A (2:34)

There's one paper, there's one paper before me, but I don't think it formalizes ADL itself, but it does kind of formalize what Bitmex did, which is made by Kyle Saska, who is the chief CIO at whatever flood's fund is called. I forget.

B (2:50)

Okay, well, there you go. Yeah, so. So you released this paper, I think, soon after the last show, and immediately after you released this paper, it triggered a firestorm. So there was all sorts of angry people. So I should note that one of the headline claims in your paper was that hyperliquid over aggressively auto deleveraged traders to the tune of hundreds of millions of dollars in counterfactual P and L that was lost by traders or something. Correct me at any point if I'm mischaracterizing what the claim in your paper was.

A (3:22)

Yeah, I will anyway later walk through the exact thing as we. Once you explain.

B (3:27)

Can't wait. Can't wait. Okay, so the HyperLiquid crew, led by Jeff, the founder of HyperLiquid, who's been on the show, Jeff, tweeted in what is now an infamous tweet, I think he said, those who can, do, those who can't fud. And basically said, tarun's paper is total garbage. Can't believe these are the people you look up to as academics. This is bullshit. Nothing in this paper is correct. Totally mischaracterizes how Hyper Liquid works. Tarun then clap back and said, wait, no, you guys are misunderstanding my paper. Blah, blah, blah. Then Dan Robinson from Paradigm. Paradigm, also a big investor in hyperliquid. They ran the big Hyper Liquid dat. Dan Robinson from Paradigm came out and said, tarun, I am also fellow academic, you are totally wrong. And the two of them just started sparring for about a week of just going back and forth and dueling threads, mostly in latex, describing each other as morons in various different ways. So it was kind of a battle of the titans. It seems to now have been resolved. So for whatever reason, all the conflict that was going on between the two of them. Now, apparently you're best friends. And Dan Robinson announced that he was going to be helping some of the data analysis of the V2 of the paper, incorporating some of the feedback and corrections that many people from the community ended up giving you. But it sounds like it was a very harrowing week for our hero to run.

A (4:44)

Yeah, my eye bags, who actually watch on YouTube, my eyebags are much larger than those normal.

B (4:50)

Yes, yes. The misadventures in talking about adl. So, Tarun, tell us, how's your week been?

A (4:57)

It was like, basically I posted that. We recorded. I posted that and then I got on a flight to Abu Dhabi and I didn't have Internet on the flight. So, you know, I landed and then I saw the Jeff response. So, like, I hadn't realized the commotion that had happened there had been like a little bit before I boarded. And then when I landed, I was like, oh, wow, I caused a shitstorm. I think it was a good lesson in, like, how research has changed in crypto, in my opinion, over the last five years. Like, five years ago, you read a paper, people are not immediately like, oh, you're fudding my bags. Like, get the fuck out of here. You wrote this research strictly for fudding, right? Like, whereas, like, now it's like, immediately assumed that you invested in lighter or drift or whatever, so you must be fudding my bag. So that was annoying. But I think, broadly speaking, I just kind of think the attacking thing has kind of gone a little crazy the last few weeks. I think, for instance, there's been lots of other research kind of trying to attack Polymarket and Haseeb and Tom. Maybe when Tom's here, he can talk about that as well, that there's sort of a sense in which some of it feels a little bit unfair, targeted, especially when there's like, things have real merit and things that don't. But the idea that all research is marketing and that the marketing exists only to fud someone, I think is somewhat disheartening because you can't really write anything without getting people mad. Now, I will try to explain in the simplest terms possible what I stated, what I got wrong, and what is still right. So Haseeb mentioned auto deleveraging and sort of this Edge case event. And so for me, October 10th was sort of like a crazy, very crazy day, I think I talked to Vlad right after and I was like, it's kind of insane that ADL was triggered 20 times in a row or whatever. You just don't see it wasn't Invented to be triggered multiple Times. And on October 11, there was a tweet, a thread by someone who, I think this is their first ever tweet thread that at least I recall, which is don Wilson from DRW. Who large market maker, very big TradFi HFT player, does a lot of stuff, person I have a lot of immense respect for. And he had a very good poignant point, which is like, if crypto can't fix adl, if they don't make it so that it's friendly for traditional players, we're just going to have the FCM system, which is like the clearing system that we have in normal tradfi, where you have these counterparties, these third parties who basically you have to effectively custody with in some indirect way and they take the risk of these kind of rare events, like they can lose their capital, but then you have to pay them fees to do that. And I think that tweet thread just made me say, like, feel this thing about decentralized exchanges, which is, okay, we have to have something that's better than this. Otherwise, like, yeah, if we end up in the SDM land, then we're basically back to centralized exchanges because fcms, again, it's like it effectively means someone custody is your assets. And so actually a couple weeks before I put out the paper, while I was still working on it, Arthur Hayes wrote this really good blog post about the history of liquidations and ADL in perpetuals markets. I highly recommend. He has a lot of posts that maybe are just price posts, but this one is actually a history of why the perpetual future had certain things in it that it did. And so the history of ADL is in 2015, Huobi, the exchange added this algorithm for dated futures, like futures that expire monthly, effectively. And what had happened is prior to 2015, Huobi had basically said, hey, if you get into this. So ADL is basically when no one can liquidate a position profitably, effectively. And so the exchange, in order to stay solvent, has to take some of the winners who have positive profits to cover some of the negative profits. So it's sort of if I'm a winner, I made $100 on the exchange, but there's this person who has negative equity, sort of like their profit is much less than the collateral they put, and no one is willing to liquidate it, then me as the winner, I socialize the loss, I cover some of their loss. And that is, that is sort of at a high level what this lost socialization type of thing is. It's just like in a case where you can't remove insolvent positions, you take money from the winners and try to regain solvency.

B (9:34)

So Quobe before that, it's also opposed to the old school loss socialization is literally everybody gets a haircut. The loss is split parada among everyone. That used to be the way that it worked before ADL's became the predominant way that it's done. One analogy that I've given to somebody who's very not in crypto about ADL's that ADL's. It's a little bit like when a company declares bankruptcy. It's like a stop the world, nobody move. We are now going to reorganize the debts to make it so that this company can start operating again. And then it's like, okay, great, now we've cleaned everything up. Now this company, it's too bad, so sad that you didn't get everything that you wanted. But the company's in bankruptcy and you send it out in the world again. And if immediately after you send a company out after bankruptcy, it declares bankruptcy again, it's kind of like, okay, this is really not designed to do this. And I think that's kind of the core of what Tarun's paper was getting at, if I can give my kind of 90 IQ.

