← Unchained
Unchained
The Chopping Block: Inside the $19B+ Perp Crash, ADL Explained, Binance’s USDe/Staked-Token Depeg, and the Hyperliquid Whale Debate
00:58:17
$19B+ in liquidations, ADL chaos, and Binance’s USDe/staked pegs snapping while APIs buckled and gas hit $400—we dissect the Hyperliquid whale, delta-neutral blowups, HLP vs. LLP incentives, and ask: should perps protect vaults or winning traders?
Loading summary
Transcript206 lines
- [00:00]
Haseeb Qureshi
How did you experience it?
- [00:01]
Tarun Chitra
I guess I had a different experience. I actually think it's funny to think of like, VCs as like the FEMA showing up after the hurricane type of thing. That's like kind of like podcasts and.
- [00:12]
Robert Leshner
Tweet threads about it.
- [00:13]
Haseeb Qureshi
I did put on my little red cross. And yeah, not a dividend.
- [00:17]
Tarun Chitra
It's a tale of two point.
- [00:19]
Haseeb Qureshi
Now your losses are on someone else's balance sheet.
- [00:21]
Tom Schmidt
Generally speaking, airdrops are kind of pointless anyways.
- [00:24]
Haseeb Qureshi
I named trading firms who were very involved.
- [00:27]
Tarun Chitra
Alec Eth is the ultimate puzzle.
- [00:29]
Robert Leshner
Defi protocols are the antidote to this problem.
- [00:33]
Haseeb Qureshi
Hello, everybody. Welcome to Chopping Block. Every couple weeks, the four of us get together and give the industry insiders perspective on the crypto topics of the day. So, quick intros. First we got Tom the Defi maven and master of memes.
- [00:42]
Tom Schmidt
Hello, everyone.
- [00:44]
Haseeb Qureshi
Next we got Robert the crypto connoisseur and czar of Superstate.
- [00:47]
Robert Leshner
Good morning.
- [00:49]
Haseeb Qureshi
Next we've got Tarun, the Giga Brain and Grand Poobah at Gauntlet.
- [00:52]
Tarun Chitra
Yo.
- [00:54]
Haseeb Qureshi
Joining us today we've got special guest Doug Colkitt, liquidity luminary at Ambient Finance and Fogo.
- [01:00]
Doug Colkitt
Hello.
- [01:02]
Haseeb Qureshi
And I'm Haseeb, the head hype man at Dragonfly. We are early stage investors in crypto, but I want to caveat that nothing we say here is investment advice, legal advice, or even life advice. Please see Chopping Blocks at XYZ for more disclosures. So we were actually planning to take this week off, but unfortunately we were brought back on the job on overtime because we have just experienced the most wild weekend potentially in crypto history. The single largest day of liquidations ever. It was an absolutely chaotic weekend and it all really happened on Friday. So Friday, for those of you who are somehow not paying attention to what's going on in the world, I wasn't paying attention. Trump. Trump issued massive threats against China for withholding rare earths from the US and basically said, we're coming back at you. New round of tariffs, and this time they're bigger, they're badder than ever. He elicited this threat late on Friday evening, US Time, and this caused markets to immediately crash. And crypto markets crashed much more severely than almost anything else because, of course, crypto could continue trading even as markets closed on on Friday evening. So we saw the single largest day in liquidations ever. Over $20 billion liquidated. Over 1.6 million crypto traders were liquidated over that time. Mostly longs liquidated. Much more so than Shorts. Now these losses you saw across the entire crypto market, caps so large caps fell about 27% on average. Smaller caps lost 52% on average. During the course of the day peak to trough, we saw some tokens completely cratered to basically near zero. Atom very notably hit less than a cent on Binance over one Crazy Wick. Multiple assets went down over 80% and that did later recover in the day. But insane levels of volatility that we have almost never seen, rivaling only the COVID crash and the FTX crash. Of course, Covid is much, or sorry, crypto is much bigger now than it ever was during the COVID or FTX crises, making the dollar notional amount that were liquidated in this, in the sell off the biggest that we have ever seen unequivocally in history. We saw gases, gas spikes all time high on Ethereum, more than two times previous high. There were people showing screenshots of them paying $4 gas thousand dollars gas to get transactions into Mainnet. And according to some people, there were potentially even more liquidations that were reported because the single largest venue for liquidations was Hyper Liquid. Hyper liquid showed over $10 billion liquidated on hyper Liquid alone. But of course Hyper Liquid is nowhere near the largest venue, implying that probably many of these other venues are underreporting their liquidations. And so probably the real number is much, much larger than what we saw. At the same time we also saw most of the centralized exchanges have major downtime. So Binance famously was down for like roughly an hour, not fully down, but had massive instability in their APIs. People were not able to get inventory in and out of Binance. It's a large part of the reason why so much of the assets that that massively drew down were only on Binance as huge spreads that we saw between exchanges. Market makers just completely unable to quote. Over 25 million, sorry, 25 minute blackout, huge API throttles. And Binance ended up refunding over $250 million of people who are incorrectly liquidated over the period of that day. They came out and said mea culpa, we messed up, we're going to make it right and try to refund some of the people who lost money. But overall it was a massive day of wallet instability. Multiple APIs, RPCs that were failing. It was a crazy, crazy, crazy day. I'll stop there. People get the overall picture. Love to get color from you guys. How did you experience Friday and the ensuing aftermath going into the rest of the weekend?
- [04:41]
Robert Leshner
I'll Go first because I feel like this is the easiest answer and it's a bit of a cop out and I'm going to be like furniture in this episode. I wasn't really trading. I didn't have any positions on. I saw a bunch of telegram groups. I was in blowing up and people were freaking out and I was like, yeah, I'll take a look at this tomorrow. So I missed the whole thing. I didn't sweat it. I'm not a trader.
- [05:02]
Haseeb Qureshi
If you were holding spot bitcoin, you don't even notice anything happened because I think bitcoin went from like 120 to like 108, I think, and then ended at 115.
- [05:13]
Robert Leshner
Yeah, like I, I pulled up coin gecko, I saw a lot of red stuff. I was like, you know, I'm not using any leverage, I'm not trading any perfs. Like, I'll see what all the commotion's about later. So I missed the whole.
