Episode 18
Carbon: The next frontier for on-chain liquidity management

Our guest today is Mark Richardson, Project Lead at Bancor. Carbon is a new trading protocol developed by Bancor that enables advanced trading strategies through the use of on-chain limit and range orders.

In this conversation, we explore the evolution of the decentralized exchange (DEX) landscape, what problems have emerged from the existing automated market maker (AMM) model, and how Carbon was built to address these issues.



Hey everyone, this is Dawei from Exponential and you're listening to Degen Responsibly, a podcast where we invite protocol builders to showcase their innovations, how they work, as well as do a deep dive on risk. Exponential is an investment platform that makes it easy to discover, assess, and invest in DeFi yield opportunities. We want to help you understand the tradeoffs and opportunities so you can degen responsibly. Our guest today is Mark Richardson, the project lead at Bancor. Bancor was the first protocol to introduce the AMM model in DeFi and pioneered features like single-sided liquidity and impermanent loss protection. Carbon is Bancor's new trading protocol that allows on-chain limit and range orders, enabling traders to execute advanced strategies with low fees and high capital efficiency. In this episode, we'll talk about the evolution of the DEX landscape, the problems with the current AMM model, and how Carbon was built to address these issues. Hey Mark, excited to have you on to talk about Carbon and what you guys are building. But before we get started, just curious about your crypto journey, how you got started, and what led you to working at Bancor and Carbon? Yeah, so my background is actually in organic chemistry. So I completed my PhD at the University of Melbourne. I did my first postdoc at UCI in California. I came back to Australia and started working for the Australian government for an agency called the CSIRO building medical sensing technology basically. And I was there, I had a three-year contract beginning from 2019. So I spent about a year on the projects that I was working on. I had some research funding to go to Berlin and complete a pretty significant leg of work under the supervision of Peter Seberger at the Max Planck Institute for Colloids and Surfaces. And my flight out of Australia was about a week after Australia closed its borders to international travel during the pandemic. And so that wiped out a pretty significant swathe of work already completed. And the city that I was living in and living in still actually went into about a 24-month contiguous lockdown. I think it's a world record. For those that aren't aware, you know, Melbourne and the Australian government in particular were rather draconian in their lockdown measures during COVID. And so that kind of ended my research career in science, unfortunately. And so during 2020 and into 2021, I was looking for other ways that I could be useful. And so I spent a very significant amount of time studying cryptocurrency and DeFi was really on the rise. And this is sort of the beginning of DeFi summer. Everyone remembers that it more or less coincided with the beginning of the pandemic. And so I was actively involved in a large number of different DAOs. I was fairly attracted to decentralized exchange infrastructure in particular, just because I think that it's probably the most important primitive for blockchains to master if decentralized finance is going to be something that is of benefit and becomes an important component of the world's financial system. And so I was in the Kyber community, I was even a part of the Uniswap community, but I was more active, I think, in Bancorp's community, especially during the launch of its DAO, when B2.1 started to go live. And yeah, over the back end of 2020, I started writing some of the first DAO policies. And I met the Bancorp founders and continued to work with them up until January of 2021, when one of them asked me, like, why do I have so much time to commit to researching these things and working with the developers, because I was just a volunteer at that stage. And I said that I'm technically still under contract to the Australian government, I'm just not allowed to go to work right now. But as soon as the lockdown measures are alleviated, I will probably not be able to continue working the way that I have. So they reached out the day after that conversation and offered me a full-time contract as a researcher. And so I left my job with the Australian government and started working full-time as a contractor for Bancorp in January of 2021. Later that year, I think I was the head of the research department. And then in October of last year, I was asked to helm the project, essentially take over the CEO position. So it's been a rather, you know, this is, you know, career advancement on crypto time, essentially. But yeah, during that time, I've been involved with essentially every aspect of the project's development, looking at, you know, the major mathematical sort of conceptualization of what we do, and also sort of speaking to smart contract developers and front-end developers and helping to develop the products from a high level. Nice. Nice. I think that's a really interesting journey there. I think the first person I've heard that had a science background kind of moved into that crypto landscape. But yeah, before we dive into Carbon, I wanted to maybe frame the conversation a little bit today in terms of where you see the state of the current VEX landscape, what are some of the things that worked so far with the current models, and what problems still persist in your mind? I think all of the models, all of the models work. It's just a question of who they work for. So the original AMM design, so sort of the Bancor V1 days, liquidity provision was seen as the responsibility of the project that has the tokens, that's generated the tokens, you know, in a way. And it wasn't really designed to be, I guess, a financially performant device onto itself. Certainly, you know, the ability to trade with that liquidity as a taker is, you know, entirely dependent on the trader's ability to predict market movements and profit from them. But liquidity provisioning in sort of the passive context was never really anticipated as being very profitable. And so those protocols weren't really designed to function that way. And so at that time, you know, there was no front end component that allowed users to provide liquidity directly to a decentralized exchange. The first protocol to allow users to do that themselves was Uniswap. And that really sort of kicked off this idea that being a passive liquidity provider can be or should be expected to be an investable vehicle. Of course, isn't the, you know, as many have pointed out, myself included, and, you know, many other commentators on DeFi, liquidity provision and AMM is generally a losing bet. It can be expressed as a short gamma option, or this is at least one of the ways that we can interpret using regular financial lingo, the actual instrument that you're buying. And for those who aren't familiar, you can think of it as essentially offering the rest of the market a type of insurance contract, where you are guaranteeing them or offering them compensation if the price of the asset that you're providing liquidity with deviates from whatever the price was after you signed the contract. So it's not really, it was never really intended to be financially