What Drives the Rapid DEX DNA Evolution?
Untitled Investment Talk from May 29th, 2023, with Darren Camas, CEO of IPOR Labs, and Dr. Karl-Michael Henneking, Founder of Untitled Investment Expertise
The interview consists of four segments:
1. Intro Darren & DEX Basics
2. Why and how did the DEX DNA evolve so rapidly? Where is the DEX market heading?
3. Risk-Free DeFi Yields, Interest Rate Swaps & IPOR
4. DEXes and Regulation
Enjoy the read or listen in on Spotify, Apple, and Google.
1. Intro Darren & DEX Basics
Karl-Michael: Darren, how did you come to the crypto space and what was your moment of truth?
Darren: There was kind of a strange moment of truth. I would say I’m a very old person in crypto years. I got started in 2011. So I guess the moment of truth was when I met some guys that were starting up one of the world’s first crypto exchanges. I was at that time only Bitcoin and they told me that we’re starting a Bitcoin exchange.
I said, what’s Bitcoin? They said it’s like crypto anarchy for money. And I didn’t have any idea what they were talking about, but I went and studied all the resources I can find about Bitcoin. At that time, it was really fascinating. I did not understand maybe the vast majority of it, but I was hooked.
And so from that time on, I’ve been building in the crypto sphere.
Karl-Michael: I think we’re gonna talk a little bit more about how you came to the decentralized exchange space later, but before we do, so for all our listeners who are not so deep into DEXes or decentralized exchanges, let me run you through a short introduction.
Other than centralized exchanges like Binance, Coinbase, or OKX, decentralized exchanges are peer-to-peer marketplaces for trading digital assets. Crypto traders make transactions without any central entity involved. So, no Binance, no Coinbase, no OKX. Instead, transactions are facilitated through smart contracts. It’s all software, it’s all automated.
Decentralized exchanges are usually non-custodial, which means users keep control of their wallets and their private keys. That’s another major difference to centralized exchanges.
Decentralized exchanges gained strong popularity over the last couple of years, especially over the last couple of months. The trading volume on DEXes currently stands at around $60 trillion per month, and this is a 15 to 20% market share of the overall crypto exchange market. And in April, Uniswap, for example, as the dominant decentralized exchange, even overtook Coinbase in trading volume. But it’s not only Uniswap, although Uniswap really dominates the market, there are other DEXes like PancakeSwap, Dodo, or Curve Finance.
The question is, where does this popularity stem from? Why do DEXes become more and more popular?
In essence, there are, I would say, three advantages of DEXes compared to centralized exchanges. One big advantage is that more tokens are listed, so there are more digital assets available. It’s simply because of the fact that this classical listing process that you have on centralized exchanges doesn’t apply here, so there’s no regulatory compliance check. This is why new projects can be listed more quickly, or more or less, can be listed immediately after launch. That’s a big plus for DEXes.
Second, there is anonymity and privacy, since there is no KYC required as it is with centralized exchanges.
And last but not least, you could say there are fewer security risks for traders since DEXes are non-custodial. The FTX, Terra Luna collapses, etc., really made it obvious to users in the crypto space that this non-custodial feature is a big pro for DEXes.
2. Why and how did the DEX DNA evolve so rapidly? Where is the DEX market heading?
Karl-Michael: To dive a little bit deeper, I would need your help, Darren. There are two main types of decentralized exchanges, right? One is automated market makers and the other is order book decentralized exchanges. What is the difference?
Darren: The order book exchanges, they’ll be very similar to the centralized exchanges where you have a set of limit orders that are placed by active participants. And this is probably the most efficient way to discover price. But the reason that you don’t find this on most decentralized exchanges is that with blockchain you have these huge trade-offs. With Ethereum as the global computer, there’s a cost, every computation has its cost. An order book exchange, it’s very good at price discovery, but if you are very actively managing positions on something like Ethereum, you are going to go broke very quickly.
So, the vast majority of DEXes is using an automated market system. And what that is, is that you typically have a liquidity pool in which you deposit typically two assets. It can be more. But let’s use the simplest version. It’s a pool of two assets, let’s say USDC and Ethereum.