- [05:26]
Haseeb Qureshi
Good move. Good move. Tom, what was your experience?
- [05:30]
Tom Schmidt
Yeah, I mean, in some ways also as a vc, you're also somewhat of like a non combatant. You know, it's like we're pitting portfolio companies and making sure they're okay and kind of assessing the damage. But I also don't have a big position on one of these perp exchanges, so I'm not as much sweating it personally. I do also maybe have some sense of calm, I guess, similar to you guys Been in the industry a long, been through, I think pretty much all these other big wipes. And I think for all of those there was always kind of a. You don't know where it stops. Right. Like during COVID or during. You don't know what the next one's going to be. This felt very idiosyncratic and like, okay, there's this weird Trump tariff scheme, but there wasn't anything more, I guess fundamental to the market that would make you kind of freak out.
- [06:12]
Haseeb Qureshi
Tarun, how did you experience it?
- [06:14]
Tarun Chitra
I guess I had a different experience. I actually think it's funny to think of VCs as the FEMA showing up after the hurricane type of thing. That's kind of podcasts and tweet threads about it.
- [06:26]
Haseeb Qureshi
I did put on my little red cross. Yeah.
- [06:30]
Tarun Chitra
Yeah. So I had a kind of different experience because Gauntlet, we run something like 4 or $500 million of delta neutral strategies. So I had direct experience with getting auto deleveraged in different places and dealing with downtime. And I will say I think the, the war between Binance and hyperliquid over who is more transparent seems to be a Pyrrhic victory for whoever wins that, because the ADL policies are not at all transparent. And we. We'll talk about that later. But I actually think this kind of entire episode showed a lot of secret things that are not. Are like, purposely kind of opaque in both centralized exchanges and decentralized exchanges. And those things that were opaque were the things that caused the most trader harm. And of course, I think the most hilarious thing about this in hindsight is, as always, this taco trade thing of Trump always chickening out happened almost in 12 hours. It was like, unreal how fast it happened. But I do think, like, no one kind of comes out looking very well, looking good. And we can talk. I can talk through how I think, like, everyone kind of has egg on their face from this.
- [07:41]
Haseeb Qureshi
Yeah.
- [07:41]
Tarun Chitra
But we're lucky that the taco thing. V shaped.
- [07:45]
Haseeb Qureshi
Yes, yes. Although not a complete V for a long time.
- [07:47]
Tarun Chitra
Not a complete V, but like, enough that to Tom's point, you know, during the ftx, every day you were like waking up with some heart palpitations of like, is it finally fucking over? Whereas, like, this was like.
- [07:59]
Haseeb Qureshi
Not that that's true. It was a pretty quick. It was a pretty quick, you know, okay, it happened and then, okay, it's over and. And we can tell that it's over.
- [08:07]
Doug Colkitt
So, Doug, if you didn't see the chart, you wouldn't even know. Probably you just saw the number. Right?
- [08:11]
Haseeb Qureshi
Like, right, right.
- [08:13]
Robert Leshner
Yeah, I, I looked at it a few days later. I'm like, wait, something happened. Besides, now I've sort of.
- [08:19]
Haseeb Qureshi
What is the. Yeah, everybody now.
- [08:21]
Tarun Chitra
Flash crash. You remember the.
- [08:25]
Doug Colkitt
Yeah, it's a lot like 2011 flash crash.
- [08:27]
Tarun Chitra
Yeah, it feels like that more than a real.
- [08:30]
Haseeb Qureshi
Right. So everybody now has learned the term adl and they've learned it probably mostly through you, Doug. So do you want to explain for our audience what is adl, especially for those people who don't trade perps, don't really know what any of this stuff is talking about. Explain auto deleveraging and what role it played over this weekend.
- [08:48]
Doug Colkitt
Yeah, yeah, for sure. So. Right. Like the, you know, the. I think the most important thing to remember about perps are they're. They're imaginary. Right. So you go long or short on btc perp. There is actually no BTC in that system. So your ability to be short is only predicated on the fact that there's someone else willing to be long. They have cash in the system. So if the system, you know, normally that works fine. People get liquidated, someone else comes in, buys lower margin, calls people step in the market, works fine. You know when it doesn't, when there's not someone willing to step in or willing to step in at like a reasonable price. Right. Like eventually you have to start taking people out of the system. There's no, there's no one. If there's no one willing to be long. Right. Like you have to start closing out shorts or vice versa. So I mean the simplest way to think about it is just like ADL is what happens when a perp exchange runs out of counterparties.
- [09:40]
Haseeb Qureshi
So how did ADL actually play out? Like so yeah, you were saying that.
- [09:45]
Doug Colkitt
Yeah, so what happens like what happened on, on Friday is that you great, like obviously BTC dumped. I think you had a lot of leverage in the system maybe because people were like farming points and they were doing a lot of these delta neutral strategies, long on one, short on another. But right. What happened is you started getting this cascade of liquidations, especially in hyper liquid and, and Binance too. I think Binance had a bunch of ADL as well. And you start taking eventually. Right. There's nothing left in the order book. And the way the system's set up is it will not fill a liquidate at hyper liquid. At least it will not fill a liquidation at a price where it would make the position have negative equity. So if you only have X amount of margin in your system and your bankruptcy price is let's say 105 and the price in the market is like 103, it pick someone on the other side and it says, well number one, you're going to be a force counterparty and number two, you're going to be on the other side of this at kind of a distorted price. So ADL does really two things. Like one, it takes people out of the system and two, it has like this socialized loss function that keeps the exchange from being insolvent. So what happened is you started hitting a lot of ADL around like 108, 107, 106. And so that started taking a lot of longs out and people are doing this delta neutral stuff. So they, you know, would be short BTC at you know, hyper liquid, long BTC at lighter and kind of getting points on both sides or, or maybe they were short BTC and long eth on hyper liquid. But you take the, you take the short side out and now you have a bunch of people who are naked long and they get liquidated. You know, they get ADL'd at 108 on their short and they get liquidated at 102. Well, that's like a major, a major loss. So you think you have a hedge position and you're saf, but you're not, right? Like, you come unhedged and at bad prices.