performant. There were several iterations afterwards that I think have kind of moved things in maybe a better direction, or at least have tried to give liquidity providers, passive liquidity providers to AMMs, some sort of agency over the decisions that are made, right, how their liquidity is traded. So for example, in Bancor V2, we introduced this concept of what we called amplified liquidity. And then I think about nine months after that, Uniswap came out with a similar product that they called concentrated liquidity, but it's essentially the same idea. And this is that instead of providing liquidity across an infinitely enumerated range, right, between zero and infinity dollars, instead, you can just concentrate your liquidity within a certain price bound. And if the market happens to move outside of those bounds, then you are going to be left holding the bag on whatever the least valuable asset is of the pair. But if you are clever enough to keep that liquidity in range, or if your specific outlook on which way the market's going to develop is accurate enough. that the price of those assets remains inside of the balance that you've defined, then you can actually earn significantly more fees compared to the liquidity that you're providing because the bid-ask spread is still relatively low compared to the slippage that you're providing to the market. And so this was really the first, let's say, step in the direction of trying to address the implicit or explicit profitability of liquidity providers in an AMM. But it still didn't really work out. I think that I authored, along with my colleagues, Stefan Lösch and Nate Hinman and Nicholas Welsh, I actually did, I think, the first full dataset analysis of Uniswap v3 back in, I want to say, 2021, probably October, November of 2021. So we literally analyzed every single position on Uniswap v3 to try and determine just how effective a financial instrument this idea of bounded ticked liquidity actually is. And we were rather generous, I would say, in the way that we handled the data because there are multiple perspectives that you can bring to that kind of analysis. And just the fact that liquidity is no longer in range that it used to be can mean that the IL, or the effective profits and losses, at least the loss component, can be quite dramatic on a concentrated liquidity protocol. It spikes very, very quickly. But you could also make the argument that these aren't liquidity positions, but are limit orders, in a sense, right, by the dip. And certainly that's one of the arguments that's espoused by the Uniswap v3 wallpaper authors. And so we chose to observe that, like, let's give these liquidity providers the benefit of the doubt and analyze these positions such that their losses essentially stop as soon as the market price reaches the border price of their liquidity range. Which should give the most modest estimate of the profits, or the most modest estimate of the losses, right, the financial harm done to liquidity providers by the rest of the market. And even in that context, where we essentially stop the IL calculation as soon as the market reaches that price boundary, we found that no more than 50% of liquidity providers on Uniswap v3 were profitable. And more importantly, the 50% that weren't profitable had lost a lot more value than the value that had been accrued to the liquidity providers who were profitable. So it still didn't really look like it had done anything to address the financial instrument that is represented by the AMM. And of course, that makes sense, right? These concentrated liquidity protocols, you know, Bancorp v2 included, are just leveraged AMMs, right? You'll see in the Bancorp patent that describes this process, and even in the Uniswap v3 whitepaper, they talk about virtual token balances. Basically the idea of taking the liquidity that's in an AMM and pretending that there's more liquidity than there's really there. So if you want to take like a 10x amplification, for example, and you've only got 100 ETH or something at a certain position, the bonding probe will behave as if it's got 1,000 ETH or 10,000 ETH or something. And this is how you achieve the better slippage rates, but also is how you end up being completely depleted of your risk asset or your cash asset, depending on which way the market's moving. So if you have a look at things like Milamidigliani theory, it says that everything else being the same in a tax-free system, a leveraged capital structure is going to perform exactly the same as a system that has no leverage. And I think that that's more or less what we saw in those analyses. And we had a slightly different view. So after recognizing that this amplified or leveraged liquidity of the AMM was not really making things any better, we instead tried to build a model that was socializing profits and losses. So in essence, using the BNT token as a way to act as a numeraire, I suppose, across the entire protocol, we actually implemented something like an insurance strategy where the performance of pools that were doing very well could offset the losses of pools that weren't performing very well, at least relative to the model position. And we actually went through two iterations of this. The first one was B2.1, which was pretty gas inefficient, but still held up relatively well. And then with B3, which was sort of a re-imagined optimized version of that protocol. And things were actually working pretty well under that scheme. And it's possible that in an alternative timeline, that model actually could have sustained itself indefinitely. The problem that we faced was that over two thirds of the liquidity in the system provided by 3R's capital and Celsius. And so when they were facing bankruptcy and other issues, when they started to withdraw their liquidity, it ended up in what was effectively a bank run, and which massively impacted the token supply of the project and kind of wrecked the insurance mechanism that we had in place there. And so now having been through sort of two different models. The first one being just leverage the liquidity so that you can get more fees. And then the second one being, well, maybe we can sort of socialize profits and losses across the entire system and help to compensate liquidity providers fairly that way. Having failed both of those models, when I was asked to redesign the DEX infrastructure for the project, I was really starting to focus a lot more on how people actually trade. The weird thing about an AMM is that liquidity providers don't think about themselves as traders even though that's actually what they are. In a sense, things like impermanent loss or at least the impermanent loss that makes people feel upset is a direct result of the fact that they treated it badly. But it was their choice to provide liquidity to the exchange in the first place. And it's the exchange prescribes for them a specific trading strategy, which is usually this short gamma option that I discussed earlier. With Carbon, I thought we've really got an opportunity here to learn directly from traders how they would prefer to trade. And I was inspired by conversations that we'd had directly with people that would categorize themselves like DeFi super users, but also mammoth or whale traders on Binance and Coinbase and so forth. I was inspired by YouTubers who do, sort of sponsored their own content and cast daily analyses of the charts and described how they would gonna trade them and that kind of thing. And I wanted to