Berk Ozdogan: Wall Street Trader Turned Crypto Innovator

0:00Your favorite neighborhood spot grows with Square. Indeed, my favorite neighborhood spot has quickly become Todd Snyder in Williamsburg. Todd Snyder is one of my favorite menswear shops and has supplied me with all the clothes I have needed this quite hot summer. Every business has different goals, but Square is the business platform that supports them all. From opening a new location, selling something new, or just expanding their reach. Indeed, I've seen it with Todd Snyder. In Square, also, you can get real-time insights, so don't wait for end-of-day reports. Go to square.com forward slash go forward slash realvision to learn more about how your business can grow with Square.

0:43That's S-Q-U-A-R-E dot com slash G-O slash R-E-A-L-B-I-S-I-O-N.

0:57Hey everyone, if you like this podcast, go behind the paywall to get privileged access to the smartest minds in finance. Join the Real Vision community and learn how to become a better investor. Visit realvision.com slash RVpod and use the promo code podcast10, that's podcast10, to get 10 % off our essential membership for the first year. Now, to the top analysis of today's crypto markets.

1:29berk os dogan dexalots head of trading welcome to real vision crypto daily briefing hey ash thanks for having me how are you i'm doing great lots to talk about berk it's a pleasure to have you here uh decentralized exchanges dexes uh and much much more to cover in this conversation but first let's take a look at price action here on the day bitcoin well it's lost its 27 handle trading right now at 26 ,911, down 2.7, 2.8%, call it trailing 24 hours, trailing seven days up about two and a quarter percent. Ethereum, we've lost the 19-handle trading right now at 1 ,858 on my screen, trailing 24 hours off about two and a quarter percent, let's call it trailing seven days.

2:14It's up around 3.7 % over the last week or so. Berk, it's great to have you here on Real Vision Crypto. Much to talk about. We were chatting a little bit before the show got started. We both spent a bit of time working at banks. Tell us a little bit about your background in TradFi before we get started talking about crypto. Yeah, of course. First of all, thanks for having me. Pleasure to be here. Obviously, I'm a big fan. My background, I'm an engineer by training, of course, spent 10, 11 years in Wall Street across a few large institutions. I was last at JP Morgan, part of the FX trading team, specifically focusing on the algo trading business.

2:56Essentially, we built the software that institutional clients would use to execute their very large size FX transactions, you know, day in and day out. Obviously, one of the most liquid markets. And since 2016 or so, I've been quite involved in crypto. You know, obviously first as a hobbyist, enthusiast, and eventually now full-time. And, you know, glad to be here and looking forward to this conversation. That's how, you know, that's why I'm here now. So how did you get involved in crypto? What was it that sparked your interest as a former TradFi guy, what potential did you see there for doing all the things that we're about to discuss?

3:41Yeah, funny, multiple stories, actually. So in 2016, I had a reasonably long commute. I live in a New York City, New Jersey area. Commutes are usually 30 minutes or longer from where I live to work. And I sort of was like, hey, this seems like a new asset class. Let me get on some podcasts. One podcast, particularly from a gentleman named Emin Gunsiraj, the founder of Avalanche, back then was a professor of Cornell of distributed systems, changed my entire perspective. It was about multi-sig, non-custodial wallets that basically allowed for the non-custodial relationship to interact with various things, but also had an additional security future.

4:24I was like, huh, this is interesting and may have real applications in traditional finance. One thing led to another. I fell down the rabbit hole, the proverbial rabbit hole, and I actually left JP Morgan in 2018 timeframe, specifically to go work on what's now known as DeFi primitives, right? The most interesting thing I've done then was basically a decentralized exchange, very much like a NASDAQ or NYSE or Coinbase on the Ethereum blockchain. Obviously, this is well before Ethereum 2.0. This was right around the big ICO boom. And basically, the team that I was part of built a ground-up club, central limit order book, on Ethereum.

5:16It was intending to trade what's known as security tokens, and the user experience didn't quite work out. I ended up going back to JP Morgan, but with the full vision that whenever blockchains and whenever specifically smart contract capability advanced to the point where it can rival traditional exchanges, that I would eventually go back. That was kind of the idea. Funny enough, in October of 2020, I pitched a Generation 3 blockchain-based central limit order book exchange idea to then the head of macro trading at JPMorgan, whom I reported to effectively. And while the idea was very intriguing to everybody that I talked about, the unfortunate reality at the time was not enough regulatory framework in place.