- [11:43]
Haseeb Qureshi
So this is one of the reasons why, you know, perps when, when everything is working well. Perps are a beautiful product. People are like, oh, perps are so great. You know, why do. Why doesn't traditional finance use perps like all of us Crypto degens. And it's like in times of real instability that people realize, like, oh, perps are really not like dated futures in that there are these edge cases where, you know, like, if you, if you just have a future, you just hold it to maturity. You really can go to sleep and wake up at expiry and say, like, okay, things are, you know, my, my position will still be there. Whereas with the perp like this, this is what happens when the system goes through real crazy imbalance where everybody wants to be short and nobody wants to be long. And it's like, okay, well, you know, there's nobody, there's nobody to pay you. There's nobody on the other side of this trade. So, sorry, the trade is canceled. You're no longer, exactly. You're no longer short this thing because there's nobody to match you.
- [12:33]
Robert Leshner
You've made too much money. You've won too hard.
- [12:37]
Haseeb Qureshi
You've won so much. We are sending them.
- [12:39]
Doug Colkitt
It's actually what they do when they're picking positions to adl. They rank them and the two criteria they use are how much P and L you're up and how much leverage you're using. Right? Because, number one, people who made the most, you figure, okay, you're going to be the least unhappy if they take you out and you leverage, Right? We want to take leverage out of the system because if it shoots back up, you get liquidated on the other way. So you're trying to take leverage down. You're trying to take most profitable people. Where that breaks down, though, is the delta neutral stuff. Because you can have massive P L on one leg and then you have, you know, a big loss on your other leg and, you know, you're not, you're not actually profitable. But, like, from the ADL systems perspective, they think you're a big winner. Like, why are you going to complain?
- [13:18]
Haseeb Qureshi
Yeah, so this was, I think what was so. Part of what was so rattling about Friday was, was how Unpredictable. People felt it was, was that, you know, there were reports of people who were getting ADL'd multiple times in the same day for, okay, I was ADL'd on the way down and then I got in and I was ADL'd on the way up. And it sort of led to a total breakdown of the predictability of perps as a way to express a market positioning. And if you were in spot, to Robert's point, you were fine. You never noticed anything happened. Everybody who was in perps noticed that the market was just breaking down under all the load. And of course, a big part of why things were so lopsided was because of the fact that the APIs were not working. And when the APIs are failing and like, people can't get money into Binance, the single largest venue in crypto, then, like, you just have these massive dislocations where, you know, some people were posting like the sole price. And it was like, it was, it was like spreads of like 20 bucks across different exchanges. Because there were some places where it was trading at 200 someplace, it was trading at like 175. And that kind of dislocation, that's like a 2015 kind of discrepancy that was just like. It just meant basically that there was a total breakdown in the ability for market makers to move inventory across exchanges. So one of the notable stories from the day was a massive de peg that took place on Binance across three assets. The first was Athena's USDE, which is the stable synthetic dollar that Athena has, and then two staking tokens that WBeth and BNSol, which are the sort of staked ETH and staked SOL equivalents that Binance has and these two assets. So USDE traded down to $0.68 on Binance before finally recovering. Now for other venues, and I wrote a big tweet thread about this. Other venues actually did not have this problem. So if you look at Bybit, Bybit traded to like 95 cents for USDE for like 5 minutes and then quickly recovered the peg. And if you look on Curve, which is the most liquid venue for usde, it was at basically a dollar the entire time. But Binance had this huge leg down, which got pretty widely reported as, oh, shit, USD is DE pegging, blah, blah, blah, it's breaking at the same time. These two other assets, the. What is it? WP eth and to be in Sol, they went down like 80% before regaining their pegs. Now, it was claimed by YQ, who's a founder, Asian founder was saying that he thinks actually what was happening was a coordinated attack because USTE was used as collateral for WBE and bnsol. And if you could push down the price of USDE on Binance at a time of instability, you could liquidate a bunch of people's wbeth and be in sole positions or something like this. There's something about the index weighting where they were not using the redemption price, but they were using the order book price. I don't know if you guys follow the story, but there's a lot of theories floating around about what exactly caused this and whether this was actually an attack.
- [16:23]
Tarun Chitra
I have a more Occam's razor slash, less adversarial theory for why the USDE USDT order book was so skewed. One really big event in the last two weeks for USDE was the getting added in Binance as collateral, but also being used for farming plasma. And in particular there was Binance Earn where users could just click a button that said hey, use the stable asset. I think actually it did say stablecoin in the doc, but obviously maybe not. It shouldn't be kind of described as that. And earn up to 30% on Plasma and Plasma was at ICO was quite high market cap wise like 20 plus billion. Now Plasma was one of the first assets to have this huge collapse. Had a really violent collapse on Friday and that was right around the time the USD EUSDT order book kind of sort of unwound. And I think there's a more boring explanation than there are all these people trying to snipe the index construction methodology is there were a lot of users who've never used Athena before didn't really know they're using Athena. They clicked a button on Binance that said get 30% and they see XPL going down and their 30% looks like 6% and they're like exit, exit, exit. I want USDT and this kind of mass of users who had never, they might not have even known they were using Athena, right? There was $2 billion that went at least that went into Binance Earn. And I'm sure a lot of those users didn't even know they just had tether on Binance. And so I think there might be more naive explanations like that of some of these really low liquidity assets really had these collapses that in the UXS and in a lot of centralized exchanges looked like people's yields went down 80% and then they exited and they hadn't, you know, it was a new like the timing around that to me was like kind of very interesting.
- [18:37]
Doug Colkitt
You think that happened fast enough for like.
- [18:39]
Haseeb Qureshi
Yeah. You think like people are, I think they're slamming like I'm not getting enough yield button.
- [18:43]
Tarun Chitra
No, no, I, I think there's, I think there's people who copy like once the, once the USCT book started lifting there's tons of market makers just copy trade it down right. Like momentum wise and that can become a self fulfilling prophecy. So like that to me is. I don't know if it's like this like index weight manipulation. It, it seems like very complicated and you have to time it perfectly. Whereas like oh, just copy trading all these people exiting this earn product. It seems like a.