build a protocol for them, right? People that aren't providing liquidity for its own sake, but people who are rather creating orders, right? People who are speculating on the price of an asset and are happy to mark out the price of what they're happy to trade for at some point in the future. And so we're kind of, it's a really thorough departure, I think, in terms of bank or narrative, because it really started out as a project that would put the needs of the industry first, right? The needs of token projects and new protocols and their tokens above all, right? And at the time in 2017, 2018, this was the priority. DeFi didn't even really have a name or a face yet. Making sure that things could be exchanged with each other, I think, was the priority back then. And it's kind of become a priority over time that these protocols should be financially performant. And with just recognizing that the game has changed in a sense, I think justifies the fact that the AML model should be retired in favor of something that is a little bit more user-centric and has a lot more flexibility and allows the person who's interacting with it to decide how and when they trade. And that is, I think, the only way that we can address the productivity of decentralized exchanges. Right? I think almost by definition, the AML will always be the thing that benefits the project, but never the liquidity provider. And in a sense, an order book-like system like Carbon, it's kind of the opposite. It's no longer the case that the liquidity provider is servicing the project by providing liquidity to it. And now it's the person who is interacting with that system that is almost influencing projects of the tokens which they are trading. And actually I prefer that system. It's a little bit more bottom up. It's a little bit more grassroots and I think it actually leads to more functional, more healthy markets. Got it. Yeah. That's super interesting. A lot to unpack there. Just to recap, from what I got from what you were saying, for the most part, DEXs today have worked to a degree, maybe not for all participants, mostly for the projects themselves, but it's brought more permissionlessness to crypto trading and they've made some improvements in efficiency from the original model with all the concentrated liquidity protocols we see today. But on the other hand, liquidity providers or LPs in default, they're still at the mercy of the market because they don't control the price and flow of assets that happens with the buyers and sellers. And so what happens is oftentimes these LPs need to rely on the fees to overcome the loss that they may incur or the protocol needs to heavily incentivize these liquidity providers with their own tokens. Right. It should be pointed out as well that even the takers, even the people that are trading against the LPs, they don't even get a very good deal out of this. And it's not necessarily because of things like slippage and so forth, because I think that that's just a consequence of indecisive markets, but rather the AMM mechanism itself can be weaponized against the person that's trading with it. And this has manifested not just in terms of front running, which was the original attack in these systems, but it's matured into much more sophisticated systems. Things like sandwich attacks in the Uniswap V2 paradigm. But then even with Uniswap V3, and it's been pointed out several times, that there are ways that you can provide liquidity out of range and then sandwich someone by forcing them off of the current price on a V3 pool and into some completely other region. And it might be separated by many tick spacings. But if they haven't set up the parameters correctly, then they will be forced to pay a much worse price than what they were expecting to get. So really, the only benefit truths, I think, out of the AMM system are the token projects that don't have to deal with listing costs, that don't have to deal with, for example, the onboarding of a market maker and other legal issues implied thereof. They're the ones that really get the most out of these systems. But the people that are providing the liquidity and the people that are trading against that liquidity, they're the ones that are getting the short end of the stick in all of these systems. Yeah. Got it. For sure. And so I guess that this leads to what you guys are building at Carbon. When I was first reading about it, it took me a little bit to understand exactly what the protocol was building, since I think everyone in DeFi is sort of used to the concept of earning yield from provision of liquidity. And you guys kind of set the model on its head where, to your point, it's less about fees, but order execution and the price in which you're willing to buy or sell these assets. So I think it would be good for you to maybe talk about exactly what you're building at Carbon, how it works, and how it's solving some of these problems that you mentioned earlier. Yeah. Okay. So yeah, I think you're right. We've kind of conditioned blockchain participants over the last few years. I think by sort of ramming AMM theory down their throat, they've come to expect that this is the only way that a decentralized exchange can operate. They're used to liquidity pools, pool tokens, bonding curves, this kind of stuff. But Carbon kind of does away with most of that. We don't refer to people that interact with the protocols, liquidity providers. We don't issue pool tokens. There are no liquidity pools. Rather, Carbon is a lot more like eBay in the sense that you might have some Bitcoin or wrapped Bitcoin if you're on Ethereum, and you think this next bull run, we didn't break 100K like we were promised in the last bull run, but this bull run, we're definitely going to break 100K. If that's really what you're waiting for, if you are sort of disciplined and that's the price target that you have, there's nothing that should be stopping you from going on to Binance or Coinbase today and then providing those Bitcoins to the order book and offering a price at 100K. Maybe if you've got a handful of these Bitcoins, you might think, you know what, I'll sell one at 100K, one at 105, one at 110 or something, just in case the price goes a little bit further up, but know full well that we might not reach those price targets or something. But this is actually a very natural way to trade. I think what's a very unnatural way to trade is to say, you know what, I think that the price of Bitcoin is going to go absolutely nowhere, and whatever the price is today is going to be the same price it is next week, and therefore I'm going to put my Bitcoin in an order system that allows me to withdraw more value at the end of that time period, so long as the price hasn't changed, which is kind of the AMM perspective. So you're right in the sense that we don't have ... There is no fee model. We don't say that people that are interacting with the protocol are there for fee accrual. What you are trying to do is trade well. If you're the same person that says, you know what, I'm going to sell my Bitcoin for 100K, but I also appreciate that after we hit that price target, it's possible that Bitcoin will then capitulate back down, and the top of Bitcoin in the last cycle was somewhere around 68 or $69,000. Maybe we're going to revisit that previous high to test it for support, and this is precisely the kinds of commentary that you get when you speak to people that are trading Bitcoin in very long time frames. I defer to some of