Markets, Manipulation & the Myth of Decentralisation with Jose Sanchez

Welcome to the Deco podcast where we bring you stories from founders and investors in Web3. Before we get started, this is not a financial or legal advice. It's just a conversation about things we find interesting in web3. Views expressed are our own personal opinions. Deco members or guests appearing on the show may have positions on assets they talk about. Welcome to the Deco Unfiltered. This is not a regular podcast episode. There is no structure, there is no script. It's just a raw conversation. But between two friends who have collaborated before, it'll be about crypto, about markets, and everything in between. Now, when we were thinking of this format, Jose Sanchez from Spartan was the first person who came to my mind. We have collaborated before. I know the areas that he works in. I know where he spends his time. So couldn't be a better person to start this format with. Hey man, how's it going?

Hey Saraf. Really good. What about you?

Collaborate? Allgood. We just took this article live. We published on Substack. It's related to buybacks versus like dividend share in crypto. And we were just looking at different stages in which different crypto projects are and what should they be doing with revenue? I mean revenue and excess revenue, right? I mean, only if there is excess like you will you try to distribute and think about what to do and so on. So like go read it. It's a nice read. I mean, we draw a lot from what happened in like traditional markets. Like how did public companies start giving out dividends? When did they move from dividends to buybacks? Why buybacks? What are the difference between the two ways of giving your profits back to shareholders and so on.

Yeah, yeah. I think it's interesting that we're starting to see more and more of these discussions happening over the last two months. It kind of shows like a real shift or more fundamentals or at least fundamental based investing. And I think it's mainly happening because there is like this push from two angles. One, the sell side research providers that are maturing, right, like you guys blockworth research and all the data providers like Dune, Defi Llama, Arkham. And then the other side is the buy side which is becoming more and more professional and more institutionalized.

You want to tell the difference between sell side and buy side before we move on.

So sell side is usually research shops that give like overview on different industries from a very research driven perspective. And then the buy side is like the allocators in the space. Usually in Shopify those would be the hedge funds and then the sell side would be like either equity research and also like some investment bankers when they prepare an IPO for a specific business. Yeah, right.

And like before we move on, like what are you doing these days? Like how do you spend your time at Spartan? What does the day look like for you?

Yeah, right now where the market is not as hot. It is mostly reading a lot of stuff, trying to look into like on chain data platforms like if you. So I've been for example learning a lot of SQL through ChatGPT and I've been spending my time like trying to do these deep dives on different protocols, like trying to build new Dune dashboards that aren't available out there and see if like there's like a change in the dynamics on a specific fundamentals of like different protocols inside a vertical. And yeah, also like following all the saga that is happening with Hyperliquid, it seems like every new week.

Yeah, it's crazy, right? Just give a like short overview of like what has happened and then we can talk about like what your thoughts are and so on.

Yeah. So it was yes, a few days ago where Hyper liquid listed perps on Jelly. And then there is this specific wallet that appeared which started to gather a lot of the open interest available on the market. And basically there was like one point in time where stat show open interest. It was so high that it was almost 40% of the actual market cap of the token. What ended up happening is this address was kind of like an attacker where it forced himself to get into liquidation. And then due to that, the way HLP works is they are the backstoppers of any like the ultimate backstop of a liquidation. So they were forced due to this trader doing that to accommodate and get into their short position on Jelly. And what ended up happening is like either this wallet on the other side of the trade started like longing the token like crazy. And other traders also seeing what was happening and HLP getting forced into that position. They were trying to push higher the price. And then this expectedly neutral centralized exchange called Binance came and listed the perps on Jelly and then CC started retweeting about it. So yeah, maybe you can follow on with this story.

But over here. Yeah, let's go ahead. This is super interesting. Right? So I think the on chain liquidity for Jelly was like two to. So if you have like reasonable amount of capital, I think the market cap was 24mil and liquidity. I don't know where it was. Like must have been early few millions. Right. Like at best like I've not checked what it was, but that's my guess. So it probably takes like a few hundred K to start like moving this price in either direction that you want. So what you basically do is like you go on an exchange like hyperliquid, take a big position there and then you use the rest of your capital to move the spot price so that the futures price on the exchange like tends to move in the direction that you want. And this is typically easy to do like when on chain liquidity is low, which is like almost for all the shitcoins that are listed like out there. But what I think went wrong for hyperliquid was I think at some point was a 200, 250 million coin if I'm not wrong. And at that point this might not have worked, but now the market cap has dropped by more like close to 90% and it just becomes easier to manipulate the on chain spot price which in a way like gives you a lot of ammunition to screw over with a perps platform like hyper liquid. So I think what did not work in their favor is maybe changing the limits as to how much you can borrow, how much you can lever. For an asset that has dropped by 90% and the current market cap is just about 20, 25 million. I think we need something like what did not work is like they probably did not like don't have this dynamic system where the leverage changes based on on chain liquidity, which means that if the price drops like by 80%, 90% for a token, you should probably change what kind of leverage that people get to trade those assets or just dealisp those assets. And yeah, like what happened later was crazy. I mean you see two sides of binance, right? When Bybit was hacked and ETH was like stolen by North Koreans from Bybit, you saw that they were ready to play ball with Bybit and you know, make sure that the customers are whole and whatever. But the behavior was reversed when something similar happened with hyperliquid. Yeah, I think that is the nature of this.