- [19:10]
Haseeb Qureshi
So I, I agree that it's an exotic, it's an exotic theory and I think like the base case is just that somebody was like hey I don't want USD, I want usdt because market's dumping that. That, that does seem more plausible and that just no market makers are able to step in and provide the inventory to, to bring the peg back. That being said, I think the, it's not implausible that somebody just had an algorithm that was sitting there waiting like look, if this thing goes down a cent, just ram it as hard as you can down and then start shorting, you know, bn, SOL and whatever else.
- [19:39]
Robert Leshner
Or somebody was getting margin called on their perps.
- [19:41]
Doug Colkitt
Yeah, I think that's speak and they.
- [19:42]
Robert Leshner
Were like I have a huge position, I need cash instantly. Like who cares about slippage if I'm going to lose way more money. Again, I don't have any clarity inside of Binance because it is a centralized exchange that is somewhat opaque. But my hunch is just that like people needed money urgently as they do in flash crash scenarios. Everyone's getting liquidated, everyone's running around like a chicken with their head cut off. Someone just hit the sell button to.
- [20:08]
Doug Colkitt
Get cold hard cash or the liquidation engine just you can use it as collateral and just force you out.
- [20:15]
Tarun Chitra
Well, I do think the USDE listing as collateral timing relative to this was like it wasn't that long ago. Right?
- [20:21]
Robert Leshner
There's so much interconnectedness that like the.
- [20:24]
Haseeb Qureshi
Well the thing is part of the reason why YQ pointed out that this theory would make sense is because there was only a couple days until they were going to change the methodology for BNSOL and WB whatever eth and after that this attack would not work. And the fact that they depegged so aggressively relative to everything else, no other staking token crashed in the way that those two assets crashed and those two assets were the only two assets that were going to have their.
- [20:50]
Tarun Chitra
I just think it, I think, I believe it's an accelerant that once there was starting to even be a small deviation, people started maybe looking at that, but I suspect it's something really dumb that started. The true first domino was probably not that. That's my strong belief that's very plausible.
- [21:11]
Haseeb Qureshi
And to be clear, we'll probably never know because I don't think Binance would publicize it one way or another. One of the things that was interesting, people were talking a lot about the microstructure and Guy was pointing out that, hey, a big part of the reason why Binance dislocated so much beyond just Binance's own instability and API issues was that Binance, unlike many of the other exchanges that integrate usde, Binance won. They were using their own order book as the Oracle price, as the index price, which is not what you should do if you're not the primary venue, which, you know, Binance was not for usde. They were not looking at any of the other prices which caused them to liquidate USDE way too aggressively during the flash crash. And that liquidation, to your point, is further caused more selling and just drove the USDE price down. And then second is that all the other exchanges have integrated Mint and Redeem for usde. And so if you're a market maker and you have inventory on Bybit, you can directly mint and redeem and do PEG arbitrage for Athena. Binance was the only major venue where they didn't have that integrated. And so as a result, if you're a market maker, you can't get inventory in, you're not going to take assets off exchange and then go do PEG arbitrage. Right, because it's just like, especially when the APIs aren't working, the API is off like nothing. Yeah, exactly, exactly, Exactly. If the APIs are failing, like you're not, you're not going to click your way into like bringing your assets out. So it was just, it was kind of a perfect storm of everything going wrong with respect to the market structure on Binance. And so this is also part of the reason why specifically Binance only gave refunds, $250 million worth of refunds to people trading USDE, BNSOL and WBE. Those three assets were the ones that got 250 million plus worth of refunds from Binance, who identified that, hey, these improper liquidations, these improper liquidations, we're going to make you whole. That to me seems like a sign that this was not just like, oh, somebody innocently got liquidated and it wasn't really an issue on their part. That feels to me like something must have been wrong that they identified like, oh, we fucked up beyond just the oracle construction. Okay, so one of the other stories of course is who did well and who did poorly. And this is what a large part of what's been debated online is like hey, you know, defi obviously everyone's like oh yeah, aave was great and you know, morpho. They liquidated stuff without. With no bad debt. Which of course is like, you know, okay, it's easier for them to get through this than it is for the perp Dexs. Love to hear your guys reflection, especially maybe Doug, because I know this is something that you've spent a lot of time on. Tell us how you saw how Defi, especially on the perp Dex side, how it fared relative to what happened in cefi.
- [23:55]
Doug Colkitt
Yeah, I mean I think right, like it depends on the perp Dex. Some of them had API issues too. And I think right. Like this kind of opens up how decentralized are some of the perp Dexis if they're. They're just an API or a box that's sitting somewhere, as far as I know. I think Hyper Liquid held up pretty well. I think it dragged a little slow. Lighter had a bit of offline time but you know, luckily that was after kind of the craziness so people could still get stuff in. So I think in terms of performance, the perp Dexs did pretty well. I think they were pretty comparable to centralized exchanges. Obviously the transparency is a big issue with centralized exchanges. I think that kind of stands out. There is that there's some stuff around ADL or what the methodology is in finance and some undisclosed stuff and learn happy about that and whatever you want to say about adl, at least Hyper Liquid is very transparent. Like these are the formulas is what we use, everything like that in terms of the end users. You know one thing about perps are they're, they're zero sum. So for every winner there's a loser. And so I think like not any significant number of people made money and they're just maybe not yelling about it on Twitter because it's bit tone deaf to do that.
- [25:06]
Robert Leshner
Yeah, it's a very poor form to be celebrating when everyone else has lost bigly.
- [25:11]
Doug Colkitt
Yeah, yeah, exactly. So I think there are some people who are still fine and then the people I think who really got screwed are the delta neutral farmers. I kind of said this before, but if you were, you know, you thought you, a lot of people have gone into this strategy long, long on one per deck, short on another, and you're collecting, you're collecting points. And I think that produced a lot of kind of artificial OI in the system that was probably, probably a big contributor like to the kind of the violence of the deleveraging.