6:10And even if we build it, great, we're going to make, you know, I can't remember the exact numbers, but it was like we're going to make 50, 100 million something. And we're going to get hit by like a potential regulatory penalty in the billions, right? Right. Therefore, let's sleep on this. Right. That was sort of the that was the answer I got at the end of 2020. Of course. Right. As 2021 rolled around, other businesses started showing interest. And, you know, similar to other banks, JP Morgan also started building things. But as you can imagine, without the regulatory clarity, you know, with what SEC as well as other regulators are doing globally, you know that project is kind of you know was dormant at which point yeah which by the way we should point out when you say sec and other regulators they are not currently at least in the united states harmonized on the view of how this industry works uh by the way i very much like the way that you started that uh which is uh sitting in george washington bridge traffic on route 80 is enough to change anyone's view of the world i know it did absolutely that's exactly how I was trying to consume as much information outside of my own world as possible because, look, these things are real.

7:25And, you know, people, there's a lot of people out there that are complete nonbelievers. But I'm one of those people that this type of technology, once it can become the fundamental building block of some other service that I use, it's going to change things. Right. And that's how I ended up where I am now. Full time in crypto, if you will. Let me ask you this. What was it that really caught your attention the most? Was it the potential to trade digital assets natively on these decentralized exchanges? Or was it the idea of what might be coming down the pike in terms of the digitization of everything, tokenization, the ability to trade securities tokens?

8:02Which was it that really caught your eye from the perspective of use cases? Yeah, quite honestly, it was not tokenization for me. Like tokenization makes a lot of sense for certain use cases. I'm not one of the folks that believe everything should be tokenized. I think there are a lot of illiquid markets that needs to be tokenized, and eventually there will be. But what really caught my attention was how trading, as we know it in traditional finance, could change if we integrated blockchains into the business. And the answer is not a single thing. It's multifold. Number one, I'll talk more about the recent stuff, right?

8:41Think about the FTXs of the world. I'll think about like the exchanges that disappeared, created losses for people, created a lot of headaches for regulators. Those problems can actually be solved by transparency, number one, and then non-custodial relationships, number two. Transparency, because if you actually have full transparency on where trading activity is taking place, which wallets have what positions and who's doing what, I think it levels the playing field. You no longer have individual private companies potentially doing bad things with other people's money. That sort of becomes extremely difficult.

9:25Second, regulators, and believe me, I've been at the end of a lot of regulator questions in my time in Wall Street, will have a significantly easier time looking at data and understanding what's actually happening in the markets. The way this worked, at least for me, in TradFi is we had a lot of algos. They would go out in the market. They would make certain decisions and execute. And then a few months later, maybe a couple of years later, you would receive a question saying, what did your algo do here and why did it do it? Explain to me and make sure it's not nefarious. The reason why they had to ask those questions were, you know, in OTC markets, information wasn't as readily available, some of your liquidity trades and dark posts, etc.

10:08And it was making it very difficult for regulators to get a true picture and therefore regulate with, you know, a lot of information in their hands. I think blockchain-based trading, if and when it becomes as efficient, you know, that's my caveat, is going to make it much easier for these markets to operate. Those were the two things. And of course, the non-custodial angle where I don't have to put my assets on a specific platform that is owned by a private company, which makes me sign all sorts of, you know, subscripted contracts, you know, is an added benefit in my opinion, right? If I keep, if I continue to keep my assets in my own wallet, I think that's a big plus.

10:52And those were the things that sort of made me believe that this is the way forward. Hey, everyone, we're going to take a quick pause and hear a word from our partners. We'll be right back.

11:05So let's talk a little bit about that trajectory forward. You started off talking about building CLOB centralized limit order books. So let's talk a little bit about this transition from a traditional centralized limit order book controlled by a large financial services institution to the type of trading that we see on centralized exchanges in crypto today, ultimately through to decentralized exchanges or DEXs, which you're most interested in. Talk a little bit about what that journey looked like in terms of your own view of the space. Yeah, absolutely. The journey is continuing, to say the least.