A (10:30)

No, no, that's actually a good analogy. I like that. The interesting thing just in hindsight is huobi apparently was pre2015 was just paying out the exchange took the risk on their balance sheet. So when the bad debt happened, they would just cover it with their profits. And then apparently some event happened in 2015 where they just lost so much money and there was sort of moral hazard. And Arthur's post talks about other exchanges where this happened, where by everyone knowing the exchange would pay out, people started taking riskier and riskier positions because they were like, the exchange is going to cover it anyway. It's like heads I win, tails you lose type of thing. So anyway, I was sort of this mindset that hey, if this can't happen again. And people have basically oh. And so Quobi invented this algorithm. Bitmex adopted it for perpetual futures in 2016 and then Binance copied it in 2019. And then since then it's kind of effectively been the standard where you order all the positions from largest to smallest according to some metric. And the metric varies a little bit, but most places it's the profit times the amount of leverage. And then you start haircutting people one by one until you cover now the goal of the paper was like, hey, are there other methods we could use that would do better? Are there other algorithms that you could do? And so the paper shows there are other algorithms. There's also trade offs, like some things you can never fully get rid of, like this moral hazard type of stuff. As an exchange grows, you're inevitably going to have that. And also some notion of fairness to traders. This algorithm I just described, where I order them one by one and then knock them out is kind of unfair to the top person, right? Like if I'm the top trader by leverage and pl, I'm the first one to get cut out. And I think this was actually on social media. A lot of the complaints about ADL was like, some people certainly made money, they happened to have the right direction, but some people happen to have the wrong direction and where they felt penalized for that.

B (12:26)

Vlad, I want to bring you in from what you saw from both what happened on 1010 and your learnings in terms of how LiDAR implemented ADL and, and also from your discussions with Tarun. What do you take away about the design space of ADL that we've, that we've learned or that you've learned personally?

C (12:42)

Yeah, I mean, I think that tarun's paper, to me, like there was a lot of like, interesting analysis of how different algorithms could work and you know, simulations of those algorithms and different scenarios. Like I think there's a lot of cases to think through with like this algorithm where kind of like this greedy algorithm, right, where you start with like the biggest profits, biggest positions. I mean that from a certain perspective that makes sense. On the other hand, like if you're doing some kind of like pairs trading position where, you know, you're long eth, short btc, then that algorithm isn't very fair to you. Right? Because like you've actually, yeah, like you have this one leg of the trade that did very well and maybe you can say, well you could take a bigger haircut than other traders, but you actually had a full loss on the other side. So I mean, I think to me part of the point of the paper is you can't necessarily design a perfect algorithm. So there's also the question of, I guess from where we sit at lidar, we want to verify that everything we did is fair and to do that we have to compute all of these claims and generate CK proofs on them. So if the algorithm is too complicated, that creates other problems in that you can't really prove it in, you know, anything close to Real time. So our algorithm actually is also pretty, pretty heuristic based. I guess one difference is that it's just less aggressive.

B (14:18)

So it's not the vanilla Q algorithm.

C (14:21)

That you know, it, is it? I mean, I think it is similar to that as well. But, but there's, there's parameters to it, right? It's like there's, there's the Q, but then there's a parameter of like when do you start pulling from that queue? So you can kind of use the insurance fund, which for us is the llp like up to a certain point and then only after that do you start pulling from the queue. So there's parameters like that as well. I think like in a perfect world, if, you know, if the computational throughput wasn't a factor at all, then probably some combination of the kind of the greedy algorithm and the socialized algorithm would be best. But, but there's, yeah, from our perspective there's also the computational aspect to consider. But in general we've tried to make the system kind of just like friendlier to traders. And I think you saw that on October 10th.

B (15:17)

But you mean that with respect to LLP taking on some of the liquidations as opposed to rejecting them?

C (15:23)

Yeah, that's right, that's right. And I mean LLP did take a hit that day, but a lot of traders didn't have losses that they would have had otherwise. Right. In fact, like, I think we've made this point before but you know, you know, LLP was down 20 million that day. But that also means all the other traders collectively made 20 million since, you know, perps are zero sum. So. And by now, you know, LLP has made most of that back as well. So something like that happens every five to 10 years. From our perspective, that's the right trade off, but I think that's a little bit of an orthogonal point about. Yeah, go ahead.

B (15:59)

Yeah, well, one of the points that we were making on 1010 is that part of the reason why 1010 was so disruptive was because it, it was very unpredictable. Right. People who thought they had hedges in, those hedges exploded. And a lot of this of course is because you can't just net on a single exchange and understand what somebody's exposure, what their risk profile is. Because if I'm super long on hyper liquid, but I'm super short on lighter or vice versa, somebody you can't tell in isolation, oh, this person is like, this person looks really risky, but in reality they're taking what they think is like actually a very safe position. Across two exchanges. But the moment that you blow out one of their hedges on one side, all of a sudden now this person is taking a lot of risk and they start going topsy turvy and all sorts of chaos ensues. And so the, it felt like this is one of the points I think that Anand made in response to your paper Tarun, is that one of the most important things is just predictability. It's just the ability for large traders, market makers, people are doing carry trades for them to be able to understand and predict when ADL is going to happen and what happens to their positions on Net when an ADL takes place.

A (17:10)

Yeah, look, I don't disagree with you, I just think. Or with that claim. I just think there haven't been many open source implementations. The documentation for this doesn't. If you read everyone's docs, they don't really lay out what they do. And I think the controversy is because of how I estimated the number, because I made a mistake in the paper where I haircut the entire position. So profit plus principal versus just haircutting the profit. And it turns out that the space of algorithms doesn't change. All these improved algorithms that the paper claims still work in the profit space. But generally speaking, the winners on hyperliquid were over collateralized in the sense that they had 13% of their positions were P and L on average. So I sort of overestimated roughly by a factor of seven. I'm working on getting the exact number, but you can roughly think of like there is the. The paper said 650 million, it's probably more like 100, but there's still this amount that was over liquidated by this algorithm. And fundamentally I think that is a sense of unpredictability, especially if you can't verify the state of the algorithm because it's not open source. There's no explicit description of what the code is doing. It just kind of tells you I'm going to rank you and then liquidate you. And to Vlad's point, there's a ton of parameters to that. How you do those when you maybe go to a liquidity pool like an LLP or hlp, the timing, like how aggressive you are, maybe instead of fully liquidating, you partially liquidate. There's so many parameters that if you don't see the source, you don't really know what's going on. Right. So I think that's certainly why the hls went after me and still are. But I think there's still a fundamental directional truth Right. Even if the magnitude is wrong.