- [25:37]
Haseeb Qureshi
So okay, actually I, it's funny because I actually brought up this point in a panel I was doing today at das, which is the point you just made, which is that look, purpose is your sum, right? And so to the extent that this was a derivatives driven event, half the people won, half the people lost. Mathematically that has to be true. And yet that doesn't feel right at all. Right? It doesn't seem like half the people won, half the people lost, because that should be true in any given day.
- [26:05]
Tarun Chitra
I think there is actually a question of the types of agents who, who won and lost. And like this is just a fundamental thing about exchanges and their users in these tail events. They are extremely adversarial to each other. Exchanges make decisions that either choose to preference the exchange right now, that is keep the exchange solvent right now and not lose any revenue like instantaneously at the expense of sort of traders whose positions get stopped out to cover very lost positions. Right. That's like the socialization. And this is sort of not something unique to crypto. I think crypto people just think this is kind of like unique to perps. But like the whole Basel 3 framework is about interest rate swap clearing and interest rate swap clearing has this exact same problem of like when there is a sort of liquidation event, how much do you get apportion the loss to the exchange and how much do you apportion the loss to the set of traders involved? And in general, the different exchanges took very different perspectives on this. So in the case of hyperliquid, hlp, depositors were preferenced in a lot of ways over a lot of profitable traders in the sense that they had very aggressive auto deleveraging and that basically HLP got to kind of make their profits from that. And in some sense that, that is sort of this belief that your zero sum kind of always in the sense that it's either the exchange that loses and goes insolvent or it's like the profitable traders who kind of lose because their positions are cut early. But I think there's kind of this, there's this old saying from poker, I'M trying to remember this guy Amario Slim. I remember kind of a long time remembering, which is you can shear a sheep many times, but you can skin him only once. Referring to kind of like the difference between kind of like poker sharks and minnows. And there's a sense in which if you preference the exchange's preferences over your traders who are profitable, and those are likely the highest future revenue generating traders, they may not come back to your exchange or may lower their size because they're like exactly at these tail events when I'm making the most money, because I predicted the directionality the most correctly and you stopped me out like why the fuck would I come back? And so I think this is going to be an interesting design choice that we're going to see. And fundamentally I think a lot of these ADL mechanisms are very non transparent. Right. It's like in the case of binance, it's like your relative P and L times effective leverage is how you get ranked. In the case of hyperliquid it's just like your relative P and L. But these are sort of like greedy algorithms. They don't actually necessarily optimally match losing positions with winning position like the minimal number of winning positions to get solvent. They're actually just greedy algorithms that are like pick the biggest thing, pick the biggest thing, pick the biggest thing. And in these cases where there are, to Doug's point, these kind of people who are doing these cross trades, oftentimes their positions get picked off the most even though they're not contributing the most delta or risk to the order book in the case of insolvency. And so these might sound like technical, nuanced things, but like literally half of the fucking Basel 3 is about this type of stuff. Like that's why it's like that does sound very technical. But I think all of the perps exchanges are kind of slotting themselves on a spectrum of like how much they value the exchange and how much they value their highest value traders. And they're making this decision between instantaneously, it's zero sum all the time or I view the long term thing of like trying to retain users, but maybe the exchange has to eat some, some amount of insolvencies. Right. And that's, but there's, but there's no.
- [29:59]
Haseeb Qureshi
Exchange that was internalizing a loss.
- [30:02]
Doug Colkitt
Yeah, I don't think any lighter. Lighter had less adl, but I think that was partially because LLP was providing liquidity.
- [30:10]
Tarun Chitra
But that's what I mean by the equity holders.
- [30:12]
Tom Schmidt
Yeah, not the Exchanges.
- [30:15]
Tarun Chitra
Yeah, yeah, sorry, sorry, sorry. In Dexs, right, the LP vaults and exchanges are separate, but in centralized exchanges they're kind of the same often.
- [30:23]
Doug Colkitt
But actually in the original Bitmex model, I think Binance has insurance fund too, but who knows when it actually attaches. In the Bitmex model there is an explicit insurance fund that sits there and does take insolvent positions up to a certain point. I don't believe any Herp Dex has that.
- [30:41]
Tarun Chitra
I think drift is the only. But yeah, all the other ones. So my point is there is this thing where you're apportioning the loss, you're splitting the loss somehow and everyone's choosing a different split. And I think until this event, because ADL's are so infrequent, people were just like, ah, whatever. We chose a greedy algorithm, it probably works. I don't think there was a lot of very thoughtful design in this choice of these ADL ranking functions. And now there is insane amounts of people looking at this because clearly those preference certain types of users quite a bit to others.
- [31:16]
Haseeb Qureshi
Well, I think it is important though to underscore that ADL is extremely rare. Right? It's happened very few times in the history of these perp Dexs for centralized exchanges only. Really these crazy tail events where that happens. Normally the funding rate is supposed to solve the problem of bringing in people on the other side and balancing out the longs and the shorts. So there are times when funding gets really lopsided. But adl, like ADL is genuinely an edge case and it's important to not have people walk away with the impression that like, oh, you know, ADL is like a very fundamental part of how perps work. It's. It's a fundamental part of how they can work in the tail events. But normally this should never happen. You know, normally there should be enough latency to bring the other side into the market.
- [31:56]
Robert Leshner
Right? I mean, I don't think ADL's are going to stick around as a major conversation point, you know, for the same reason that very few people trading on the CME are talking about like the clearinghouse rules, you know.
- [32:07]
Tarun Chitra
And like, I actually don't agree with that because during COVID all the clearing agents blew up and then these funds had to die because their clearing agent up and didn't post collateral. Ronin Capital is the greatest exam, if you want to Google the greatest legal case on this, Ronin Capital versus CME to 2020.
- [32:24]
Haseeb Qureshi
Totally.
- [32:24]
Robert Leshner
But that, but that's like of hedge funds of the 99.9% of the people that are trading perss with 50x leverage, okay. They are not professional hedge funds. Okay. So I think that what's going to happen is 1% of the users are going to ask hard questions to all of the different venues that never really existed before. Just like when smart contract hacks started happening for the first time, everyone really started asking about Oracles and smart contract audits and all of these things that they never knew to really ask about before. Now a small number of people are going to get very noisy about these edge cases, but 99.9% of society, I think, is going to forget about ADL until maybe the next mass liquidation event of this scale.