Yeah, I think also another interesting dynamic was that Jelly Jelly reached so much higher market cap because once people started knowing what was happening, they tried to like long jelly jelly like getting the bandwaggle of jjelly to try to make the HLP vault insolvent. So so as soon as HLP took the measure of the listing Jelly jelly and force liquidating the trader's position at a last look, Oracle price was what said like unilaterally, right. The market, the token price of jelly just dumped because the trade was over. So it was like this way of expressing a trade by people, traders that didn't mostly like hyperliquid, to try to make them insolvent, which was also kind of like, wow, yeah.

But what do you think of. I mean, of course there was a lot of criticism for hyperliquid, saying that, you know, it's not like decentralized and six validators that sort of convened and like agreed to take these measures were just like Jeff's different computers and whatever. What's your take here?

My take here is hyperliquid, since its inception and since it launched its first product, have has always been open with the fact that it's not truly decentralized yet. It's on the path to become more decentralized over time. But if it has to like make those decisions, it will make them. And like, apart from that fact, this is also like not entirely new. I also see it as the DAO hack when Ethereum had to roll back the chain and now create two different Ethereums. So in a way this unilateral decision of deciding the Oracle price is similar to doing a rollback on a chain when something crazy happens, like an immensive hack or something like that. So it's like even on decentralized systems and networks like Ethereum, we have seen this happening and obviously it's not the desirable thing to do, but it's just a matter of trade offs, I guess.

Yeah, I mean, and like just a caveat, right? Like before eth maxis come at us, we know that like these things happen with Bitcoin as well. But the point that we are trying to make is no system is decentralized to begin with. Decentralization is a spectrum. Like you start extremely centralized probably for almost any chain out there, and you gradually decentralize over time. Right. So nobody's claiming that issues were only with Ethereum and not other chains. No, like every chain has had these issues at the beginning and they have taken measures, like even Bitcoin, right? They, they have taken measures to ensure that the chain survives first and then we can talk about decentralization and all those things. So yeah, it makes sense to have like hyper liquid, you know, I mean, and them owning it is fair, I guess, saying that we are centralized and we do whatever it takes to make sure that the product exchange survives first and then we can do whatever we want with desimalization. It's all fair. Yeah, that was good. I don't know, I mean, these attacks, like, obviously they'll learn from this. I think something similar happened to Hyperliquid a couple of weeks ago, but at a smaller scale. Yeah, but I'm sure that like they'll keep at it and make sure that these things don't repeat. Going by what is happening with Hyper Liquid and how the product has evolved, like it isn't like unfair. It isn't too much to say that like they will learn from this and there will be mechanisms in place to ensure that this doesn't repeat. All right, so we've written I think couple of articles both related to Intents and Oracle extractable value and so on. So you do spend a lot of time in the intense space. So like I think a year ago, before AI and like Meme coins like sort of took over all the Mindshare, there was a phase in which like Intents kind of ruled. Kaito was not there. So I don't know what the Mindshare was. But yeah, like what is happening in the space? Like was it also one of those cases where we overestimated what impact it could have at the beginning and like just tell us what is happening.