B (19:01)

Well, I mean, the other reason is because also you're an investor in later, which I think you mentioned that. But just to underscore it even more, that's a big part of the reason why they went after you.

A (19:10)

Yeah. But I want to kind of underscore one final thing, which is the goal of decentralized exchanges is to improve upon centralized exchanges. And one thing is a lot of people who worked at centralized exchanges messaged me, DM me and said like, hey, these ADL repeated events have happened before at centralized exchanges, but no one knew because you can't really see it. There's no public data. You only see yourself getting ADL'd, you don't see other people. So you can't measure how big of an event it is. But I talked to some people from Huobi who were working at Huobi 2016-2018 and they said this occurred multiple times and Huobi just paid people out. So it basically reverted to insurance fund once they did the adl. And then if it's sufficiently large, they had enough complaining traders that they paid them out. And you can't really do that decentralized exchange. Right. And I think that's sort of why.

B (20:03)

You need it do all sorts of stuff. If you are sufficiently motivated, you've been at DEFI long enough, your token is an insurance fund is like the meme. Right?

A (20:12)

That's true, that's true, that's true. I guess, yeah, maybe that's fair. But it's interesting that this was the first time you could see this event. Right. We should kind of be.

B (20:23)

And that's to Hyper Liquid's credit. Is that okay, Maybe you can't read the exact ADL algorithm, but you can see all of the ADL's. You can see the record of this person was closed at this price at this time on this chain. Otherwise you wouldn't have been able to do this analysis at all.

A (20:36)

Yeah, exactly. And I think that's already an improvement.

B (20:39)

The other element of this, of course, is that you were sparring not just with the Hyper Liquid hls, you were sparring in particular with Dan Robinson. And I want to ask you about that because the fact that Paradigm was a big investor in Gauntlet, wasn't it?

A (20:55)

Yes, still are, I guess. Yeah.

B (20:58)

Still are. Yes, presumably still are. So any reflections on getting into a sword fight with your lead investor?

A (21:07)

I guess you recently had a podcast series about the same type of thing.

B (21:12)

Did I? Wait, what? Wait, who did I get in a.

A (21:15)

Sword fight with Santiago.

B (21:17)

Oh, Santi. Oh, that's true. Yeah. Okay. I'm his. Okay. Yeah, you're right, actually.

A (21:21)

Yeah. Right. Isn't that the same.

B (21:23)

Yeah, yeah, that's right.

A (21:24)

That's right.

B (21:25)

Fair enough.

A (21:26)

I just think, obviously I'm not trying to create more beef. I have my own opinions and thoughts as to what. What motivated the level of aggression, but I won't share them. But, you know, I think things could have been handled differently. But, you know, I don't know. At some. At some level, I'm just kind of happy it's over because, like, it was just a lot because I spent. I feel like I spent two months doing this as a labor of love and, like, you know, like, there's.

B (22:00)

There's nothing I.

A (22:02)

Expect to get out of it, right? It's like, I'm not.

B (22:04)

Right.

A (22:05)

It's like, literally a public good. And I feel like I was like, you know, that meme of, like, the open source software developer who, like, there's like 500 people in their GitHub issues saying, like, why didn't you fix this thing? Why didn't you fix this issue? And it's like, this person did this for free.

B (22:20)

Yeah.

A (22:20)

And is not really gaining much, is in fact only losing based on. Based on the last.

B (22:25)

Well, there's. There's two. Well, there's two elements of it, right? Like, one is that it's like that. That saying from the wire. You come at the king, you best not miss. Is that like, look, you come after hyper liquid. Like, you better come correct, you better bring the receipts. You better, like, have all your shit together. So that there's a part of that. That's true. There's another part of that that, like, yeah, people don't know what the culture of crypto was like before the kind of modern token industrial complex. You know, like, in a way, it's almost like I feel like there's a. There's a little bit of. I don't know, like, think like pharma research, right? Like, I. I could not imagine reading pharma research and not just being immediately jaded. They're like, obviously this person is fucking paid by pharma to, like, get some result. They're like, oh, antidepressants are like, great. And I'm sure there was some time when people were just in pharma research for the love of the game and people were experimenting on themselves and like, oh, I wonder if this thing will reduce my dandruff. And their haircuts are on fire. There was a time when that was actually how it worked. And I remember when you were writing papers about defi, about the staking rate versus the organic yield that was on chain and how that produced instability and consensus, you were writing papers like that. It wasn't like, oh man, why are you running at compound? Or like, oh, why are you talking shit about Makerdao? It was just. Yeah, we didn't know. And so we were all exploring this giant terrain together and now it's like, well, everyone's got a bag. Everyone's got. Yeah. And look, it's different now. You're an investor and so it is genuinely different. And like we have genuinely grown up and we are in the pharma era. So like you kind of, you kind of got to respect that. Like, yeah, you can't pretend to be the guy, you know, putting dandruff thing on his, on his own head and just being like, oh, what do you know? Look, head and shoulders is the best dandruff shampoo. It's like, okay, well, you got to, you know, you got to know how people are going to perceive it. So I can see both sides. I can see both sides. That's my.

A (24:13)

I clearly did not know that. And I was really, I will say I was mainly inspired by like, I said this Don Wilson tweet because, like, he's been around a lot, right? He lived through March 12, 2020, which in TradFi had a number of really crazy events, including one where like these people who are supposed to post collateral on your behalf, these FCMs, a bunch of them went bankrupt and trading firms who were profitable went bankrupt because the person who was supposed to post collateral on their behalf for this kind of lost socialization didn't. And then the exchange closed everything. Vlad and I both know some people who are indirectly hit by this kind of blow up. And I think it shows the weaknesses in the tradfi system too, right? So Don has seen that and I think he has probably the best perspective of anyone in crypto because he's been on both. He's kind of been in crypto for a while. Cumberland's been trading here forever, right? So he's kind of seen this grow up. So I just like had a lot of like. His statement basically felt like this almost existential thing, right? Which is like, either you fix all this shit or we're just going to go back to fcms. Which is like kind of a sad to me. That's like, if crypto ends up just with fcms, like these like clearing Agents like that's sort of just like a sad ending for derivatives. Like the point for bearer assets, self custodial stuff is like you don't need that, right? You shouldn't need that. And to me that's like fundamentally the main thing I was kind of cared about and why I kind of went down this rabbit hole.

B (25:47)

Vlad, what's your take? FCMS Is that FCM's bearish?