- [33:15]
Haseeb Qureshi
Yeah. So adl I think is reflective of a design choice, but the deeper design choice, if you look at lighter versus Hyper liquid, going to this conversation with a caveat that I think most of us are investors into lidr and at least Dragonfly owns some hype if you look at OLP versus hlp. So these are the vaults on LIDR versus Hyper liquid on lidar. The idea, and this is something that I think Vlad was on Twitter actively arguing like, no, no, no, no, this is good that we did this. OLP lost money, HLP made money. And this is because, I'm sorry, llp, not llp. Llp, LLP made money. And sorry, the other way around, LLP lost money, HLP made money. And now from one perspective, you could look at this as a depositor and say, that's great, I want to be in the vault that makes money. I don't want to be in the vault that loses money. But this is zero sum against your users, right? So that means that more HLP traders lost that money and more OLP trader, or sorry, more lighter traders made money because LLP was stepping in there to take some of those toxic liquidations as opposed to having those traders be fundamentally in a worse position. Because all that P&L zero sum, right, it's coming from somewhere and the somewhere is the traders. And so the question in some part is also, what is the purpose of these vaults? Why are they designed to be the way they are? Tom, I see you nodding. I'll hand it off to you.
- [34:37]
Tom Schmidt
No, I mean, it kind of reminds me a little bit of the Jelly Jelly episode with hyperliquid a couple months ago, where it was clear that HLP was being preferenced over traders or exchange solvency or contract solvency. And I think for a lot of people, including myself, that left kind of like a sour taste because I was also under the kind of impression this was supposed to be kind of this general public good, almost kind of like a investable insurance fund like Bitmex, as Doug was saying. And I think this sort of episode sort of showed that that is true, that this isn't supposed to be a true profit maximizing strategy that you can invest into. It's like it is supposed to be kind of this backstop for the exchange and generally that works out pretty well. But in some scenarios you might lose a little bit. And I think obviously people also love drama when also both of these vaults are up pretty massively over the past year. So it's like a 5% drawdown or a 5% gain on a single day doesn't really impact the P and L of the vault strategy. But the point is that your P and L is coming from somewhere. And so would you rather have it come from the traders or rather have it be directed into the vault? I also like this data point from Conor Grogan at Coinbase who does a lot of on chain forensics that you know this is sort of like a little mini extinction event for I think you looked at hyper liquid but I'm sure it's true for a lot of perps, dexs or perp exchanges that like almost half of the top traders, top wallets on HyperLiquid lost like 50% of their worth and like 100 got totally wiped out to zero. And so it is like you know there's a small number per trading and trading in general is like so head concentrated that you do kind of want to protect those, those traders and like have them continue to trade. And so obviously when there's this sort of huge catastrophic event like ideally you don't want them to be kind of wiped down to zero.
- [36:20]
Tarun Chitra
This is why I brought up that quote about the sheep because it is really boils down to like your future round. This is a multi round game, right? Like if you liquidate those users that's future revenue you're not getting back either because they're like tastes in their mouth is like I don't want to go to this exchange that over aggressively liquidated me. Or they're just out of money, right? And so I think there is kind of this very distinct trade off in these, in between the vaults and the, the traders. But, but in centralized exchanges between the centralized exchange and the traders, right? Because they, they kind of are the ones enforcing these rules.
- [36:55]
Doug Colkitt
Well I, I think one, one major difference between HLP and LLP now is that HLP originally was both liquidity provider in normal times and liquidations when the order book couldn't absorb it. So originally like HLP was very involved in the book and now it's like down to 2% of the order book and LLP and lighter is still provides a large fraction of liquidity. So I think it's actually not at least my understanding is not as much a function of the liquidations are handled that differently. Both of them actually kind of do a similar thing when the order book can't absorb the liquidations. I think the difference was that lighter was in there as a market maker and actually supporting the books outside the liquidations and hyper liquid and everywhere else. When normal market makers went offline, you just, just didn't have, you didn't have anything in the book outside of that.
- [37:44]
Haseeb Qureshi
So LP was just getting run over like a market maker.
- [37:47]
Doug Colkitt
Yeah, I think LP was just sitting there and absorb, you know, as it was going down by, by, by, by, by like you can see, like you can still see there are some like long tail coins. Like the last one is like Kanye coin where it's kind of stuck in this it's tiny position. So it's not gonna affect anything, but it's like stuck as like half the OI in Kanye coin. And it's like no.
- [38:03]
Tarun Chitra
So, so I, I, I. One thing I will say is I kind of think the Jelly jelly incident that Tom mentioned, which we did cover on an episode because it's a very embarrassing fact for Sam lesson which I want to remind everyone of that that coin exists. But the Jelly jelly incident I think made sort of hyperlogre get way more aggressive on their ADL policies after that. And I think they maybe overfit to the small cap coin ADL but then applied it kind of everywhere. And that was sort of very apparent in the difference between, between the two. And I just think like this is a, this is the kind of weird thing that, you know, I know Robert's.
- [38:43]
Haseeb Qureshi
Saying, hold on, I thought Jelly jelly was not adl. I thought they just like paused the market or something.
- [38:48]
Doug Colkitt
Changed the Oracle. They force changed the Oracle.
- [38:50]
Haseeb Qureshi
They changed the Oracle.
- [38:51]
Tarun Chitra
Yeah, yeah, yeah. But, but the point is they couldn't really do ADL because it took too much oi. It was like too like it would have caused more issues if they actually executed right. So if they did it when it was smaller size before the position got big, it would have been. So that incident will make you refit your ADL model. What quantile you start liquidating on.
- [39:11]
Doug Colkitt
I think in general makes you think of what's going to happen if I can't get out of a position?
- [39:15]
Tarun Chitra
Yeah, exactly.
- [39:16]
Doug Colkitt
ADL or liquidations or force Oracle or anything, right? At one point you have to get out of a position.
