he was about to do like a retail trade and he was about to click like order or buy. And then he saw the price move on his screen and he was convinced that it was because, you know, Goldman Sachs was front running his tiny retail trade.
Arthur: (48:15)
And it's bizarre because it's also presented, I think like in 2010 as like, oh, I'm uncovering this big secret. And I'm like, I started working in Goldman Sachs in 2009 because people were talking about this on TV and I saw it was cool so this is, you know, it wasn't CNBC. It's not like they were uncovering a big conspiracy. And, you know, people adapted, of course. So they start using execution brokers, where, you know, they will say, no, we'll take your order, chop it up and then put it on the exchange. But so now as a result, what happens is you’re getting transition of flow. So you get this retail flow, right? And if you send retail flow to the exchange, the retail is gonna get a worse price because the exchange doesn't know anything.
The exchange says, whoa, whoa, whoa, wait a second. We don't know who you are. You could be retail. You could be a hedge fund. We're gonna be giving you wide quotes because we don't know. Whereas as a broker, you know, the order flow is not toxic. That's a word that's used for like, yeah, informal flow. It's not informal order flow, so you can give it the better price. So you're gonna match it internally. You're gonna give it the better price in the NBBO, which you're allowed to do that. You're gonna report the trade to the exchange, or you're gonna give the customers NBBO but actually you internalize it for something better. So the customer feels like, oh, I haven't paid anything, you know? But actually you've made money because you can quote tighter on your internalized flow than you could on the exchange. So retail really win out out of this, like the small retail guy who's like just buying single in stocks. Market's much more liquid for that. I would say the losers out of this have been non-institutional traders who now have to either use execution brokers or dark pools.
Tracy: (49:46)
So here's a very broad question based on that very long and detailed answer. But what is crypto trading like behind the scenes of exchanges? Like how transparent is it and what sort of execution are people getting ?
Joe: (50:02)
Right, and all the MEV aspect, right? How big are some of these?
Arthur: (50:08)
It's equivalent to me is putting your giant order on the exchange and everyone knows, you know, you're about to do. And so people are gonna get ahead of you. The difference is that if I put a giant order with my broker, at the very least, the broker has a fiduciary responsibility towards me, you know, I’m his client. And so they have to do things a certain way, whereas I don't have any fiduciary relationship with a miner. Now there's nothing stopping people from building these relationships. You know, as a trader, you could go to block producers and say, I want you to promise me that when I send you an order, you're never going to include, you know, you're never gonna show it to anyone else. Like you could do that. People don't do it.
Joe: (50:48)
Do you think that's gonna happen?
Arthur: (50:49)
No. No, I think what's gonna happen is people are going to build better protocols, which are going to remove some of the extractable value.
Tracy: (50:56)
Do you think regulation will come to the space? Like will the SEC for instance, get in interested in best execution for retail traders in crypto?
Arthur: (51:05)
I think it's possible because people have made MEV some sort of moral issue, which I think is dangerous. You know, the world saying like, oh, the miners are exploiting people, right. They're taking advantage and so on and so forth. One thing that I think would be more helpful to frame as, and again, I said is like, if you want a fiduciary relationship with your miners, you should ask for it. What I would worry about is a regulation that says that by default as a miner, you have a fiduciary responsibility with people, which I don't think you should have.
Tracy: (51:36)
What about if I have a fiduciary responsibility with a brokerage, that for instance, is providing me crypto exposure through my 401k, as I think someone is doing now, I forgot the name, but, you know, various people are starting to do this in a sort of more regulated way. How does that fit into the execution.
Arthur: (51:52)
Well, in that case, of course, you know, the broker should not like favor another client or themselves, you know, in doing the transaction. And just to be clear, this happens for on chain trading. When you use the right protocols, it's not actually an issue with trading on centralized exchanges. I think, you know, first of all, a lot of it comes from the latency of change and the more latency you have, the more MEV there's going to be, because quotes are not adjusted in real time. So it's partly a function of that. And the way in which protocols have been designed, if you are in a system where you can't address codes really rapidly, you need to build batch transaction mechanisms. And they're harder to program, but once you have this, you can reduce the MEV on the exchange a lot. And I think we'll see that happen.
Tracy: (52:38)
So just going back to the turtles at the beginning of this discussion and plastic bags on the blockchain. I was packing up my apartment in Hong Kong a few months ago, and I noticed that a bottle of balsamic vinegar that I had that I had never used, it was on the blockchain, right. You could trace it to test its validity and big existential question here, but is the future just everything on the blockchain and how do you differentiate between the importance of having something like plastic bags being tracked and traceable versus having something like a store of value that people are using to hedge against inflation or for whatever reason.
Arthur: (53:20)
Well, I think there's more value in the store of value aspects than the supply chain applications. A lot of the supply chain applications we see on blockchain honestly, are stone soup, you know, you'll see, IBM go to these people and say, put everything on a blockchain, but if you digitalize your entire chain, right, where you have, you know, like you can scan bar codes and do all of this and put it in a system, and then, you know, like you've, solved 99% of your problem. And then you put it on blockchain, you know, why not, but that's not really what's driving the value, I believe there's value in digitizing and having cryptographic signature on supply chains. Absolutely. I don't think the blockchain adds a whole lot to this.
Joe: (53:55)
All right. Well, Arthur Breitman of Tezos thank you so much. That was a great conversation. I feel like we have really good conversations with like the people who actually did some time in the TradFi trading world. That seems to be like the sweet spot for people who can explain things. So thank you for coming on Odd Lots.
Arthur: (54:13)
Thank you for having me.
Tracy: (54:14)
Thanks so much, Arthur, that was great.
Joe: (54:29)
I guess I said that at the end there, but I do feel like the best explanations and conversations we have on this space are often with someone who did some time trading in some legacy institution and can sort of, I guess I would say like translate some of these ideas back and forth.
Tracy: (54:47)
Absolutely. But I also feel like there's a tendency in crypto for people to talk about all these issues as brand new when in fact they have been very diligently thought out many years before by traditional financial institutions. And there's a reason why, you know, they do it a certain way, but execution, for instance, is something that's been top of mind for a lot of people for decades.
Joe: (55:08)
I still think people should just chill out. It's like, it's just a penny. Or it's like, it's just a fraction of a penny. What's the big deal.
Tracy: (55:14)
The other thing that I liked from that conversation was the description of DeFi. You know, the question of what is the export here? And we've talked about it before, but like, what exactly are we doing other than dealing in more DeFi?
Joe: (55:26)
Well, you used to make fun of me because early on in some of our first DeFi conversations, I would be like, well, the whaling expedition and it's like, is the protocol actually, you know, it's like VC got its roots, kind of, in funding whaling expeditions.
Tracy: (55:41)
I would never made fun of you. I never made fun of you, I just questioned why you were obsessed with whaling expeditions all of a sudden.
Joe: (55:46)
Fair enough. But anyway, you know, as Arthur pointed out, it's not that exciting if it' just tokens trading tokens and more tokens. Or it's not that world changing
Tracy: (55:59)
Or boxes being opened.
Joe: (56:01)
Tokenized boxes. Yeah.
Tracy: (56:04)
All right. Shall we leave it there?
Joe: (56:05)
Let's leave it there.
You can follow Arthur Breitman on Twitter at @ArthurB.