Yeah, so I think what happened with Intents is that they were very useful and they basically made bridging very cheap and fast. And that was like the main unique selling proposition. And everyone was like, wow, okay, so this really works. And the user experience vastly improved like a lot. So now it faced like kind of like you don't really know where they sit, where they had like a little mini round maybe when defi. I don't even know how to pronounce it but decentralized finance with AI, these chat LLM prompts and chat based interfaces which kind of had like a first wave of innovation but then died off. And the main reason why they died off is because the product wasn't what users expected it to be. No one was to really like type into a chat based LLM what they want to do on chain. It has to be like the application itself providing recommendation in defaults. Like hey, this is happening right now in the market. Would you like to long this specific meme coin that just launched rather than you yourself having to be the proactive one and saying so today I want to grab my Solana, grab my Sol in Solana and like go in base and buying this token. So it has. I think they failed because one, they weren't that interactive and it required the user to be the proactive one and two, their user experience in actually doing the actions cross chain. It wasn't the Best because they were not connected to solvers yet. So the way they did it is still using the message based bridges to go cross chain which they are not the fastest at least for the medium to lower value transfers. So like the medium average transfer of 5k or 10k, that's where intents excel and yeah, so that's what happened with Defi and why it died. Also what was interesting as of this last month was the recent push by the Ethereum foundation for this initiative called Open Intents Framework which is being led by hyperlaying across all of these interrupt based teams building on ETH layer 2s. It kind of shows that Ethereum needs like becoming aware of the problem, the problem it has with all the layer two's fragmentation of liquidity. And it kind of also connects with this unichain framework and the super chain native Interop that the OP stack is developing to go live by the end of the year. So it's like great to see all of these different efforts being developed by the foundation and them acknowledging and creating these groups.

But yeah, yeah, let me just interject here, like you said a lot of words, can you just simplify this? I mean for whoever is listening, right, as to what do you mean by Interop like and like why is it that Ethereum foundation did not recognize this before? Because it's not like a year ago the situation was different. Right. I get that there has been like change in card with Ethereum foundation and there has been personnel change and you know, there's this new entity that is for profit and all of the changes like just take us through.

Yeah, yeah, yeah, yeah.

All of that has happened.

So I think this has happened because Ethereum foundation, all the community members and all have realized that intents are the way to go for asset based interop which is like 90% of the use case. When you want to do a cross chain action, you want to breach over to a different layer two to buy a specific token there or because on a specific pool on a different layer 2 it may be the most liquid one venue or avenue to trade a specific token. And so that happened, Intents penetrated the market a lot. I think right now they make for 50% or more, 60% of all breach volumes that go between layer twos. And the reason why it happened is because it was faster and cheaper when you breached over less than 20k, 25k. When you go over those amounts you kind of like get into some scalability problems where the liquidity networks are so much better.

But yeah, but it makes Sense. Right. Because for someone who's breaching like over 25k or 50k, they are not as sensitive to price pricing as someone who's moving like $1,000. Right. I mean optimization in fees basically. And you know what kind of price you get. Whatever is in terms of percentages obviously affects the people who are trading like with large volumes, but as a sheer number it probably affects more to guys who are trading like five hundred or thousand dollars.

Yeah. And also like fundamentally an intent based protocol cannot really serve those large transfers because the way it works is so intense. They need like this middle layer called the solver layer which actually does the transfers for the user. Right. And these solvers, they have given the upfront capital on the destination chain, they have inventory. So given they don't hold a lot of inventory, else it would also have to be like scattered around all the different layer twos, they have like capacity problems. So they work very well and they optimize for lower value transfers because they hold lower inventory in each chain because they can rebalance and utilize it more efficiently versus holding like 1 mil on every layer 2 where they would only like utilize maybe 10% of that. So it's better to utilize a 50k inventory of a solver doing like 80% utilization than having one meal inventory on every layer 2 doing only single percentage utilization. Yeah.

Would like retail supermarkets be a good analogy here? Because like you want to buy goods of goods and like whatever products from different companies that are like located in like different geographies in your country or outside. So like without solvers it is like you going to individual company and telling them what you want and them shipping all these things to you. But with solvers it's like you are going to a local supermarket and just like buying all the stuff that you need and like. So instead of interacting with the companies directly, you interact with them through these intermediaries that are the supermarkets. And say you want coffee, right? You want 10 packets, 50 packets. Probably supermarkets can help you. But if you are looking for 100,000 packets, you would probably have to circle around and go to a higher center. Like is that ballpark like decent analogy.

Yeah. I also think that's the aggregator thesis where you would have like this layer on top which basically makes sure that the trades, depending on what amount you want to breach, just get routed to the most efficient methodology, whether that's intent based or that would be like the most, probably the most efficient for the lower value amounts or message based or even mint and burn for the higher value. Like one mil reach orders. Yeah.

All right. Going back to the Ethereum thing and how it's evolved and what the problems were and how it will change now. Like so what is the new infrastructure? In what ways will it change the current fragmentation of liquidity across various MTOs.

Yeah, yeah, this is interesting. So the way Ethereum layer 2s have developed is now you're having like these different clusters of layer two. So you have the OP stack, you have the arbitrum orbit stack, you have the CK sync elastic chains and then you have. Yeah, and then you would have the Polygon AG layer. Right. And each of these different teams is developing their own native Interop solution for message passing. So like from passaging messages from one another, that's not asset based, that's just message passing.