C (25:51)

Yeah, I mean I think there's definitely. It's an interesting problem, right. It's a hard problem. I mean I think there's different solution spaces here. Right. Like, I think for example like options. Like maybe like one. One thing I've always wondered about is like why, why isn't there more emphasis on options and decentralized markets? Like, like I think options would actually make some of this stuff cleaner in the sense that you could take leverage in ways that aren't going to bankrupt the whole system. So I, I mean I think that may be one area that there'll be more of a shift towards in the coming year. But, but also yeah, I think having. I do think this kind of research is a public good. Right. I mean I think understanding this further, like certainly like our team when we looked at the research like we learned a lot and you know, it'll influence how we think about like future versions of you know, as we iterate on the AD algorithm. So yeah, I mean I think and yeah it's like people pass positions on different exchanges even within a single exchange you may have positions of correlated assets or like spot versus futures. Like, like I think that it's, you know, these are hard problems and you know we'll I think having, I think Haseeb to your point about like adl that this shouldn't happen very often in the first place. Right. I mean I think that's part, part of what we're trying to get to as well.

B (27:31)

So today, speaking of should not happen today we had another day of leverage wipeout. It was something like $400 million that was wiped out today in terms of leverage. But it does seem like in terms of the onchain volumes on chain volumes are hitting a fever peak. A lot of this is driven by anticipation of LightR's TGE which I believe is coming soon. And again, fair notice both the Dragonfly and Robot are investors into lighter right now you guys are the number one exchange by 30 day per volume as well as seven day per volume. There are some markets right now in Polymarket that are anticipating when the airdrop is going to take place supposedly at some point by the end of the year, but who knows? Polymarket obviously seems to be fairly confident in that. And there's been some news, I guess recently about some new. Well, there's been some conversation about the fee model of LiDR, which seems to come up again and again. So just to reiterate the way that liDr works. So LiDr, unlike hyperliquid, in LiDr, there are two ways that you can decide to trade on hyperliquid. One of them is with zero fees. But if you're paying zero fees, you have increased latency. So there's like a quote unquote speed bump that you're paying if you're not paying, if you're incurring, if you are a normal trader, if you want to pay for lower latency and API access, there's like a pro. I don't know what it's called, but like a. Yeah, the premium tier. Premium tier. There's a premium tier that you pay for that does charge fees if you want to be able to trade faster. And this is where more pro traders tend to trade. And so with this structure, so lighter has become quite profitable. It's making significant revenues as an exchange, but it's doing so despite the fact that for a majority of retail traders, they're actually on there paying zero fees. And this is, I think, part of the reason why it's attracting quite a bit of volume in addition to the anticipation of the airdrop. So talk to me, Vlad, about. I've seen a lot of people going back and forth about how they feel about this pricing model. Do you think this is where crypto trading is going? There's been a lot of comparisons, like. Well, you know, Robinhood kind of did the same thing where they introduced zero commission trading and they kind of took over the old school brokerages which all charge commissions. It's a little bit different. It's not exactly the same thing, but it kind of rhymes. Talk to me about how you see the pricing of fees in crypto perps.

C (29:52)

Right? Yeah, I mean, I think, you know, I was thinking of something Tarun said earlier, which is like for in the ADL context, it's like, I guess Huobi Trident and other exchanges copied it. I feel like the same. That's kind of how I feel about the fee models that have existed for, I don't know, the first 10 years of perps trading where they've all been kind of pretty much the same. There hasn't been much innovation there. Even though in TradFi as you said with Robinhood and others that follow, there has been a lot of innovation there. I think for us. I think there are three reasons for this fee structure. One is philosophical kind of democratizing finance. I mean that's part of why we started building LiDAR in the first place. Two, when you think about two sided markets, right, if you have kind of retail and you have professional trading shops, like you know the professional trading shops, they'll, they'll adjust. Like the harder part of the marketplace to attract is the retail. I mean retail has a lot of options. Like they, there are a lot of products they, they can try. Like the, whereas like the professional trading shops they're active on many exchanges. Like if they're making, if they're making 100k a day, trading on later and then now they have to pay premium fees. Now they're making 80k. Like they'll, they're, they're still going to stay right like that. That's still, they're running a business. As long as it's profitable, they're going to keep doing it. So that's the kind of second reason, right. I think third reason is it actually I think is like better for the market in the long run because it expands the universe of traders and possible trading strategies. When you have the zero fee tier, like maybe there's some traders who wouldn't be in the market at all but now they're in the market and they're doing like scalping or like, or maybe someone who's retail is now maybe they would have been trading once a day. Now they're trading three times a day, right. So it's like if you expand the opportunity set, I think that's better for the market in the long run. But, but how exactly that will play out? Like I think we are seeing some of the other builders in the space experiment with V models like ours or other variations of that. But how will that play out? Is it going to take like even with Robinhood, right. I think it took five years before other major players started to try their few models. So like I don't know if it's something that will see like a very immediate shift from the crypto moves much.

B (32:42)

Faster than any other industry I've seen.

C (32:44)

It's true, it's true.

B (32:47)

The time it took from Bitmex to Binance or from old school Bitmex contracts to Binance completely changing the game was like a year and a half, right?

C (32:59)

But I guess it would be kind of presumptuous to say oh well, in a year every this will be the state of play for the whole industry. But we certainly believe in it. Right. And we're going to continue. It's nice to see because I think like six months ago when we talked about this, it was kind of theoretical and it was like, oh, this will never work. Right. And it's similar to when I remember in the early days of Robinhood where I was an advisor, like same kind of feedback. So it's nice to see that it's actually at least so far working. Right. But now that we've added Spot, right, that's also interesting right now some of the stuff we're going to be working on, that we are working on that we're going to be releasing soon, right. Is like you'll be able to do. I mean you can already trade Spot for free now, right. But with native bridging to and from Ethereum, it'll open up some interesting possibilities there too. And I think part of like, you know why? Because right now we only have one asset on spot, which is eth. I mean that's going to change soon. But even with that asset like there's been some pretty healthy volumes. And yeah, I think the fee structure has a lot to do with that.