And what happens is you have a passive participant who is considered the market maker. They place their funds there. There’s a magical pricing brain called the AMM and that’s usually a constant function market maker, or it’s basically an equation that prices the tradeoff between volume and size, volume and slippage and size of an order.
And that governs the exchange rate between the two assets. And so this is a very simple construction. It’s something that’s cheap enough to run on-chain, and that was really what kicked off. The kind of the birth of popularity of the DEXes is the constant function market maker and the famous X times Y equals K formula.
Karl-Michael: In traditional finance, these are mostly institutions that are doing the market making. For DEXes, that can be retail investors as well, right?
Darren: That’s right. In the automated market maker version, it’s extremely passive. So, anyone can be a market maker. They place two assets in a pool. They’re exposed to the trade-off between those assets, and also something that’s known as impermanent loss. But they can make the fees as people trade against them.
And so this is bringing the function of market making to the average user.
Karl-Michael: Order book market making didn’t become popular also because of the fact that when every trade has to go on-chain this consumes a horrible amount of gas fees. So it’s simply much better to move to automated market makers. And correct me if I’m wrong, but I think that drove the adoption of AMMs, right?
Darren: That’s right. It’s really very expensive to replicate an order book on-chain because, if you have to update the global state to update an order position, that becomes extremely expensive. You can have something like Serum on Solana or, there are a lot of the PERPS (perpetual exchanges) that are running on L2, on Rollups, where the transaction fees are cheaper.
Otherwise, it’s impossible on Ethereum at these current gas prices.
Karl-Michael: Next to automated market makers or a very small portion of order book DEXes, there’s a kind of third category, which we call DEX Aggregators. Maybe the best-known one is 1Inch. What are these guys doing?
Darren: They’re basically doing smart order routing. So tapping into a bunch of different order books and finding the best path between points.
You get inefficiencies in DEXes as well. Let’s imagine, if you have two DEXes that exist in the world and someone takes an order on the other one, well, the pricing is going to be out of balance and it can be opportunistic for an arbitrage trader, or it can be opportunistic for an asset buyer who is going to get their best execution.
Actually, I saw a nice visual, I think it was on Dodo. It looks like one of the online travel agency aggregators where it’s taking you through one DEX and another, it’s like finding your best path.
And so it’s basically best execution order routing on-chain.
Karl-Michael: There is this constant function market maker mechanism to find prices or determine prices. But there are also automated market makers that rely on external market prices provided by Oracles or other exchanges. What are the pros and cons of these two different routes of price discovery, so constant market making versus the external based market making?
Darren: The external price based, it’s typically, price is being pumped in through an Oracle. So an Oracle is essentially a wrapper to an outside data feed. A data feed not native to the blockchain. So for example, if you’re looking at something like dYdX or GMX, which are perpetual DEXes, they need to get a price feed from somewhere, and then they have an internal system where people are betting long or short, and it becomes kind of a player versus player mode.
And this external price feed is critical for fixing and settling the instruments. The downside would be, of course, the reliance on the price feed itself becomes external to the blockchain, so it becomes an exogenous risk. But this is also pretty common for people to rely on these price feeds, to price different instruments.
And often you’ll have the arbitrage traders with automated trading even between centralized exchanges and DEXes that are relying on various price feeds, on- and off-chain.
Karl-Michael: If I use an Oracle versus a constant function market maker, how does this compare in terms of capital efficiency? Isn’t this Oracle version sometimes or most times better from a capital efficiency perspective, or has there been an evolution in the constant function market makers, which compensated for this initial issue?
Darren: It can be, if we are talking about the simplest form of a constant market maker, like Uniswap V2.
But you have concentrated liquidity on AMMs like Uniswap V3. Uniswap V3, you can consider it a trade-off between the constant function and an order book-based mechanism. And this allows people to price much closer and much more actively towards what’s happening on external markets. But there’s a trade-off because you have to be much more active. So Uniswap V3, it’s actually more capital efficient (than pure constant function market makers).