- [39:23]
Tarun Chitra
And I think these things, even though they seem minute, I think they actually will show up in trader retention. If I look at six month, one year retention of OI and traders, I think this stuff is sticky. People left BitMEX because of this. I feel like this was like half the reason people were like, oh, they're just always purposely liquidating me and like.
- [39:44]
Haseeb Qureshi
So okay, I want to go back to the question I posed which is theoretically on any given day, not just Friday, but every single day, half the people are winners, half the people are losers. When you're trading perps, that's just how perps work. It's not like spot, right? With spot, like everybody can be winning and it's all great. So if this is true, why does it not feel true? It feels like a little bit of a paradox that everybody seemed to be having a terrible day on Friday, much more so than usual, even though the animated should not be true.
- [40:13]
Robert Leshner
For the same reason that when you walk into a casino, right, you hear a lot of complaints from the people who lost $100 and you don't hear that much cheering from the people that made $100, right?
- [40:25]
Haseeb Qureshi
But isn't that not, is that not true every single day? Yes, but like it's not true every single day.
- [40:33]
Doug Colkitt
It's right, some people win, some people lose, but people have different sensitivity to, you know, for whatever reason. So I, I, I do know at least a couple of larger firms are not doing great, took some losses. Market making firms took some losses from this because they are dependent on like we're going to hedge here, we're going to hedge there and whether the APIs are down or they get ADL'd and like for them, if I'm a guy who punts and I'm up 30%, I'm a gambler, that's just another day. But if I'm a market making firm, I'm wiped out. That's not great.
- [41:04]
Haseeb Qureshi
Okay, so maybe the way to describe it is that the P and L is equivalent. On any given day the P and L is equivalent. P and LS cancel out. But the allocation of risk got all janky is that normally market's supposed to allocate risk correctly and say, okay, you're long here, you're short here, you're hedged because you want to remove your risk. And when ADL start happening and everything starts going topsy turvy. That's when the allocation of risk just breaks and people who weren't actually taking risk are suddenly taking risk.
- [41:31]
Doug Colkitt
It's the opposite. Yeah, it's the people who are most risk averse were forced into taking risk.
- [41:36]
Tarun Chitra
Yeah, exactly, exactly.
- [41:37]
Doug Colkitt
I think a lot of times like.
- [41:39]
Tarun Chitra
I'm sorry, no, go for it, go for it.
- [41:40]
Doug Colkitt
I was just say a lot of times like financial crises, like it's not necessarily the amount of loss. It's when stuff that's perceived outside anywhere, it's when stuff is perceived as safe and all of a sudden it's not safe. Like those are the worst financial crisis when a lot of like levered hedge funds lose a lot of money. It's like, well the system can absorb that versus like a money market fund breaks a buck.
- [42:00]
Tarun Chitra
I think another thing is the exchanges made money. That's like to Robert, extending Robert's analogy, you're saying it's zero sum. But there is a sense in which the exchanges took a wider vic than they normally do. And that feels bad for people. Right? The exchanges took like even higher percentage of than normal.
- [42:19]
Robert Leshner
But to add one more layer to this, the people that made money were short. They were short going into this insane scenario.
- [42:26]
Doug Colkitt
Right.
- [42:27]
Robert Leshner
And I think the to your point to see why don't we perceive it that it's zero sum and everything that's out is because a lot of people lost a lot of money Being long crypto. Long is good. Long is what everyone is supposed to be in crypto and when they got wrecked, they all went to the timeline and acknowledged the fact that they got wrecked.
- [42:48]
Haseeb Qureshi
But it was, it was like a V shape. Right. So there's as many people making money on the long backup.
- [42:52]
Tarun Chitra
Well, it's a V shape for some people.
- [42:54]
Haseeb Qureshi
No, but everyone who got liquidated, some people got, I'm sure you know, but.
- [42:59]
Tarun Chitra
But this gets back to my point about the multi round game. Like if you liquidate all your users, they don't have capital to come back for a lot of them. Right. Like and they're not going to come back.
- [43:09]
Doug Colkitt
The other interest ADL made some people who were up actually complain. Like, like Andrew Tate was like, I was like I can't believe this is what the hell is auto deleveraging? And somebody said well if you didn't get auto D leverage, you'd actually have lost money because it liquidate, it took you out at a good price.
- [43:25]
Haseeb Qureshi
Yeah, that, that, okay, that, that, that makes sense. Is that people even who were up because of. Because they were adl. They were mad.
- [43:33]
Tarun Chitra
Yeah.
- [43:33]
Haseeb Qureshi
So I think it's just like that unexpectedness. It's the unexpectedness that people really do not like and causes everybody to complain and get angry because again, it's like this allocation of risk thing is that people who are not supposed to be down or down and people who are not supposed to be up or up, and the people who are up aren't out there being like, yeah, I made money. This is great. Because they were not out there to try to make a bunch of money.
- [43:55]
Robert Leshner
There was way more complaining in the casino than celebrating, honestly.
- [44:00]
Tarun Chitra
Look, I don't really have many thoughts about Mr. Tate, but I will say the video he made. So you wrote a tweet that says leverage trading is satanic, and he made this video is kind of hilarious. It's a video about this having a profitable position, getting ADL'd. So I'll tell you people, if you want to see the emotion that comes out of this, watch that video.
- [44:25]
Haseeb Qureshi
Boy, I do not. Wow.
- [44:27]
Robert Leshner
A quick search. I have found the video. After this show, I will watch it.
- [44:31]
Haseeb Qureshi
Okay. All right, great. So, I mean, another thing, of course, as with any aftermath like this, we saw a huge wipeout in the total leverage in the system and the open interest on almost every single platform massively dropped. Alts were hit a lot worse than Bitcoin. But you saw basically alts open interest pretty much halved. And a lot of people were blaming the massive drawdown that you saw in alts. Many alts wicked down like 80% plus. Even assets that were worth nominally 10 billion plus had huge, huge drawdowns. And people were claiming that a lot of this is because of cross margining, is that basically this is the first massive drawdown where you actually had these really robust cross margining systems where your sweep position or your IP position was getting liquidated as collateral for whatever else you had in your portfolio. And so this caused this kind of reflexivity on the way down for a lot of alts. But that open interest hasn't built back up yet. It seems like it's now starting to build back for Hyper Liquid. I think Hyper Liquid dropped the most in open interest. I think it more than halved after the drawdown.