B (34:12)

So beyond that, let's also talk the product roadmap because you know, one of the things. So right now Hyper Liquid de facto leader, you know, although right now both you guys asteroid have been leading in volumes in terms of open interest, they still remain the king. They recently launched Trade XYZ which is their RWA trading platform. You can trade by NASDAQ perps, you can buy Tesla perps, etc. Doing pretty, pretty significant open interest at this point. My understanding is that you guys also have your own offering on that front as well as Hyper Liquid just announced cross margin allowing you to margin the same account with not just usdc but also I believe they just added hype as the other margin asset and they plan to add more over time is my understanding. And you guys also have cross margin on your roadmap. Give me a sense from the inside baseball, because now it used to be that basically Hyper Liquid, totally dominant, left everybody else in the dust like them versus DYDX or them versus Synthetix or whatever. It was really like BitMEX versus everybody else. Now it's a horse race. Now it really does feel like there are a lot of credible players in the same market and there's a lot of product competition in terms of features and trying to keep parity with each other and stay at the vanguard. Talk to me about what that's like from your guys perspective as a team.

C (35:32)

Yeah, I mean to be honest it's been a lot of fun right to see new features and I think part of it is coming from Hyperlucid but also other competitors like for example Ostium I think was maybe the first to really lean in on real world assets and now we've added fx. I think it's been, it feels like longer, I think it's been a month. Right. And actually like the open interest there has been quite especially on, you know, the Euro, like this has actually been pretty, pretty high. But, but I guess to the your bigger point I think like for us like building on top of Ethereum is really important to how we think about the roadmap. Right. And so really like the vision is that any asset on Ethereum can be collateral on lighter and vice versa. A lighter position can be actually kind of tokenized and live on Ethereum as well. Right. So for example, like that's not possible today. That's. I mean I think that the tech for that is almost, is almost there. I mean the one remaining piece is the ZKVM sidecar which we've been working on that for some time actually we're going to have some announcements around that in the coming weeks. Some of those ideas have actually been inspired by some of what TARUN has written about six months ago and other folks we've been talking to in the Ethereum ecosystem. But essentially what that'll allow us to do is like right now we have these custom circuits that can do order management, risk management, liquidity management very efficiently but you can't. Like right now you can just write a smart contract and have that run on LiDAR. Like you have to implement everything as circuits but with the sidecar you would be able to do that. And so then those smart contracts will share the sequencer with kind of the lighter exchange itself. And so that'll open up a lot of things like universal cross margin, like tokenization of lighter positions and so on. I think some of the.

B (37:44)

So should I imagine that is like that's a substrate on which LiDAR is going to be shipping new features that allow you to compose things with your core LiDAR positions, et cetera. Or this is like a developer platform a la Hyper EVM that's meant to attract a broader set of developers. How should I think about the lighter vm?

C (38:03)

Yeah, I mean some of it will be, there's some features that our team will implement that we have thought about for a while already. Like for example tokenizing lighter positions or even tokenizing llp. Right. That can live. You can actually just buy shares of LLP on Ethereum. So things like that will probably build directly. But yeah, other developers like Layer 0, Morpho, Aave, all these protocols can live on top of lidar once the sidecar is done and we've talked to some of these teams already about doing that. So for example, in implementing something like basis trading, right. That's not necessarily, you know, straight up.

B (38:54)

Tokenize a basis trade within later.

C (38:57)

So that's, that's, you know to make that work though you need some form of lending as well. Right. And we probably like, you know, they're already very established lending, you know, protocols on top of Ethereum. It's like kind of makes a lot more sense from our perspective for those to to live on lighter as opposed to like us building, you know, lighter land or whatever.

B (39:17)

Right, that makes sense.

A (39:18)

Yeah. One thing I actually think is interesting about these RWA perps. So Vlad mentioned the Euro is quite liquid on hyperliquid or sorry on later is there's significantly more volume in the FX perps than the history of the euro stablecoins. If you go look, if you compare. And it sort of makes me wonder if people are just going to trade perps of FX and then only use USD so they want exposure to euro, they'll just do a basis trade of holding spot stables and long or short the per. Because it's actually some of these rw like the RWAs are funny because the ones that everyone thinks will be the highest volume seem to never be right. Like Tesla, everyone. I feel like Tesla has been this meme in crypto for like the last five years. Everyone who's like stocks on the blockchain would just be like they're Korean people who want to buy Tesla. Like I feel like I've heard that pitch 5 million times since 2020. I don't know if you haseeb you've gotten that also.

B (40:17)

Yes, the kajillion times.

A (40:19)

But then you look at the actual volumes on hyper liquid and lighter, it's like Nvidia is crushing it. Or the index. I didn't think people wanted to trade the index on chain. Right. It's sort of like. So I really enjoy watching the RWA perps because the leaderboards show me things I never would have expected to see on chain.

C (40:38)

No, I mean I think it'll be interesting once you have for some of these assets. Like once you have the perps and the spot kind of on one platform, you know, to your point about kind of FX basis trade, but also even with, you know, we're talking to the Robinhood team about kind of tokenized stocks. Right. I mean, there's some interesting stuff there as well, right, with, you know, spot versus perps, but also for, for imagine like for fixed income, right. Like, we haven't. That's not an area that has been explored that much yet on chain. But like you can do some like, imagine that you can have like a yield curve on chain and at the same time like trading the, the perps too. Right. Like, so, like, I think there's a lot of interesting stuff and yeah, I, I don't. It's hard to say right now like which ones will get traction, which ones won't. But I think once you have the spot and the perps, that could change because at least in tradfile, a lot of trading shops, their whole strategy only makes sense in the context of the basis trade.

B (41:45)

So I totally hear you guys on that. At the same time, I also don't want to be Pollyanna ish, which is that whether it's the RWAS that are being traded on trade XYZ or whether it's the forex that's happening right now and lighter right now, all this stuff is pre tge, right? Trade XYZ hasn't launched a token, obviously. LIDAR hasn't launched a token. And we've all been in crypto long enough to know that there's some distortion, there's some gravitational force that's exerted by token coming. And so one of the questions I have for you, Vlad, is once token comes, token's coming, you can check the polymarkets to decide what date you think the token's coming. But token's coming once token comes. One thing that we have seen from every token in the history of anything is that the behavior TGE T minus 1 versus T plus 1 changes. What are you expecting, what are you preparing for? And how are you thinking about the before and after for lighter has a token?