the more sober analysts in the space, people like Benjamin Cohen, that will say things like, concede right now that you're not going to be able to call the top and not be able to call the bottom, and that you should set price targets where you are improving your valuation over time without refusing to trade stuff, or without over-hodling, or over-committing, or over-investing in particular assets. That's really what Carbon is designed to do. This is how it works. The smart contracts are very, very lean. You come to the system and you say, here are my coins, whatever they are. Maybe it's wrapped Bitcoin, maybe it's ETH, and you say, these are the amount that I'm happy to provide, and this is the price that I want to sell them for, an actual rate. If it's ETH, you might say, I want to sell these ETH for $5,000 each, and that's all the system does. It takes those ETH, it slaps a price tag on it, and it sits on chain and waits for someone to take that price from you. Now, if the price at the moment of ETH is $2,000, then of course no one's going to buy your ETH from you for $5,000. They can go and buy ETH somewhere else for $2,000, so you're going to be waiting a little while. But eventually, if you're right in your prediction that that's where the market's headed, then the price of ETH might ... It will be worth $5,000 again one day. When it gets there, people will find the price that you're offering more active. Now, I often get asked, does this mean that there's an oracle system, or when does the liquidity become active, or do you need keepers or something? Who executes this trade? The answer is, I think, much simpler than what people think, which is that no one does. We're just waiting for someone to come and take that trade from you when they find it attractive. There are steps that we can make to make sure that if the market has an appetite to buy ETH from you at $5,000, then there are ways that we can drive pedestrian traffic towards that order. I think the most efficient is to make sure that we're correctly plugged into aggregators. I'm excited that we've recently just finished a lengthy integration process with 1inch. Even prior to that, we've already been integrated by OpenOcean, which I think is the preferred aggregator of MetaMask. If someone goes to their MetaMask slot application and tries to buy ETH at around about the time where ETH is trading at $5,000, your order will be found. People will choose to buy that from you in much the same way that if you put a rare trading card or something on eBay and price it at $10,000, people might not buy it from you right away, but when the market heads that direction, someone will buy it from you. you eventually. And so we do everything that we can to make sure that that process is as efficient as possible, including back running people's trading errors. So one of the most common trading errors is that people will avoid using aggregates for reasons that I still don't understand. And they will instead trade directly via a DEX front end, which if you're doing this, you need to stop doing that right away. It's a very, very bad idea. But people will still do it. And so let's say that you're offering your ETH 5,000 and then someone goes to Uniswap and buys it for 5,100. They got a worse deal, but that's still an opportunity that we can take to make sure that your order gets filled. And so our arbitrage system will back run their transaction and then use the cash that's extracted from that trade to then fill your order and extract the ETH out the other side. So we basically have created something like this basis liquidity, this full back end of resting limit orders, a true limit order protocol that is thoroughly plugged into the aggregator space, but also in the presence of an arbitrage bot that observes those orders against all of the DEX liquidity at the same time. And so it's actually a very, very nimble, very lightweight, but also very powerful way to trade because even trading spot has its disadvantages. If you are someone that is wanting to sell your ETH at $5,000 at spot price, then you might find that because of sandwich attacks and front running and so forth that I mentioned earlier, that it's actually very difficult to get a good price on Ethereum unless you are well equated with things like Flashbots or other protection mechanisms. But even then, block builders might see your transaction through an uncooled block. That doesn't really happen that much anymore since the proof of stake consensus system on Ethereum went live, but it can still happen. And so the risk is much less if you're using things like Flashbots, but it's not zero. Whereas in our system, if you just say, you know what, I'm just going to put a $5,000 price tag on my ETH and just set it up and let someone take it from me, it's literally guaranteed that you will get that $5,000 trade if someone has got the appetite to buy it from you. And if that's the current price of ETH right now, then your confidence is relatively high. And so this means that the price from a price execution perspective, it's actually more certain, much safer to trade via a limit order system like the one that we've provided them to try and trade spot on a DEX. And I think that that is the important sort of take home message here, is that we've kind of become complacent, right? AMMs were our first and worst attempt to build decentralized exchanges on a blockchain, or at least they were our second and only moderately better than the sort of SQL style, off chain, chain mess that was things like Ether Delta, which were a really great inspiration, I think, for the beginnings of DeFi, but obviously not a very good way to build a decentralized exchange. And so we've kind of taken the, I'd say the lessons of the last like three or four years, and we've really built something that I think is very specifically tailored for the needs of market participants, not for projects. And so if you are sort of frustrated with the fact that you haven't been able to place a reliable limit order somewhere, or if you haven't been able to trade spot before and get the actual, the tokens that you thought that you were entitled to because you've been sandwich attacked or front run or something else, then just know that we have built a system that is immune to these kinds of adversarial influences, or even just inconveniences. And you get all of these things for what is actually a lot less gas than you get on competing systems. So it's not even the case that we've built something so sophisticated that it costs a huge amount to interact with. It's actually a lot cheaper than even Uniswap B3 or B2 is. So it's a slight change of pace. I think it does require a bit of re-education because, like I said before, we have spent so many years trying to teach the entire industry about AMMs, and now I'm trying to get them to- This space was downloaded via spacesdown.com. Visit to download your spaces today. It's a bit of an uphill climb, but we're certainly getting there. The people that have been using carbon so far, I think we've had nothing but a positive reception. But yeah, it's a bear market and there aren't as many people perusing the protocol offerings as they used to be, but I'm still very optimistic that we've got something very special and I think that it's going to be a very important component of whatever the next phase of DeFi looks like. Yeah, cool. So to recap, you have these users who create these trading strategies in carbon and this is what you guys