But you often see a lot of different triangular trade arbitrage, which is happening between typically a base asset — if we look at Ethereum, it would be the base asset — the Ethereum (transaction) cost, and then the dollar price. And someone is arbitraging that between the centralized exchanges, the Perps DEXes, and the Spot DEXes that have the AMM pricing.
Karl-Michael: Capital efficiency is one thing people have to keep in mind when they are using a DEX or trading on a DEX. The other thing, especially for liquidity providers is impermanent loss or divergent loss. It’s a very complicated concept. Can you explain this in layman’s terms?
Darren: The overview of impermanent loss is there is a concave payoff between two assets. In layman’s terms, when one asset is appreciating, you’re getting the other asset, right? If you have a quality asset and a subpar asset, typically your position will get full of the subpar asset. And this is an issue also with the shape of the curve of a constant function market maker.
You have different ways of addressing this that are coming out. It’s either redesigning the curve of the AMM at its core, and that could be something like Curve Finance does and I believe Bancor V3, you have something like Swapped Finance or you also have kind of derivatives of a Uniswap V2 position, and this could be something like Smilee Finance or Gamma Swap, which is essentially creating an inverse payoff.
So, you have kind of very interesting ways of solving this problem of impermanent loss or getting overweight in a depreciating asset. You have different ways to either hedge that or to reduce the impermanent loss, something like a Balancer 80/20 pool. There’s a lot of innovation around how a user can provide liquidity without being overexposed to potentially depreciating assets.
Karl-Michael: This is maybe one of the strongest innovation drivers currently in the AMM market. There are lots of new concepts and lots of new protocols popping up trying to tackle this issue. Are there any other big issues or problems you think need to be solved in the DEX space next to capital efficiency and impermanent loss? For example, what do you think about UI/UX?
Darren: We have a very different user experience. So, maybe a quick trip down memory lane: When I started 12 years ago, the UI/UX of Bitcoin was awful. It was absolutely awful. You had to download the Bitcoin core client. You had a popup that looked like a leather wallet, and then it had to sync for hours.
Fast forward to DeFi Summer, you have this crazy kind of this box trading screen, that was popularized by Uniswap, then copied by SushiSwap, which is you trade one asset for another and you get this little slippage calculation factor, and this is very basic, but the good thing it was simple enough for a retail user to basically effectuate a swap and to understand more or less what’s happening.
But you have other tools around there, like I believe IntoTheBlock was creating something like an impermanent loss calculator of the LP positions. You have something like DEXTools, which is something that I use on a daily basis, which is able to create a series of candlestick charts out of DEX trades and then putting up an interface over that, that’s really a huge improvement over like the basic swap screen.
I think that these tools will be built. I think that there’s a lot of kind of open data that you can get natively from blockchain even, calculating impermanent loss, calculating slippage, doing things like estimating a front running attack by an MEV bot, which is maybe a bit complicated here, but it’s basically getting subpar pricing. So, a lot of these things can be added in. I think it’s just a matter of time.
Karl-Michael: We have like 20% market share of DEXes now, it means that still 80% of the trading volume is with centralized exchanges. So, as you said, Uniswap is not difficult to handle from a UI/UX perspective, why is it still 80%? Is the major inhibitor here regulation or what keeps people away from using DEXes?
Darren: To be honest, I think it’s access and ease of use. So maybe the UX model that you were talking about is important. From my point of view, my perspective, most of the people are kind of crypto speculators, most haven’t actually interacted with a blockchain.
So, their digital assets are an internal database entry on, let’s say their Binance account or Coinbase account, for example. Whereas for me, I do most of my trading through DEXes, just because I like the experience and it’s easy. You don’t have to have the login. You go switch between any two assets. Sometimes you have high gas prices, but it’s actually for me a better experience than on the centralized exchanges.
And so typically I’m not using or swapping on assets that I don’t have wallets already set up on different chains. So, there’s a limiting factor in there. But I would say that most of the crypto or DeFi natives are going to be using the DEXes just because of ease of use.
And the other thing is counterparty risk. With the (centralized) exchanges, you have counterparty risk. And actually, my chief scientist, I met him first in 2017, our chief scientist at IPOR Labs, and he was building DEX software and he’s coming from the institutional side and he phrased DEXes s as non-custodial exchanges.