11:44But it started out, like I said, with Ethereum and a club that was deployed in Ethereum, right? Now, if you think about how TradFi works today, right? On almost all asset classes today, you have sub-second execution. In the equities market, you actually have sub-10 milliseconds execution for most of the cases. In effects, it's sub-100 milliseconds, et cetera. And back in 2017, 18, 19, Ethereum couldn't support that kind of user experience, right? What that means is you attempt to do a trade, your transaction is going through for the next 10 minutes while the market is doing its thing and you're sitting there waiting and taking market risks.

12:29So that's where the journey started. Even though the world computer, the Ethereum world computer as people call it, allowed a developer to actually develop the matching engine as what matches people's orders. On-chain, the user experience didn't work. It wasn't possible to risk manage effectively. And quite honestly, clubs didn't work on Ethereum, right? Then, obviously, we saw very clever folks come up with the concept of a constant product formula to eliminate the need of a central limit order book. And instead, they created these liquidity pools, AMMs, as they're called in the Web3 world, which allow people to basically tie assets together, become a market maker, and provide liquidity to the market.

13:16Speaking of that shift from centralized limit order books to liquidity pools, I think about that really as a function of strengths and weaknesses, tradeoffs in technology. When you have a TradFi centralized limit order book framework, as you said, sub 10 millisecond execution times, you can't do that on blockchains. But what you can do is something that's very interesting, which is to lock in smart contracts in ways that third parties and even the parties who entered into those contracts cannot alter. So you have essentially this mechanism where you can say, OK, we can lock money so that we can build a liquidity pool with price matching.

13:52And when you have price deltas and they move, you obviously get a change in the movement of that liquidity pool. It's really actually kind of an interesting and clever solution to a challenge and also an advantage that DeFi specifically and more generally crypto and smart contracts provide. Absolutely agreed. So first of all, in my opinion, this is one of the most clever things I've actually witnessed in Web3. Basically, you have this issue with the underlying protocol not being developed enough to provide you the speed that you need so that you can execute trades quickly. And instead, folks created a mathematical solution that sort of worked around the protocol level issue, reduced the number of transactions you would need to do a trade, and simultaneously improved the user experience tremendously.

14:40like imagine a traditional exchange experience, right? You have to deposit, wait for the deposit to clear. Then you have to enter orders. And if you're entering limit orders, you have to wait for somebody to hit your limit orders. In the case of an AMM, one click, atomic, execution happens in your wallet and boom, you're done. You only suffer one click and one confirmation time. So when you say atomic, for folks who may not understand, you're talking about essentially locking down multiple sides, legs of a trade simultaneously with a smart contract in a way that would be very difficult to do in traditional finance.

15:16Exactly. I'm sorry, I should have explained that a little bit better, but exactly. With one action, everything you need for one asset to become another happens, right? And it's a significant and fundamental shift. And I would argue that was the first and one of the biggest user experience improvements in DeFi, in my opinion. Now, obviously, with the what I call Generation 3 blockchains, you have avalanches of the world, Solanas of the world, and many new up-and-comers who are pushing the boundaries of time to finality, right? Specifically, I believe clubs are a lot more feasible now, and I'm going to compare why clubs versus AMMs, but clubs are a lot more feasible now because the user experience are coming experiences coming very close to what it was in traditional finance clubs without compromising on the non-custodial aspect i.e you don't deposit your assets anywhere other than the blockchain and then of course you still have that transparency layer where you know what's happening on chain directly on chain so this is extremely interesting let me just do a little bit of translating here for the audience essentially what you're saying is that some of the original weaknesses for centralized limit order books that cause DeFi to move in the direction of liquidity pools in the first place by these what you call third generation blockchains, which have transaction finality, meaning final settlement on the chain happening much faster.

16:44In your view, if I'm understanding it correctly, you're saying essentially that this is a universe in which centralized limit order books actually can live and thrive and therefore provide more similar functionality to what we saw in ShradFi. Really interesting. Absolutely. I mean, think about how TradFi market and how electronic trading developed, right? Over the course of the 80s, 90s, early 2000s, and now, I think the market has truly settled in the most capital efficient and the most generally everything else efficient way of trading, which is central limit order box. And also think about how big the traditional finance market is, how many actors there are, etc.