Is it only within their ecosystem that they are developing or this is like a standard that every.

Yeah, they all want to make it for every ecosystem. But to be fair, it's mostly focused, at least at first on their own ecosystems.

Right.

But the cool thing about intents is that they sit like one layer above that. And the way we work is like so just to walk through like a life cycle of an intent. So I think this could be understood better is like a user express an intent, the solver fronts the capital on the destination chain that they want and then the solver gets repaid later by the intent protocol. Right. And they get repaid sending a message of the proof of the field that was fulfilled. So this sending a message of the proof, this proof, it's currently done by different designs. And what the different layer two interop teams are building is for this proof sending to be more efficient and more fast so the solvers can get repaid faster. So if they can get repaid faster then they can provide better fields because they don't have to price that latency that they take to repaid that opportunity cost. And so once these different layer 2s develop their solutions, which will probably make this solver repayment go from like right now maybe like one hour to just a couple of seconds. What you'll see is like now solvers get repaid insanely fast after they solve a user intent. So they can provide like way better price quotes and everything gets like way more efficient. So you will just feel like the vision is like making Ethereum feel just like a normal layer one. And on the back end it would be like all these different solvers rebalancing themselves but feeling their users intents at a like 1 second, 2 second latency. So like 12 second click experience is the angle and that's the path to make that happen. And we're close. Yeah.

Okay, so who is benefiting out of this? Like what was it like without this before? Who was at a disadvantage and after this like who benefits and what can be the magnitude of it?

Yeah, I think the main obviously like the main beneficiary is the users. The user experience, it will just get insanely better and from before so it will just get really like interop could get commoditized because of that. Because right now like if you go from charging right now, every breach charges maybe on average like even though it's not that much. But 12 VPs per transfer, a percentage base fees per every breach transfer is if the system gets more efficient and their margins just keep going down, they'll probably suffer a little bit. The other way to see it is that maybe you get a lot of MEV opportunities after that so you can recapture that and send it back to the user or keep it yourself as the platform that generates the mev. But definitely the losers could be the current breaches that are charging the highest fee percentage based transfers. And the winners is mainly the users, the front ends, the people that aren't the customers because they can just have different business models. Not really reliant on breaching transactions.

But yeah, and who all are like what are the leading projects or protocols in this space right now?

Yeah, so the protocols pushing forward this are hyperlane which are also reactive in the intent space. They are both like a layer to your general messaging passing bridge and also they are building on top the asset layer, the intents layer. The other one is a cross with the standard 7 CSA3 which is like this crossing standard that aims to standardize every crossing intent and they aim to be like the sediment layer of all them all. And there is like a lot of efficiencies if they achieve that vision because you get like a lot of interesting coincidence of ones if that happens. If all like all flows through that standard are using that sediment layer. And yeah, maybe also unichain as it aims to become the defi liquidity layer where basically all of the liquidity sits on unichain and the different layer twos and the different apps can just access it very quickly and very cheaply through this standard.

Okay, so coincident of wants is nothing but you want to buy something, someone else want to wants to sell something, it'll just match you too. So for example, if I want to sell like 2 ETH for I don't know 4400, 4500. I don't know what the price is right now. And Usain on the other hand wants to buy it. So instead of going to a Dex and like using or wasting up all that cash and whatever, the matching engine will just match our orders and it'll get squared off. And what he meant was like as the scale increases, which means that as more and more users start going to across the likelihood of this. Right. Coincident of wants where orders get squared off like that increases and that makes the whole system more efficient and in terms of the pricing that you get. So yeah, yeah. Anything else that you're looking at, like what is interesting these days? I mean the markets have cooled off. What is the mode at Spartan and if you can share something.

Yeah. We have been spending more time on Solana so it's been interesting seeing that previous to all the hype and the Amin going mania and all the trading volumes, there really wasn't a lot of competition inside of Solana. Like the Dexs didn't really compete with each other. It was kind of like every vertical has its own winner and it lift off a monopoly. And we've seen how this has started to change. So the main change has been Palm Farm really launching its own Dex. Right. And Raydium kind of like counter attacking and trying to launch its own token issuance or launchpad. So just show like overall competition is heating up and in this scenario. So I've been thinking like okay, if this happens, like where do you want to be as an investor or even as a builder? I think really it all goes to like owning the user, owning the dapp, owning the front end interface. You cannot get this intermediate like Radium had like experience from Pan Fan. So we were trying to like kind of see which are these builders or teams building like cool apps for trading or for whatever defi action you want to do on chain and yeah, just scouting like the mapping and yeah, what.