refer to them as makers. And they're similar to the existing liquidity providers and current AMMs, since they're providing the liquidity for these trading activities. And then on the other hand, you have users or takers who perform these or execute these spot trades using the maker's liquidity. And these can be, I guess just users directly using the carbon front end or as you mentioned, DEX aggregators like I guess my question is, is there any cases where people have these or makers have these live strategies, but their orders don't get fulfilled for some reason? Are orders always likely to be, I guess, executed when the market reaches those price ranges? It's kind of an abstract question. There really isn't such thing as a market price, capital letters. There's really just the last price that someone paid for something. I used to give these kinds of simulation seminars on our Discord every week, where we would have a look at the price data from different API sources and compare them. And so, it might surprise you that Link was once quoted as having a price of zero by CoinMarketCap. And if we have a look at the price of stable coins, especially things like USDC for some reason, if you get the price data from some API sources, it will show these amazing wicks all the way up to $13, $14, or sometimes down to just a few cents. And I can't know for sure, but I speculate that one of the reasons that this happens is that whatever that data aggregator is using, it might be sensitive to things like sandwich attacks, which do result in these enormous spikes in the price of the thing that people are trying to trade with. And so, that's a good question, right? Let's say that you did have a limit order to sell your USDC when it was worth $13, or when it was worth 13 USDP. Even though the API that you're reading says that there was a few minutes where the price of USDC was worth $13, doesn't mean that someone's actually going to pay 13 USDT for your USDC in that time period. And why? Well, the reason is that these market prices, they're just guesses, right? Well, I mean, they're measuring it, they're looking at it, but it doesn't infer that that's kind of the market consensus on the price of that asset. And this is, I think, one of the things that's generally very poorly understood, not just in cryptocurrency, but maybe in all liquid markets. And that's the price that you get, right? When you look up the price of gold, or when you look up the exchange rate of the Australian dollar versus the US dollar, you basically never get these rates. They're constantly changing. And the very act of you taking the market at a particular price causes the price of the market to move. And so in that sense, we could have a look at other situations where if someone, say, put up some liquidity for ETH at exactly $2,500, but only provided a very small amount of ETH at that price range, then the gas adjusted price of that ETH might be a little bit different. And this would mean that you would need to wait for the market to go a little bit higher than $2,500 in order for, for example, an aggregator to pick it up. But also, if the number of units that you're selling is small compared to what someone can get somewhere else, then maybe just for the convenience of being able to do everything in one transaction, or even the gas savings of doing everything in one transaction, might mean that people will still prefer to take it from... somewhere else at a higher price. And this is something that's true of markets generally. So for example, one of the examples that I've spoken about previously is things like paper cups or plastic cups, like the type that you buy to cater a wedding or something like that, or a birthday party. There are some distributors that will sell you paper cups in plastic sleeves of like a dozen cups or something. And they might be very, very cheap, but if you're trying to cater for a very, very large crowd, then you may prefer to get the cups from a different distributor that will sell them to you by the kilogram, like boxes and boxes of these things. And even though you might be getting a slightly worse rate per cup, just logistically, these businesses do functional volume. And so in a way, the convenience saving or the logistic costs of moving so much inventory is actually kind of priced in, and it means that those things are actually worth slightly more than if you were buying them in lower volumes. And this is true in cryptocurrency as well. Even on a traditional order book, you might find sometimes, especially because cryptocurrency exchanges aren't appropriately equipped with things like clearing houses and other things that the Chicago Mercantile Exchange is, you'll actually find that some limit orders, if they're small enough, will actually remain unfulfilled, even if the prices move past it, if someone hasn't been able to close that position efficiently, they kind of get lost in the mix. So there are these other aspects to it. And I think it is helpful to remember that you are literally making a market, right? You are offering a certain amount of stuff for sale at a certain price, and the amount of stuff that you're selling is just as important as the amount that it costs to buy it. And so your neighbor might be offering a slightly worse price than you, but they might be offering 10 times more inventory. And if there's a buyer somewhere that's looking for 10 times more inventory than what you're selling, they might just buy it from your neighbor, even though your neighbor is offering a slightly worse price. So there are these kinds of things that you need to take into account. It's not a system that guarantees execution at a certain market price, because there is no such thing as a certain market price. Rather, what we do is make sure that there is every opportunity for you to find an appropriate buyer for the inventory that you're offering and at the price that you're happening to sell it at. But that doesn't mean that we can guarantee an exchange any more than eBay offers people a guarantee that if they sell their PlayStations at a lower price, that they will be bought first, because it just doesn't work that way. If you go to eBay right now and sort PlayStations highest price to lowest price, you might find that for some reason or another, you actually prefer to buy a PlayStation that's a slightly higher price than the cheapest one available, either because it's closer to you in terms of shipping or something else. It could also just be random. Some people will just go to eBay and just click PlayStation and just buy the first one that they find without comparing prices. And so no, we don't offer those guarantees any more than eBay does. But we do make sure that we're plugged into the correct aggregation technology and that users can see how liquidity is enumerated and priced as effectively as possible, so that there are the best chances of being executed. But there's always some nuance, there's always some fuzzy region around what the quoted market price is that means that things like actual market demand, trading volumes and other things, that's really out of our hands. It is really up to the market to decide. But everything other than for those talking points, we've got those bases covered. That makes a lot of sense, but do you think this is maybe a function of Carbon maybe needing more makers to have more strategies, more volume, more liquidity? No, it's not actually that. Carbon would work with only one maker. With only $100 in TBL, it would be just as efficient if it had $100 million of TBL. It's completely immune to the number of makers that are there. But in terms of driving more volume to Carbon, that's also no impact in terms of the amount of makers? Yeah, I mean, correct. It's actually better for us as a protocol. Obviously, as the project lead, I want Carbon's TBL and volume to be as high as possible. Even if I consider things like nascent TBL to be mostly a vanity metric, I still want my vanity metrics to be good. Volume obviously is not a vanity metric. This is directly out and monetize the system, so I want the volumes to be high. But it's important to recognize that this is exclusively from the protocol and business side of things. The actual makers themselves will not benefit at all if the TBL is higher or if the volume is higher, because at the moment right now, essentially every time someone places an order, it gets executed basically right on the button. There might be a couple of a block delay or something while an effective arbitrage route is still being established. But yeah, like I said, even if Carbon was barren and there was no one on it, and you decided to put $100 worth of ETH up at the current market price, it would still get executed in the next block, regardless of how many makers are in the system or not. We would only have $100 of TBL and we would only bend to $100 of volume, but it would still execute every bit as efficiently as if we had $100 million of TBL. Right. Okay. That makes sense. I guess moving a little conversation to the maker's side in terms of what are you seeing today in terms of the usage of Carbon? What kind of strategies are the most popular ones being deployed? Yeah. There are a couple of sophisticated makers that are working on what they refer to as institutional grade trading strategies, because they're looking to onboard clients essentially through their vault system. Some of these strategies look like things like trading ETH versus the ETH Flexible Leverage Token, or the Flexible Leverage Index. There's a project called IndexCoup that has a product where ETH is effectively contributed on or borrowed from a compound, I think, and then sold for something like USDC, and then they re-borrow ETH with that USDC, and they basically perform that winding as many number of times as necessary to give the index token that it represents, or sorry, the index token represented by it, a true X exposure to the Ethereum price. This just means if the price of ETH goes up 10%, then the FLI token will go up by 20%, and similarly, if the price of ETH goes down by 10%, then the FLI token will go down by 20%. But really, you don't really want to hold the FLI token for very long, but in terms of trading it on a daily or weekly timeframe, it's a really, really convenient way to introduce leverage into carbon. Those types of strategies I've seen work extremely well. The person that is running that strategy regularly told me at Paris that he expects to 10X his ETH holdings over the year, thanks to carbon and the FLI token strategy working for him. So that's been extremely performant. There's also, I think the ETH-USDC trading pair has been extremely popular. You can see on our app, there's actually a new feature that's just recently been added called the Explorer, and the Explorer allows you to search either by wallet or by token pair, and you can kind of peruse all of the different strategies that people have been setting up, what their ROIs are, and what their current status is, and so forth. But yeah, I think that there are the super sophisticated, well-informed trading strategies such as the leverage token one that I just mentioned, but then the same vanilla strategies that you yourself might be running on your Coinbase account are also extremely popular. Things like people saying, you know what, I'm going to scalp ETH when it gets back to this price, right? Or people looking at the Bitcoin shot and then putting WBTC to sell at 30 and then to rebuy again at 25. So there are aspects to our system that we haven't had time to explore, which is the recurring limit orders and rotating liquidity. and maybe we can discuss that next, but these are the strategies that have proved the most popular. It's the big, the commanding presence in cryptocurrency and all of the flavors of the week, YouTube technical analysis trends, these tend to be the tokens that get traded with enormous size. Yeah, you mentioned the recurring trading strategy. I think when I saw that one and you guys provided some strategy backtests on it that were really pretty impressive when compared to like a traditional AMM. So I was curious if you could talk a little bit more about how that strategy works with the buy and sell ranges and what you guys saw in your backtests. Yeah, so I refer to carbon as an audiobook-like DEX construction. I say audiobook-like because an audiobook is really just a subset of what carbon can do. So to clarify what I mean by this, in a conventional audiobook, you would say, here are the number of units of things that I'm trying to sell and here's the price that I'm willing to sell it at. I mean, in reverse, right, if it's your cash asset, you can say, here's how much cash I'm putting up and I'm happy to buy this other asset when it's exchange rate is whatever. But it's still like it's one price, right? It's a certain number of things and a price for all of those things. So it's quite a sort of an immutable bucket. And so the way that it would operate if you're executing sort of like a manual grid strategy is that you might say, based on the current Bitcoin price, for example, that I've got a bunch of Bitcoin and I think we're going to go back to 30K soon to sort of retest that level, but I'm not confident that it's going to break through. And so I'm going to sell two or three Bitcoin at the 30K level. And after that trade happens, right, so I put my Bitcoin in the book, I walk away for a couple of days or a couple of weeks, I see that the price of Bitcoin gets to 30K. And so now I come back to my Coinbase account or Binance account or whatever, and I see that I've got now 100 grand or whatever it is from the sale of my Bitcoin at 30K. Great. But now I want to buy that Bitcoin back when it gets lower, right? I'm not like exiting from Bitcoin permanently, I'm just trading the current volatility. And so now that I've got this 100K back in my account, I'm now going to say, you know what, I'm going to buy that Bitcoin back when the price gets to 25K, for example. And so the process here is kind of arduous, right? You're going to interact with the book and Coinbase charges you for that interaction. And when your Bitcoin sells, it charges you for that as well. And then when the money ends up back in your account, you're kind of charged for that, I guess, on a taxation level, but I don't think that Coinbase charges you directly. But then you're going to put that money back in the book now at the 25K level, Coinbase charges you for that. And then when your trade is executed, you get charged again. But there's kind of a lot of steps here, it's a very hands-on sort of approach. And I think what's been lacking is the ability to just express your desire to remain inside the book at the levels that you want to trade at. So imagine being able to say, you know what, I'm going to put a limit order on my Bitcoin at 30K, and then I'm going to place another limit order to buy Bitcoin back at 25, except I'm only going to fund the sell, right? So