C (42:42)

Yeah, I mean, we're certainly going to continue to build and roll out new features and we'll. I think there'll be a lot of interesting functionality that native token will unlock. Right. But yeah, I mean, just from like an expectation setting standpoint, it's. The old cliche is like it's a marathon, not a sprint. I mean, I think the team and the early community is not necessarily expecting something that will be like A rocket ship on day one. But we'll kind of start in a healthy place and then kind of trend upwards from there if we continue building products that the community likes. I guess the goal obviously isn't just to maximize value on day one, but to have something that's healthy relative to the current state of the market and continue iterating from there because it's not, I think we've seen a lot of, we've seen some tokens that have this spike and then don't do well and then others that continue kind of a steady grind. Right. Which is, I mean I think we've had a very strong growth cycle like the last few months. But at the end of the day we're in this for the long haul, right? We're here, there's a lot of technology to build here, there's a lot of innovation. And yeah, I think sometimes people might get excited about certain parts of the product for the right reason or the wrong reason. But some of it may have to do with like we're trying to, for example, some of what we're doing now, right, is if we see egregious examples of points farming which are basically like civil attacks. Like we, that's not, I've seen some.

B (44:39)

Chatter on the timeline of you guys just slashing people who are just flaggingly.

C (44:43)

Yeah, exactly.

B (44:43)

Behavior. Yeah, yeah, exactly.

C (44:45)

So there's, there's stuff like that that happens. Right, but, and then you know, there's anything from that to like real traders who may be trading a bit more aggressively than they would otherwise like. But, but then there, there is, I think across the space there is a lot of real innovation happening and there will be for the years to come and we, you know, we want to be part of it.

B (45:05)

It's clear this is a big market. One, one last question on, on this front as well is like, I think the last time we spoke must have been in the summer. You told me that your North Star metric is not volume, it's not open interest, it's tvl. Is that still true?

C (45:21)

Yeah, I, I, that is still true and actually I think that the events of October 10th were pretty good validator of that because there was a lot of noise around open interest and volume across a few platforms. Whereas I think TVL was actually pretty steady because I think at the end of the day TVL is a metric that represents capital that customers real opportunity.

B (45:55)

Cost.

C (45:57)

Trust us with.

B (45:58)

It's funny because we were talking about this with the L1 debate and we've over the years invented so many new metrics for layer 1s. Like, oh, there's rev and there's the Metcalf law and there's all this other stuff. The first metric we ever had for layer 1s was TVL. And it has turned out, I think in retrospect, to actually be the most robust metric that in the short term you can game it, but in the long run it's really hard to have billions of dollars in tvl. It's just you can't will.

A (46:27)

Also, in perpetual exchanges, the TVL is mainly stable. So it's much. It's even. It has less of the cycling.

B (46:35)

Well, not anymore. Not anymore. Not with cross margining coming to a perfect change near you.

A (46:39)

Sorry. Yeah, I guess we should add that.

C (46:42)

Yeah, I guess even. Right. Like we. Now that we have ETH, I think like 2% of our TVL is now in ETH. But. But yeah, I mean it's interesting because when I first got into the space, because TVL isn't really a thing in TradFi, right? So like when I first got into space I was like this, like to your point about new metrics, like, like this is kind of a new metric for me and it took some time getting used to it, but now I actually really see the value and how robust it is.

B (47:11)

Yeah, it is very much that like 70 IQ, 140 IQ meme I think is TDL. I feel like it's aged very well. Okay, so one more kind of meta story that I wanted to get into that relates somewhat to this discussion on tokens is there's been a lot of drama lately about the dichotomy between tokens and equity and what happens particularly in the case of M and A. So there was a story earlier today that Axelar got acquired by Circle. And when I say Axelar got acquired by Circle, of course, Axelar is a decentralized bridging network protocol thing, you know, cross chain message passing. And it was not the token that got acquired, but rather the dev go that got acquired by Circle. It was basically structured as an Aqua Hire kind of thing where Circle took this thing and basically like, yeah, token, you know, go have fun. The token immediately went down something like 15%. I should note. We're investors in Axelar. Turun, are you guys investors in Axelar?

A (48:07)

Yeah. You are?

B (48:07)

Okay. Yeah. So we're all investors in Axelar. So we all had tokens go down. I don't know if you guys already sold, but we certainly did not. So we had a fun little day today. And then of Course, there was another story a few weeks ago about Tensor, which of course was one of the large. They created Vector Fun as well as, you know, they were originally a big NFT exchange on Solana and they got acquired by Coinbase. And interestingly when that happened, the token, although like same story, acquired the Devco, did not acquire the token itself. In that case, the token went up, whereas when Axel got acquired by Circle, the token went down. So maybe like Circle, Circle is not as bullish as Coinbase as an acquirer. I'm not really sure what that means. And then separately to all this, there was a third drama around aave and basically aave. So of course AAVE has the protocol. Aave, as everybody knows, has a big token, it's a big lending protocol. There's also AAVE Labs or Avara, which is the company like the Devco that actually employs all the people and owns the ip. And Avara was actually. I forget what the specific story was, but they were basically collecting fees off of something that was happening via the front end. Do you remember what this was?

A (49:22)

Cal Swap integration?

B (49:23)

Cowswap. Cowswap. So basically one of the, one of the functionalities on the front end is that you can swap and basically originally the fees were going to the protocol, but now those fees are getting swept into Avara, the company, much akin to what was happening in the early days with Uniswap. Uniswap Labs versus Uniswap, the protocol. So all these three stories, kind of loosely related, but they all triggered a broader conversation about what should be the right relationship between tokens and equity. We talked about this at some length actually, when Uniswap did their unification, where they united both the DAO as well as Uniswap Labs and stopped charging a separate fee on laps versus versus the dao. But the weird case here, I think with aave, it's actually relatively straightforward. Maybe, I don't know, we could talk about that. But I think in the case of M and A of Devcos, that does feel like also actually a pretty different situation than a sort of, you know, a Labs double dipping with respect to getting paid something that maybe should be owed by the token holders. So that's the backdrop for this conversation. We've had a lot of people going back and forth about it. Tarun, what's your take on this whole drama between tokens versus Equity, M and A? How does it all fit together?

A (50:36)

The M and A thing actually reminds me a lot of what's happening in AI, where someone will buy the Management team and then dump the rest of the company. Like what happened to Character AI or Adept or any of these companies?

B (50:51)

No, but in those they did pay a big fee, licensing fee or whatever. Yeah, yeah. So the shareholders still got a payout.

A (51:00)

Yeah, I guess that's true. Although. Yeah, it's true. I don't think the payout is uniform.

B (51:07)

I don't know if it's uniform, but like, if you literally just like acquired the people and pay them a bunch of money, there would be shareholder lawsuits. The fact that it's not M and A, that was mostly a workaround for like Lina Khan and all this M and A kind of scrutiny, but investors were still getting paid. So I think that's a different story.

A (51:23)

Actually. That, that's, that's a good point. There's no, there's no like token holder lawsuit rights, which I, I think it's true. Maybe that's like the real, that's like the real problem. Like.