Do you think is more difficult? Like building a Dex or building a launchpad. I mean it's not an easy to ascend, right. Otherwise you just go and buy whatever. Right.

But yeah, yeah, from a tech standpoint probably building a Dex is more difficult but again you could just argue it.

Can be just a fork of radium or whatever like spoke of ice rover or whatever.

Everything that's already been invented techwise on the text space. So you can just fork. So yeah, I think the most difficult thing is like a launchpad is a form of a product. I think the most difficult thing is like building a very good product that people want to use. And like, if you do that, you're just going to own everything. And easiest, like found or Found was very successful because they spilled a very good product that their users liked and then obviously the time was massive. Making launching tokens very cheap and fast and easy happened to be a very insane time.

So, yeah, even many projects tried to replicate pump fun and it did not work for them. Right, Take your pick. Like Pump fun for xyz, right? Yeah, yeah, that did not work. Like, there were attempts on base, There were attempts on like every other chain, but it just did not take off.

Yeah, usually the copycats don't really take off. Pandotfang was the first and that's why probably the first movement advantage that they got was insane. Yeah.

Do you think it is also a function of liquidity that the others did not take off? Because, like, for whatever you said and done, like Sonala had this like crazy liquidity anyhow. Right. I mean, you saw that there were weeks when the stablecoin market cap on Solana was just like, I don't know, few hundred millions were getting at it like every week. And that was not true for every other chain.

Yeah. But my take is actually like Pandora Fund was the reason why it's like a cat and mouse game. Like, in my view, Pandorfan is the reason why Solana.

Why the liquidity. Correct.

Yeah. Yeah. And like why all the liquidity actually ended up being on Solana. Yeah. That's one way of how really like a very good, successful app can make for a chain. Like the impact it can have on a chain.

100%. Okay. Cordells, is there anything else?

Yeah, I guess the other theme that I'm looking into on how to play it is stablecoins. It's really been the only vertical or.

Consistent vertical that just keeps growing and.

Growing, just going up. It's crazy. It's like obviously like you cannot own tether or circle, else everybody will be like very happy with it. So it's like trying to see how do you play this? How do you play this? Do you play like the decentralized stable coins? Do you buy maker because of dai? Do you buy curve because that's where stablecoins go to earn yield. We've been toying around with the idea that we think pendle. It could be very interesting if the stable coins keep going up because if you actually look at the TVL and how stablecoins are being utilized on chain, the main product or protocol that they Go to is Spandle because they want to either park their stablecoins to earn the yield from the other side of the yt, the PT to earn the pt or because they just want to trade the yield. Like people saying Athena SUSD is too cheap. The yield that it offers. I'm going to long it buying YT and yeah, it's been showing its hand as the best protocol to just capture all the stable coins that want to earn yield on chain.

Yeah, that makes sense. What were your inhibitions towards Athena?

In what sense?

That it can't scale. Like what are the issues with the model? Because there can only be so much open interest in the market, right?

Yeah, more than that I think it will just perform well when there is like a decent bull market or volatility because it depends on the funding rate. And the funding rate, like open interest is a derivative of the funding rate. If like a lot of leverage happens, the open interest is going to go up, funding rate is going to go up and then Athena can earn a lot of GL dollars, revenue and all of that. So it's like a leveraged bet on crypto. That's my view on Athena. It really is like a leveraged bet on crypto.

I think one of the things that they had published to counter this, there were very few times in history when the funding rates were actually negative. For the most part the rates are positive. And if you look at yields in general, of course they drop across the board when the market cools off. Like for example, there are days on apps like Camino and AAVE where you see like 25% yield on USDC, USDT, whatever. Those are very short lived but I think it all comes down to 3%, 4% and stuff like that. So when the markets are not very hot, I've also seen like 0% and 1% on these apps. So I don't know.

I mean, yeah, right now it's very low too. I think it's one of those like, like one every hundred days periods where the funding rates are not very high and even low. And it goes along the idea that there is like all of this insane reflexivity in crypto where you just need like a small flame to kind of like ignite it all up again. And we're still trying to find that flame again after these last two months. But as soon as like something like, I don't know, a new innovation happens or that then it's like everything goes crack up.

Yeah, that makes sense. All right, I think this is good. Unless you have something else to talk about. If you have anything for me, we can.

Oh, yeah, that's pretty much it for me. Yeah.

All right, cool.

Awesome sounds.

Let's see. This was fun.

Same.

Thanks for hopping on, man. Thank you.

Yeah, no problem, no problem.