17:24For them to move to this space, I think while AMMs are cool, novel, and interesting, their capital efficiency, or I should say capital deficiency, as well as the way they operate are so different with non-deterministic execution pricing, meaning once you click, you think you're going to get a price, but you actually have a margin of error, right? This is what people talk about slippage. And, you know, the important thing I think to understand here is that on a limit order book, you can do precisely that. You can place limits against individual trades and obviously a whole series of other order execution management type of parameters that you just cannot do today on liquidity pools.

18:08Exactly. And the fact that, you know, everybody already built big infrastructure APIs, workflows relating to Clobs, I think we... API for people who may not know, application programming interface. This is essentially the ability to allow computers to talk to each other programmatically in a way that you can develop an open standard, you can publish that open standard, and then you can have different systems integrate without necessarily having centralized coordination. Exactly. So think about all the infrastructure that traditional finance institutions have built. If you can give them the exact type of connectivity, communication protocols, APIs, application programming interfaces, and the general behavior, while giving them the added benefit of non-custodial relationships with the counterparties that they're trading with, plus the transparency, I think there is a real shot on decentralized central limit order book exchanges taking off, especially if you think about the events of 2022 and all the bad things that happened because of a, quite honestly, a handful of bad actors and the domino effect they caused.

19:17I expect the market and the trends to shift in that direction. Well, that plays precisely into the point that you made about non-custodial trading. Talk a little bit about what that is for the layperson who may be struggling to get their head around it. Sure. So let's talk about what a custodial relationship is. Essentially, what that means is you're allowing another entity or a participant to take custody of your assets. You are depositing your assets to a wallet that belongs to somebody else, and you're essentially trusting them to safeguard your assets and direct your assets in the same way that you would tell them to direct it.

19:55Right. You know, to further break it down, when you deposit your assets into a centralized exchange, they will take it into their own wallet and they will create a virtual account with credit for you to trade with. You know, one way to think about it is the exact link between the balances that you have on the exchange that you see on the front end of the web page and the actual balances you deposit. The link between them becomes blurry and maybe disappears. Right. That's what we've seen with some of the exchanges that unfortunately went bust where the number of assets that were deposited versus the amount of trading and the volume that was occurring was not the same.

20:36Decentralized exchanges eliminate that trust into an individual institution, and it typically replaces it with either your own wallet in the case of one transaction trading, or in the case of central limit order books, typically replaces it with a smart contract that you can always look and see the contents of on the blockchain. Hey, everyone. We're going to take another quick break and hear a word from our partners. We'll be right back to the Real Vision Crypto Daily Briefing.

21:12Yeah, and it's important to point out that the risk on this is not just technical, but it's also currently legal. In other words, If you have a deposit on an exchange and the exchange goes bust, it's guess what? Get in line and stand behind the rest of the creditors. Figure out jockey for position in terms of creditor classes, in terms of where you get paid out in the bankruptcy. That's a challenge and obviously something that, as you point out, noncustodial trading, noncustodial asset solutions more generally have the capacity to significantly, significantly change the landscape in a way that, quite frankly, is very novel to this space.

21:46And I think it's just incredibly exciting. Let's talk a little bit about the underlying technology that you guys are using at Dexalot. Why AVAX? So I mentioned this part earlier as well. Time to finality, which you eloquently expressed as time for the final settlement to occur in Avalanche is quite quick. Avalanche under the covers has a very novel consensus mechanism, which allows a blockchain and transactions to become final, as in non-changeable, very quickly. The biggest and probably one of the most important reasons for us was the fact that Avalanche doesn't have reorgs. Even in Ethereum or other Ethereum-like blockchains, you have something called reorgs where once the transaction is confirmed, until a few more confirmations happen, you have a probabilistic chance of potentially losing that transaction.

22:47I'm simplifying, of course, but Avalanche does not have that problem. And if you think about, obviously, if you exaggerate the problem and think about trading, imagine you do a trade, right? 20 minutes passes, and all of a sudden that trade, that risk that you have recorded disappears because the blockchain reorgged because there was a longer chain, longer source of truth. Somebody took over and your transaction disappeared. This was a real risk, and we didn't like it from an exchange perspective. We felt like none of our users would like it. And of course, ability to finalize a transaction within at most two seconds was extremely important.