So that’s quite fascinating where the smart contract is your custodian. So instead of having the centralized exchanges that have a licensed regulated custodian, or let’s say the more cavalier ones, that we know like FTX was, it was judge, jury, and executioner, it was your custodian. In the DEX case, for better or for worse, the smart contract is your custodian.
So that changes the locus of trust, and to be honest, I prefer that.
Karl-Michael: I see more and more specialized DEXes, like Curve Finance for stablecoins, dYdX they offer margin trading, you as IPOR you’re offering interest rate swaps. Do you think that this kind of specialization in DEXes is going to continue, or let’s say in three or five years down the road we would rather see a concentration again because otherwise liquidity would be too much fragmented?
Darren: I think it has to continue, the specialization, and I think we can have better capital efficiency. So, there are these guys in France called Mangrove and the whole idea is you take a certain margin or you take a certain collateral and you can use it in multiple venues at the same time, up to a certain limit and (the protocol) does the capping.
The liquidity, I think, is a different beast that we’ll get to. We need different infrastructure layer tools. That’s also about capital efficiency. But I think the specialization has to continue. I think we do have an oversaturation of DEXes that don’t have really any fundamental value add. And we really need a maturation of the financial tools that are out there for the market to mature.
3. Risk-Free DeFi Yields, Interest Rate Swaps & IPOR
Karl-Michael: This is a good bridge towards IPOR I would say, because you are a specialized DEX, specialized in interest rate swaps. I think you need to let our listeners know what IPOR is doing. And by the way, what does IPOR stand for?
Darren: IPOR is the Inter Protocol Over-block Rate. So, you can consider a little hat tip to the LIBOR. But we embrace the best of traditional interest rate benchmarks and apply the best of decentralized finance. So it’s a block over block or real-time cost of capital inside of Defi.
And how it’s constructed is a protocol-to-protocol call. So, for example, the IPOR index, it reads from Aave interest rate contracts, it reads from the Compound interest rate contracts, and it creates a weighted benchmark rate that’s published to chain. And that publishing to chain is like this kind of Oracle construct we talked about, but it’s native to chain, so it’s like the risk-free rate inside of Defi.
And so, on one side, IPOR is a benchmark interest rate, and on the other side, IPOR is a derivatives DEX. And the derivative DEX in simple terms allows you to trade the direction of the IPOR.
So, in TradFi, if we look at some of the collapses in the US banking system recently, they were related to two things: Lack of risk management around interest rate exposure, specifically in the bond portfolios, and duration risk. And, actually, the two parts of IPOR are: This IPOR rate can be fundamental in building out term structure and managing duration risk. And then the interest rate derivatives help people manage their interest rate risk instead of the DeFi credit markets.
Karl-Michael: So that means IPOR enables users to go long or short on interest rates and therefore to hedge their interest exposure, you allow for arbitrage between interest rates and users can just speculate on interest rate trends. At the heart of all of this at IPOR are interest rate swaps. How do they work on IPOR exactly?
Darren: A traditional interest rate swap would be between two financial institutions that want to trade cash flows. One is betting that the rate is gonna go up over time, and when the other is betting that the rate is going to go down, typically you’ll have two different sides, a lender and a borrower and they’re exchanging cash flow. So over time, they’ll exchange the fixed leg for the floating leg, or the fixed rate for the floating rate.
In the IPOR context, you have one passive party. So we embrace the idea of a passive liquidity provider. For example, they put down USDC. And that USDC is out there to underwrite the interest rate swaps in USDC.
The AMM is a bespoke AMM built specifically for the IPOR protocol that essentially will bring in the different market dynamics such as models out volatility and in the V2 max notional exposure, max draw down, the different kinds of market dynamics that are happening in the interest rate market.
And it basically gives a spread above or below the IPOR and a trader can pick it if they want to either pay fixed or receive fixed. So, on one side of the trade, you have a passive liquidity provider who’s essentially the market maker. The AMM prices the risk, and then the trader decides to take the rate, and these two parties, the liquidity provider and the trader, are exchanging cash flows over the duration of the instrument.