B (51:34)

Yeah, yeah.

A (51:36)

Because like, actually I kind of, I, when I saw this, I was just thinking about this AI thing because I was like, oh, it seems like hiring the dev co is a little bit like hiring the manager, like some subset of the company, but because like the foundation might exist and have employees. Right. Which is like the rest of the ecosystem. I, I think the tensor one seem, I don't, I don't know, it seemed fine because like, isn't the product kind of like this kind of sunsetting or kind of the trailing off? Whereas like, Axel, I feel like it's a little bit weirder because in theory the network has a lot more assets secured by it. You know, like there's like aum. I don't, I don't know what happens. What are the rights to that if it turns out like the value?

B (52:18)

So they handed off the, the.

A (52:20)

Oh yeah, yeah. They have a foundation, right?

B (52:22)

Yeah, there's a foundation that's still, that's still capitalized. And that foundation is now being shepherded by another organization. What's it called?

A (52:31)

Common prefix. Right, common prefix.

B (52:33)

Common prefix. Yeah, yeah. You know, common prefix.

A (52:35)

Yeah.

B (52:35)

Do you know them? Okay, yeah, yeah. So common prefix is now taking over. It's a dev shop that's worked on multiple different chains. And so maybe this is now their business is like taking over zombie chains and just running them.

A (52:46)

Well, they were trying to launch this like non chain chain, right? POD or whatever.

B (52:51)

Oh, is that right?

A (52:52)

Or pod. Yeah, but, but, but common prefix is actually a bunch of like, you know, academics who work in consensus quite deeply. I think their Name is one of the best names in crypto, but it is a very nerdy name.

B (53:03)

It looked like a very impressive team, but it did. It was like this like, oh, it all clicked moment where I'm like, okay, there are like, you know, one of the things people complain about with FTX and the FTX bankruptcy is like all these like bankruptcy vultures who just like show up and charge insane fees. And I'm like, why wouldn't there be something like that for crypto? Is that like, you know, okay, we say the chains can't go bankrupt, but like, the people certainly are like, yeah, fuck, you know, fuck this, I'm out. And in which case, shouldn't there be somebody who's just like, yeah, I'm going to charge you an arm and a leg, I'm going to take over your stupid foundation, I'm going to pay myself all the money in there and I'm going to manage your stupid chain. And I'm like one of the only people who can do it. Not to say that actual stupid, but.

A (53:45)

I prefer the perspective of this because devs beat the lawyers.

B (53:49)

I'm sure it's a better system. I'm sure it's a better system, but to be clear, I don't know, Common prefix. I just learned of them the first time and I'm going to chat with them soon. But it would make sense that that exists. There ought to be something like that. Just given the way that crypto works. If it exists for companies, why shouldn't it exist for protocols? Because the normal thing is that it just goes into disrepair, right? It just like, it's basically in the junkyard of old dyno chains that nobody even knows how to start up anymore. There's just some validator in some cloud somewhere that's still running and if anyone needs to fix it, it's like, oh, fuck that. That one's just going to get killed by a quantum computer. No one's ever going to go bother to go update that thing.

A (54:27)

You know, like, look, I think you're right, there is something to this. Like vulture crypto has not had many vulture M and A advocate type investors and I know a bunch of people who are starting funds right now for doing this type of stuff of like token holder advocacy and maybe, maybe. No, no, but, but I think there's a version of the world where you're talking about the, the, the people who made money off the FTX estate. Imagine a fund that's doing token holder advocacy. But I'm actually just trying to Buy the. The tried to get my company whether my dev shop to be the one that's. So I use my fund to buy.

B (55:05)

Okay. So now we're introducing conflicts of interest and like self dealing into the whole. Okay, great, perfect. We created all of traditional bankruptcy.

A (55:14)

I think that's what's going to happen. That seems like.

B (55:16)

I can see, I can see it. Okay, so I do have, I do have a theory on this. But Vlad, I want to, I want to bring you in. What's your perspective on this whole shareholder token holder drama?

C (55:25)

Yeah, I mean I haven't been following the specific cases. I guess we've been pretty heads down as you guys can imagine. But I guess one thing I'll say is I feel like, I mean again, I don't know the motivations for these specific deals but like I feel like there's a lot of untangling going on now from like, like, you know, I think the last kind of kind of under the, the, the regulatory environment in the past there was a lot of kind of like, I think what someone called like centralization theatrics. Right. Of setting things up for. And not necessarily, not like running a chain in the way that's just most efficient for its users, but just having all these different entities. So I don't know but my guess is that maybe some of this stuff is meant to untangle some of those structures. But yeah, I mean I think from like from our perspective, you know, it is like I think we do want kind of one, you know, everyone who's involved, whether you're a team member or an investor or you know, an early community member, like you're kind of all in the same boat and have allocation from the same pie. I think having when incentives. And to me like that's part of like the magical things about crypto is like that the opportunity to do incentive alignment when once you lose that then yeah, I mean you get into, you might get into some unhealthy outcomes. But. But yeah, I mean I, I have totally agree. Zombie chains that, that part. That's maybe a separate point here. I, that I'm curious what might happen there.

B (57:10)

But anyway, yeah, no, I completely agree and this is the, the, the drum that I've been banging for a while is that you should have one asset and you should not have some smearing of value across two different assets, two different sets of investors, et cetera. This was the criticism that I levied against Uniswap for a long, long time on the show as well as in other venues and I think this is largely what's being echoed by a lot of people who are criticizing these deals. Now. We're investors in Axelr. And so actually, I want to clarify that despite the little story I just gave about, oh, maybe there are these vulture firms, I actually think what happened here is totally fine. So I think, like, Mike Dudas as well as a few other people were like, oh, my God, this is terrible. You know, Circle's a bad actor. They're like cajoling people into taking these horrible deals. Look, you know, Axelar, I think it's a great team. Their Token's down like 85% or something from where we invested. Like, normally when a startup is down 85%, like, it's time to call it quits. It's like, basically, hey, didn't work. We didn't build the thing, we didn't succeed. And you go, you're not enslaved by your VCs or by the people who've brought you together. And to say, well, we have to just keep running this forever until we hit all time high again. At a certain point, you just say, hey, it's time to cut our losses because we're talented people and there are more useful things that we could be doing with our lives than just slaving away in the startup mines. And as an investor, it's kind of your job to be mindful of the human capital that is actually embodied in a startup and that is best released when the startup is not working. Obviously, you try at it for a while and you might have to pivot. There are times when you have to go through the dark night of the soul in order to get to the promised land. It happens. But there are times when it's like, hey, this market didn't materialize. We thought interoperability was going to be huge. It's not huge. You guys build Axelar. It's great. Market doesn't think Axel are that valuable. That's okay. It's totally okay for that to happen now if you're a retail investor. Retail investors don't think that way.