23:26And the last reason, which is where Dexalot is now, is that Avalanche allowed scaling through something called subnets. Right. Subnets are essentially blockchains that you can spin up quickly, eliminate the sharing of the bandwidth with other projects. Imagine if you're an exchange, a very successful degen-loved NFT project drops, the chain gets clogged, right? I think I read something about, you know, ordinals are quite popular in Bitcoin today. The Bitcoin mempool is so full that it's going to take a few days to clear, right? You don't want other successes on the blockchain affecting your product.

24:07And by spinning up a separate blockchain that's still connected to the main chain to run your application was incredibly important to us. We leave there now. It's a perfect metaphor, too, because obviously the consternation that that's caused in the Bitcoin community. Essentially, you have transaction finality being delayed for all types of transactions while the mempool gets flooded with requests for ordinals or whatever else. We don't want to pick on anyone here, but it's just the premise that you can have one sort of category of transaction clogging the ability to execute across the board on the rest of the chain.

24:42Talking of which, talking about risks and vulnerabilities, one of the things that we've seen on AVAX Avalanche here in the last, call it 60 days, is challenges with performance degradation and some mini network outages on the C and X chains. Talk a little bit about that in terms of the scaling solution, in terms of the risks that it provides, but also in terms of the opportunities that folks in the Avalanche community see from this multi-chain world. Absolutely. So I actually do keep up with most of the Gen 3 blockchains. I can't say I keep up as much as I keep up with Avalanche. But Ava Labs, the team behind the blockchain, is one of those teams, particularly on the engineering side, incredibly fast in delivering additional features, right?

25:29Unfortunately, like there is a tradeoff between shipping new features fast and making sure everything doesn't break. And quite honestly, no matter how much testing you do, I think it is sometimes difficult to create the exact condition of a mainnet. And there was an outage relatively recently after an update that was done on the platform chain on Avalanche. Now, the cool thing is that subnets, the separate application-specific chains, were not affected by this outage. The outage affected the platform chain and the C chain, which the two are connected. Can you explain those two chains and what their functioning is in the network?

26:11Sure, of course. So Avalanche as a platform, I'm going to call it, consists of three main chains. First chain is what's called the X chain. It's a blockchain purely for transfers. X comes from exchange. It's not exactly a traditional exchange, but it allows people to send value from one wallet to the other very quickly and very cheap. That's the first chain. The second chain is called the P chain, stands for platform chain. And this is the chain where the network nodes and the node-related, the validator-related operations take place. This is the chain where the entire validation of the network occurs.

26:51And then you have the C chain, which stands for the contract chain, where it is the EVM clone or EVM-compatible chain. This is the Ethereum virtual machine clone. This is the Ethereum virtual machine where you deploy your software and let the validators run your software. Most of the activity, of course, occurs on the C-chain. And essentially what I'm saying is subnets are similar chains to the C-chain with specific customizations designed to support typically a subset or maybe a single application. In the case of Dexalot, it's a single blockchain that is designed only to do trading in a central limit order.

27:37Yeah. Obviously, for folks who are relatively new to this space, this can seem a little bit confusing, a lot of moving parts here. But let's talk about the general framework for how you think about the scaling of this network and what you think the potential is to start executing at scale that we can see, for example, in TradFi. You mentioned FX. We talked about equities. Obviously, this type of trading, absolutely massive scale in terms of what the throughput is currently in traditional finance. Absolutely. I mean, the speed with which this development is occurring is incredible, right? The original Avalanche consensus, which still stands today, allows you to scale to millions of nodes without significant degradation in time to finality, ability to finalize these transactions quickly.

28:30Right. That's one aspect of it, i.e., you can push through a lot more operations per second without compromising on the fact that once transactions occur, they cannot be changed. Right. That's number one. Number two is this subnet scaling solution. You can spin up as many subnets as you want. And it doesn't have to be one subnet for application. Imagine the case of an exchange. Typically, for example, you mentioned equities, right? In the equities world, traditional finance trading systems will strike their infrastructure by buckets. For example, in the equities world, the letter A, the stocks that start with the letter A, typically have higher volume.