I'm only going to provide the book my Bitcoin. And so if the price of Bitcoin goes straight back down to 25 before reaching 30, it's like, damn, right? I didn't get a chance to sell any and so I can't buy any. But if that first limit order hits, then now you've got some cash that's like ready to buy Bitcoin again if the price comes down to 25. So what this means is we actually do have a way with carbon where you can create a limit order with money that you don't have yet, that's contingent on the sale of another token at a completely different price point. And what's interesting about this is that you can trade that essentially in perpetuity. So you can say, you might be of the mind that this 30K to 25K range is likely to persist for some time longer. And I'm not saying that this is necessarily my position, I'm not providing trading advice here. I'm just saying, it might be the case that someone is convinced that we're going to bounce between 30K and 25K for like the next year or something. And with carbon, you can literally constantly sell the Bitcoin that you have at 30 and rebuy it again at 25. And what's really interesting is that when you run these backtesting simulations, the ones that you were just referring to are the ones that I said that I used to give these seminars on every week. What you'll find is that this mentality that people have where the aim of the game is to trade exactly twice, to buy the exact bottom and sell the exact top, is not even the most profitable way to trade. You can miss the bottom and the top by miles and trade the middle and still end up more profitable than if you did somehow magically call the exact bottom and exact top. And it's an important lesson, I think. And so yeah, that's exactly what this backtesting simulator is designed to, to help you investigate on your own. So it doesn't provide any predictions, it doesn't try to guess what the future price targets should be. This is something that you have to do on your own. But it does help you realize that calling the bottom and calling the top or even trying to predict these things is kind of besides the point. All you're trying to look for really is where the volatility is going to be in between those two things. And carbon is perfectly situated through this kind of rotating liquidity principle to take full advantage of those kinds of market movements. And if you go to the Explorer and have a look at how some of these strategies have been behaving, remember that the protocol isn't really that old. I think we only launched back in April. And you can see that despite not having any incentives at all, there's no liquidity mining in our system, there's no inflationary token rewards, there's none of that. But you're still seeing people are printing these kinds of 25%, 30%, 40% APIs, or not APIs, sorry, ROIs. We don't actually report APR or APY. We decided to stay more squarely in the zone of certainty. I think a lot of people get confused about how APRs and APYs should be calculated. So return on investment is the number that we report. But yeah, the point that I'm making is that compared to AMMs and compared to liquidity mining systems, I think we're actually, the people that are using our system are actually doing extremely well. So yeah, this kind of rotating liquidity paradigm, it's powerful enough that I think that we should be confident in leaving behind these kinds of inflationary tokenomics models in the past and moving ahead with just studying and offering products that help traders who are looking to take a positional direction, help them make a better decision and profit from that kind of intraday or intraweek volatility. Yeah, super cool. Since you brought up the point of these token incentives with liquidity mining, because carbon doesn't have, it doesn't actually, I guess, these fees for the makers and takers, they don't go to anybody, but I guess it goes fully to carbon itself. So my question is, what is sort of the mandate on this fee that the protocol is building? And also, I think you guys don't have a governance token right now for carbon. And so is there some sort of roadmap for the decentralization of carbon down the line? Yeah. Carbon is already decentralized, so the governance token is still BNT, the Bancor token. So carbon is a named product of Bancor. We're at the stage now where Bancor has a fairly decent IP stack and a large number of different products that it's contributed to over the years. And I'm very interested in sort of dusting off some of these old contracts and modernizing them and having a protocol suite rather than a single version number that represents everything that the project does. It just doesn't make sense anymore. The carbon is the Model S, Bancor is Tesla, right? There's an organization and a DAO that represents the project, and then there are going to be a suite of different products that are represented by that, but it's going to be a single DAO. So BNT and VBNT at the moment is still the governance token for carbon as it is for everything that Bancor produces. With respect to the fee system that you were referring to... and what the mandate is there. Essentially if you offer your ETH for $2,000, let's say you put up one ETH and say, I would like $2,000 for this ETH please, you will get exactly $2,000 for that ETH. And so your profit motive is still going to be quoting prices that you think that the market will take you at. And if you're using the rotating liquidity, essentially being able to correctly manage that intraweek volatility. So if you're selling ETH at $2,000 with the expectation of being able to buy back at $1,500 for example. But that's what you get, right? In that particular case, that's like a 33% ROI, which is pretty good. The fee that we charge on top of that, that's something that the taker essentially pays to the protocol. And this is a markup, meaning that if you're demanding $2,000 for the ETH that you're selling, the taker, the trader that's going to be interacting with your position will pay $2,000 plus 20 basis points on top. So that's going to be what, $2,002? And so it's pretty minor. Given the volatility of a lot of these assets, it doesn't change very much. And for assets that have a stable price, things like USDC, USDT, I think we only charge one 10th of a basis point. So it's practically zero. And so what that means is that your profitability, it's not at all determined by fee structures and it's determined exclusively by the range that you're trading at. So if you're lucky enough to buy Bitcoin at $3,500 and lucky enough to sell it at $69,000, then that's your profit. That's your 20X profit. But if during that process, the person that traded with you at $3,500 and the person that traded with you at $69,000, they will be paying the protocol 20 basis points on top of whatever your core price is. So it's relatively insignificant. The discussions that we had with the DAO before launching Carbon was where to set that fee level and 20 basis points was kind of where we settled down on just because having a look at the overwhelming majority of assets that people trade with, 20 basis points isn't really enough to move the needle with respect to decision making and how people will choose to trade, but is significant enough that we think that if the protocol becomes sufficiently popular, that it could be a sustainable source of revenue for the project in the long term. So that's really the point here, is that the fee model in our system actually goes through