Karl-Michael: And as a liquidity provider, what kind of yields can I get on IPOR currently? You have USDC, DAI, and you have Tether pools, right?
Darren: All three pools right now should be trending around 5% per annum as a liquidity provider. So let me just read for the past 30 days: USDC, this is USDC on USDC returns has got 4.89% APR. USDT is 5.56% and DAI is 3.89%. And where this yield is coming from is the stables. They sit in the pool by a depositor, while they’re sitting idle they go out into the money markets to yield, and then they earn fees from people trading against the pool.
So the fees can come from opening contracts, the net outcome of the contracts. And then finally asset management yield. And so it’s a simple way that the LPs are earning money.
Karl-Michael: And these yields are partially stablecoin yields, and the yield is partially coming from your own token?
Darren: No, so that’s actually stablecoin on stablecoin yield. This is what we consider the real yield. There is a liquidity mining function here. And if we’re to add to that, then the base yields would be anywhere between 14% and 18%. So that would be the liquidity mining yield earned on top of the base yield.
And so there’s an entirely other function, that I think it’s too much to cover here, but, the 3.8 to 5% is stablecoin on stablecoin.
Karl-Michael: That’s how liquidity providers and traders can make money. How are you making money as IPOR?
Darren: Well, so as IPOR labs, we build the protocol. We were contracted to build the protocol. In the future, we will ask the DAO treasury for funds to continue. But really the goal is to have everything happen at the token level.
The IPOR Labs team designed a new liquidity mining primitive or tokenomics primitive called Power Token. This is a modular architecture where you can have one token, you can stake that token, and then you can delegate that token to different modules.
So in IPOR, currently one of the modules is this liquidity mining module where you can get that extra yield from the liquidity mining rewards. That same token can be delegated to a governance module, it can be delegated to multiple modules. So the whole goal around IPOR is to build kind of a self-sustaining ecosystem.
Another part of that is governance around the IPOR index, right? This is, from a conceptual perspective, the most interesting thing where we have this risk-free rate that’s native to blockchain, that’s printed on chain, that serves as a historical record of the risk-free rate in Defi in the past, and in the future is used to structure different deals and derivatives.
So it’s this an on-chain — what we call public — good, that any protocol can call to, to get the risk-free rate at that moment, and that they can use to structure their own deals.
Karl-Michael: The Power Token, this is already part of what you call V2?
Darren: The V2 is coming out in Q3.
The Power Token was launched at the beginning of January. So the Power Token would align with our boost in TVL. Recently, there were two governance decisions around the Power Token. One was to change the shape of the boosting curve and the other one was to reduce the emissions by 30%. And that has resulted in a stabilization of the liquidity mining.
The V2 is looking actually to build in more of the utility around the interest rate swap. For example, what you mentioned, the interest rate swap really can help create kind of a fixed rate product. V2 will focus on not necessarily trading of the underlying instruments, but building them into larger positions: for example, a fixed rate borrow that’s pulling from the deepest liquidity pools in the DeFi ecosystem and giving the fixed rate borrow at the best available rate on the market. So this is really the purpose of the V2.
Karl-Michael: You’re currently operating on the Ethereum chain, right? Which comes normally with higher up to very high gas fees. I’m asking you a question, which you have heard maybe a lot of times, recently: When Arbitrum, when Optimism, when Layer 2?
Darren: We really had a lot of discussions inside the community. Also checking out the test nets. The big thing is that we’re waiting for this kind of V2 construction to go through audit, and then there’s a possibility that we could deploy on both Ethereum and Arbitrum test nets to make sure that everything’s working correctly.
It is a fairly complex architecture that also involves the asset management into the underlying protocol. So that needs to be tested. But if there is an IPOR V2, let’s say on Arbitrum, I imagine that a lot of the trading volume could go there while a lot of the kind of structured products would go on Ethereum L1. Because as a derivative protocol, it’s underwriting the most liquid credit markets, which exist on Ethereum Layer 1. And so I imagine, that these larger structured products might live there while a lot of the more trading volume would be on something like an L2.
Karl-Michael: The IPOR index currently is based on stablecoins, so you use Compound and Aave as the base interest rates. Principally you could do the same with Ethereum. Think for example about Liquid ETH Staking which is one of the next big things in the DeFi space. Are you planning to expand your range of indices?