A (59:11)

Right.

B (59:11)

Retail investors are like, how dare you. You created a token, you lost me money. You better, you know, like, keep, keep working on this until I'm finally back in the money. Otherwise, I'll hound you till the end of your days. And that's part of the reason why people historically take venture capital instead of retail capital, because they know that venture capital is okay with the idea that, hey, you gave it your best shot and didn't work. It's okay. You can move on. So when Axelar got acquired by Circle. So I should also clarify, I don't know all the details of exactly how this was structured, this and that. I know the high level of what we're getting in this outcome, but from my perspective, it's fine. I think the Axelar guys are talented. They should go work on something important. Circle's building important stuff. And if the market thinks that Axel are not that valuable and basically there's not that much more to do on Axelar besides just keep it running, then, you know, so be it. Good game, fair play. You guys put multiple years of your lives into this. That's all that an investor can ask for. So from my perspective, I don't think. And if you're Circle, you are under no obligation. I mean this is this really Mike Dudis points. I'm kind of subtweeting Mike Dudis right now. You are under no obligation to pay token holders for anything. They're not buying a token, they're acqui. Hiring a team. This is basically an acqui hire. So if you're aqua hiring a Devco and the Devco makes no money, the Devco is just like a corporate shell. So if you're acquiring a Devco, you don't owe anybody anything. You don't owe token holders something. You don't owe the investors a big outcome. You don't owe anything. If the people who own the equity in that thing, which if it's all safe, then the only equity holders are the founders. If they agree to the deal, it's a deal and token holders don't get a say in that vote. Now, I do think that there is a good argument to be made that that is perhaps a flaw in the system. And to your point to Rune of like, hey, why don't token holder lawsuits is basically, why would a token holder lawsuit make sense? The answer is because maybe token holders should have a claim on governance of the Devco. And I imagine that maybe there's a way that you can build that in that perhaps there's like a protocol level rofer in the Devco docs that if somebody goes in and tries to do a change of control of the Devco, that the docs are written such that the Devco has to go and run a competitive process with the protocol and say like, hey, these people want to acquire us, but as a protocol, as a dao, you have the right of first refusal. You can buy us back. If like somebody is convinced there's a better use for this Devco talent than the protocol. But my guess is that would the Axelar protocol pay up for the Axelar team? One thing that we've learned from these daos is that they hate paying for talent. I think probably one of the worst parts of DAO governance that we've seen is that they're just so cheap. DAOs are so incredibly cheap. They hate paying above market salaries or even market salaries. They abhor market salaries. And every single individual line item is overwhelmingly picked apart and criticized. Where if you look at a corporate board or you look at VC governance, we don't do that. That's stupid. It's a really terrible way to retain talent and to incentivize a team. And so I don't know. I'm curious, Tarun, what's your take on this concept space of letting daos play a part in the dev go retention?

A (62:30)

It seems inevitable it'll happen. I mean, I feel like you've just created five jobs for players who are listening to the show who are going to like open class action lawsuits tomorrow or something.

B (62:42)

Well, you can't. I mean, you can't. No, I don't think. I don't think you could. I don't think you have standing to sue. If you were trying. I think that would be. It'd be such a stretch.

A (62:50)

You can make life horrible for the dev code.

B (62:53)

You can be annoying. Yeah. But like, you have to pay lawyers. I don't think anyone's gonna take that on contingency. That's such a stupid lawsuit.

A (62:59)

Having been part of.

B (63:00)

I don't know, maybe I'm wrong. Eat my words. Maybe we have accepted this idea now into people's heads. Berwick Law is going to wake up and suddenly like to have laser eyes and be like, boom. If you own intel, let us know.

A (63:12)

Haseeb, you and I both are party or. Well, I guess I was party in the lido thing, which. Yeah, what. What? I don't. I like. I think there's lots of lawsuits that exist in the world that might be frivolous, but take seven years to adjudicate.

B (63:28)

Totally, totally.

A (63:29)

And.

B (63:31)

Yeah, but also it was a different regulatory regime at that time, so there was more sympathetic. There was more sympathetic judges. There was, you know, the like. If, if you went through motion to dismiss today on a lot of those lawsuits, I think they would have succeeded today, whereas they failed at the time that they were actually initially adjudicated. So.

A (63:50)

Yeah.

B (63:50)

Anyway, we're up on time. Vlad, thank you for joining us. What should people look out for if they want to see what's happening with Lighter TGE coming up soon. Give us some. Give us some alpha.

C (64:02)

Just, you know, follow us. Lighter xyz, you know, we'll. We'll keep everybody up to date on the next two weeks. There'll be some. A lot of announcements.

B (64:15)

Okay. And lighter points prediction for lighter points for their value. Yeah, Yeah.

C (64:24)

I mean, I think the market, it's starting to trend up a little bit again, so we'll see if it hits 100 pretty soon here.

B (64:32)

Okay. Is that the prophecy? 100?

C (64:36)

I think we'll see some trades at 100, but, yeah, I guess those points markets are. Although that's been kind of an interesting thing to watch throughout this process, but yeah, we're pretty much just. Just staying heads down and there's. There's so much to do. I mean, so, yeah, definitely. Thanks for having me, guys, and thanks for coming.

B (65:01)

Well, next. Next Tarun paper is going to be on points markets, so.

A (65:05)

Dude, I think after. After. After this time, I think it's the first time I've ever felt like I needed.

B (65:13)

Have they beaten it out of you? Have the HL boys beaten it out of you?

A (65:16)

I just think it's not just that. It's just like, no. Like, I feel like there's no value for this type of thing. It's like, you might as well go.

B (65:24)

Through the next show. Like, no, you can't just be like, this is the beginning of your expert witness arc.

A (65:31)

It's just kind of like. All right, if you're just going to be kind of.

B (65:35)

All right, all right. We've got to end the show on something positive. Tell me something positive.

A (65:40)

I don't know if I have anything.

B (65:41)

For you, so you force it. Force it. Anything.

A (65:47)

$100 million in LLP.

B (65:50)

There you go. There you go. His Excellency. All right, good note.

A (65:55)

All right.

B (65:56)

All right, thanks, guys. We'll be back in a couple weeks.