the protocol. It's a true protocol fee, whereas the fee model in traditional AMMs is really not quite even a fee. It's more about determining that bid-ask spread for the liquidity that you are providing. And the simulator, the back-testing simulator that I've created and that you mentioned before, I think does a pretty good job of highlighting exactly how that process operates. Got it. Yeah. So 100% fees goes to the protocol. Makes sense. So what's next for Carbon, I guess? What are your plans for things like cross-chain expansion or ways to drive greater usage to the platform? Yeah, absolutely. So yeah, an enormous amount of my time recently has been spent on B2B development, and I'm speaking with other projects, especially people that are running on-chain hedge funds and similar products, asking them what their needs are, helping them to integrate, and that kind of thing. Specifically for cross-chain stuff, this is a fairly nuanced and abstract sort of question. I think in general, the participants in the industry are of the mind that you kind of just decide on a specific blockchain and kind of deploy there. And I guess that's also true, like you can kind of do it that way. And I think it's a bad practice to just kind of spray and pray. I prefer to treat these cross-chain deployments as being of strategic importance. So for example, one of the projects that I'm speaking with right now is very, very interested in Carbon. It's precisely the type of product they've been asking for or looking for. And so they are, I guess, a type of yield aggregator, but they're trying to coin the term for themselves, real yield, which I've heard thrown around the industry a bunch of times. I don't think that they're the first to think of it, but they are certainly taking it seriously. So they're kind of almost like a Wi-Fi, but with a more tri-Fi leaning, and are looking for ways to generate their own alpha. So they refuse to use things like liquidity mining programs and other things, and are explicitly looking for other financial instruments to use to generate value for their protocol and for their customers that way. And so Carbon, as opposed to conventional AMMs, and as opposed to liquidity mining programs and so forth, is much more attractive to them. And so when I was speaking with them, the first questions that come up are, which blockchain are you using now? It's something you didn't have to ask two years ago, but now it's like the first thing, the most important thing you ask. And they said that they're currently sort of in a friends and family mode on Arbitrum, but still with a few million dollars worth of liquidity to use, and not a whole lot of stuff to use on Arbitrum with it, but looking to move to Ethel one eventually, so that they could do things on both of those fronts. And so it's these kinds of partnerships that you look for. And so I can say now, finally, okay, here's our launch partner for Arbitrum. So I'm sort of waiting for a commitment there. And so these B2B processes are a little bit longer, a little bit more structured. But also I kind of prefer them, because it means that when we're setting up our deployment and getting things ready, we've got someone on the other end of the phone that we can talk to and be like, hey, the thing is about to be live, and which things do you want supported, or what are you waiting for, that kind of thing. So it's these kinds of partnerships that I'm looking for. And so I've been speaking with projects on Mantle, I've been speaking with projects on Polygon, Optimism, I've been speaking with the founders of Shardium, of Monad, of Conflux, which is like a Chinese EVM chain that's gaining a lot of traction now. There's so many opportunities, but I refuse to just kind of deploy the contract and just kind of market it from Twitter. It's just not how business is done. You really need to find the right people that want to use your protocol on that system or secure a grant or some other type of agreement with the blockchain itself, so that you know that when you arrive there, that you're really hitting the ground running. I think the biggest mistake that a lot of projects have made over the last couple of years, and I don't mind holding SushiSwap, I think, up as a really good example for this. And I don't think that they would mind me doing so. But I think that at one stage, there are up to like 19 deployments on different blockchains, and I think only two of them are still active, or three of them are still active. And so that's the kind of thing that we're looking to avoid. So we're looking for very stable, long-term, sober relationships, either with the blockchain itself, or with a very important project that is really enthusiastic about carbon and about using it from the get-go. And I think that this is, yeah, these conversations are going very, very well. So yeah, all of the blockchains that people like are the same ones that are in my gun sites. So if I can, I would love to foster relationships with projects on Arbitrum and Optimism, on Avalanche, even on Solana. I've wanted to go to Solana for years, but it's just been difficult because of all of the drama with FTX and other things that have been happening, and the instability of their chain at times, which thankfully is not true anymore, but it was once true. There are all kinds of things that hold these things up that I think residents in DeFi just don't get to see. But yeah, the strategy is the same as it's always been, which is... is find a compelling business case to be on a specific chain and then go there. And I'm pursuing every available lead, um, you know, with, uh, with all of my energy and enthusiasm. Um, but yeah, it's, uh, it's always going to be a slow process. Awesome. Yeah. I think that's the, you guys have the right strategy in terms of finding that that product fits first on the, on the chain before deploying there. Um, but yeah, super excited for, for the things to come. And I think it would definitely benefit for some of the lower gas costs as well on these, on alternative L1s and L2s. But yeah, I think that's a, that's a good place to, to wrap up the conversation. Um, wanted to move it to QA now, if there's anyone in the audience who has any questions or Mark or I. Wanted to say it as well. If anyone is, uh, say if anyone suffers from a stage fret, I don't mind if you want to DM me, uh, privately via Twitter, um, or via telegram, but also, uh, Jen, one of our listeners up here, um, hosts Twitter spaces, I think about three times a week now. Um, and so if, you know, if after the call is finished, you think, Oh, you know, this would have been a perfect question. I wish I had asked that. Um, you, I don't mind if you want to join one of the, uh, one of the other Twitter spaces we have organized this weekend and ask your question there as well. There doesn't look like there'd be any other questions. Um, I think this has been a really informative discussion on the DEX landscape and, um, really interesting stuff that you guys are building at Carbon. Uh, appreciate you taking the time to talk about everything about the landscape and protocol and yeah, excited to, to keep up on all the new updates. Thank you so much for hosting me. It's been a great conversation. Awesome. Well, thanks everyone. Have a good day. Thank you. This show was brought to you by exponential. Exponential is on a mission to democratize access to the best yield opportunities in DeFi. Join exponential.fi now to start your DeFi investing journey.