Darren: I think that’s really the next step. So, we already have a study of the staking rates undergoing at IPOR Labs and the way that the infrastructure is built. We have an entire framework called the IPOR test framework, which is what we use for our quant modeling and backtesting against the actual blockchain system, built with the AMM so we can do essentially market simulations on-chain.
These derivative instruments are essentially volatility based, so it’s very straightforward to model them, test them, and launch them. The IPOR derivatives framework is pretty modular. So the next step is to study the behavior, construct the AMM, construct back-tested against the data and launch it in the wild. This will actually be a very big part of the v2.
4. DEXes and Regulation
Karl-Michael: Now we come to the very last question of this podcast. It is what we call the golden question. The golden question is normally a bit challenging:
DEXes and regulation, or crypto and regulation, we know that this is not the easiest relationship between the crypto industry and regulators. CEXes are facing more scrutiny, with the SEC and the US even going after Coinbase, although Coinbase tries to be quite regulatory compliant compared to other CEXes. At the same time, we hear more and more requests from regulators for DeFi companies to follow KYC and AML procedures and to apply, let’s say the classical regulatory framework to DeFi.
So, what does the future of DeFi look like? Will there be regulated DeFi platforms and non-regulated DeFi platforms in the future? And if so, will there be a dark DeFi market remaining where liquidity is provided by illicit funds? What’s your view on that?
Darren: I think that we’ll have a convergence. I think it’ll take a lot more time than we think. So, if we take the regulatory actions that are happening in the US, they’re extremely aggressive, and maybe they’ll push all the innovators out, and I think that’s maybe from a competitive perspective not a very intelligent thing to do.
Where you have the EU, which is typically considered much slower and much more conservative, that has a much more open, logical framework that is not an almost a hundred years old securities law construct that is being used to regulate something completely new.
There are a lot of things that are incompatible that need to be re-looked at. For example, if we talk about liquidity pools, right? If you go by traditional banking definitions, these are co-mingled funds. And to protect clients, you need to have segregated accounts. Well, in Ethereum, it is possible, but it’s not feasible and it’s not rational because it kills the cost of transactions and makes it not useful.
The other part is the KYC. Once you have this thing that goes across borders, it’s much more difficult to enforce, but it’s doable. But you have other kinds of risk-based approaches. For example, Chainanalysis has a tool where it essentially does like a dirty money score. And this kind of a risk-based approach versus a KYC-based approach is also extremely valuable.
So there are different things about actually the technological instruction that are also challenging from a regulatory perspective because it doesn’t match. And it’s not necessarily because people don’t want to build things that work with regulation, it’s the crypto space, and people are pushing forward on utility. For example, IPOR is looking at bringing utility and kind of this base stabilizing layer to the fixed-income market inside of DeFi.
But, if we were looking at the uptake of Aave Arc, for example, which is the permissioned pools of Aave, it really didn’t have any uptake. So, from an entrepreneur’s perspective, if you’re waiting for regulation, you’ll never build anything because you’ll never get clarity, and coming from a 12-year view in this industry, if I was waiting for regulatory clarity, there would be nothing, right?
But having an intelligent and two-way dialogue makes a lot of sense because something like interest rate derivatives is a kind of institutional tool. There’s a lot of back and forth that needs to be put into the protocol design, and it makes sense that large institutions will be coming into DeFi through fixed income. So I think it’s something that’s an inevitability.
What’s the timeline? I don’t really know.
Karl-Michael: Definitely. It’s very difficult to give a timeline, but I agree with what you were saying. Blockchain as a new technology offers alternative ways of doing KYC, and as you mentioned, Chainanalysis with their KYT chain analytics is a good example of that.
These analyses are sometimes more precise than traditional systems when it comes to identifying illicit behavior. So if regulators would be open to this new technology and if they start to understand it better, they would see that the blockchain technology is even more powerful and it will be even easier for them to handle the whole thing than in TradFi.
Darren, it was a big pleasure talking to you. Thank you so much for joining me today.
Darren: Thanks for having me.