- [45:36]
Tarun Chitra
But it seems like it's 15 bill to 6 bill, something like that. 15 or something.
- [45:41]
Haseeb Qureshi
Yeah, Quite, quite large. Massive, massive haircut. And I'm sure some of this was because of, you know, the basis straight across different purpose exchanges. But it seems like some of that interest is now Coming back, obviously, it's Monday. You know, Trump is tacoed, as they say. And so, you know, all.
- [45:57]
Tarun Chitra
I just think the most hilarious thing is my group chats where there's like, crypto people and equities people. And the equities people were like, ha, ha, markets closed. Fuck you guys. And then the crypto people are like, having this, like, meltdown. It's like kind of a funny.
- [46:11]
Doug Colkitt
People aren't even back to work yet. It's.
- [46:13]
Tarun Chitra
Yeah, exactly. They had today off.
- [46:17]
Haseeb Qureshi
Yeah. Yeah. It does kind of feel like somebody threw like a smoke bomb and then closed the door and.
- [46:23]
Doug Colkitt
Well, I mean, it also probably, like, didn't help that this happened right after market close. Right. Because if you have the ETFs trading, like, at least you have some liquidity on, like, you know, the. The actual market to buy the dip when it goes down. But, like, you didn't even have that.
- [46:37]
Tarun Chitra
I. I will say I did enjoy plotting the Trump token complex against basket of other alts, and it was down a lot more than the rest. Like, during that spike, it, like, really? Trump, really? Trump and Melania, like, really sold off. Not as much as Adam, but they were like 70, 80% down.
- [46:56]
Haseeb Qureshi
All right, well, speaking of Trump, one of the other stories was talking about the buildup to this whole drawdown, which was what's now being called the hyper liquid whale. So what happened is, of course, this whole thing was catalyzed by Trump tweeting that, oh, hey, we're going to restart the tariffs against China. And apparently about something like 30 minutes before this announcement took place, there was a whale who had hundreds of millions of dollars worth of bitcoin that they deposited on hyperliquid. And they deposited this just a very short period of time before the announcement was made and close that position, I think within minutes of the announcement. And the big drawdown that took place, making people believe that this was somebody who had advanced knowledge that Trump was going to do this, and their P and L was something like $200 million that they made on hyperliquid, entering into this massive short position. That may have been the initial push for markets to start teetering over, just absorbing the shock of this massive. This massive short that started on hyperliquid. So if you remember, there was another episode like this also on hyperliquid. What was it? What was the precipitating event? It was also a truck.
- [48:09]
Tom Schmidt
I think it was also the bitcoin whale, this guy with really old bitcoin address who just deposited through unit, like, just Like a billion dollars in bitcoin.
- [48:16]
Robert Leshner
Or something like that wasn't tied to a specific event?
- [48:18]
Tom Schmidt
No, no, no, no. It was just huge. You just see this? Yeah.
- [48:22]
Haseeb Qureshi
Wait, is it the same whale?
- [48:23]
Tarun Chitra
No, no, no.
- [48:25]
Haseeb Qureshi
This is a different guy, right?
- [48:25]
Tom Schmidt
This is a. Yeah, yeah, different.
- [48:27]
Haseeb Qureshi
This is. Yeah, sorry, yeah. Yes, this is a bitcoin whale. But there was a.
- [48:30]
Robert Leshner
First time it was Tarun. This time it was Tarun wearing a.
- [48:34]
Tarun Chitra
I wish that would have been funny.
- [48:37]
Tom Schmidt
I think it was notable because that was like a super old address that just woke up after like 10 years and started to go punt some, some perps on Hyper Liquid.
- [48:44]
Haseeb Qureshi
Right. But so this, whoever this person is who did a massive short on Hyper Liquid, almost certainly they were somebody who had advanced knowledge that Trump was going to tweet this, which makes it seem that, like. Okay, well, that's, that's not a long list of people. It must have been somebody who's connected to the US Government or maybe somebody in China who got advanced. Notice that, like, hey, we're going to.
- [49:03]
Robert Leshner
I think it's circumstantial or coincidental.
- [49:05]
Doug Colkitt
Like.
- [49:06]
Haseeb Qureshi
You think it's circumstantial coincidental?
- [49:08]
Robert Leshner
Yeah.
- [49:08]
Tarun Chitra
Was it like, wasn't the Chinese announcement the night before that the rare earth thing, whatever, that he got mad about?
- [49:15]
Tom Schmidt
People were saying that like, this is like an open conversation topic in China, that like this was already happening. And so it's not like maybe there's.
- [49:22]
Haseeb Qureshi
Some coincidence shrugged it off. Right. Because this news came out the day before and news markets are like, ah, whatever. You know, China's doing something.
- [49:29]
Tarun Chitra
Crypto is a global market haseeb and it incorporates global information. The US Markets might not for this particular thing.
- [49:37]
Haseeb Qureshi
So you guys don't believe the conspiracy theory that this person had some kind of insider knowledge?
- [49:42]
Tarun Chitra
Isn't he losing money on this current short that he's doing like quite a bit?
- [49:45]
Robert Leshner
Maybe it's to cover up his insider knowledge.
- [49:47]
Tarun Chitra
Yeah, yeah, that's the theory. Sorry.
- [49:51]
Haseeb Qureshi
That actually is very plausible to me. I, I actually find this extremely plausible.
- [49:54]
Doug Colkitt
Well, at the very least, at the.
- [49:56]
Haseeb Qureshi
Very least, the amount of attention on this address is like insane. You know, it's like the, the number one story I saw on social media was like, focusing on this address.
- [50:04]
Robert Leshner
Every theory is based in something plausible.
- [50:08]
Haseeb Qureshi
Right.
- [50:08]
Robert Leshner
It's not like it's like, you know, that unreasonable of an idea.
- [50:12]
Haseeb Qureshi
Here's what I'll say. Here's what I'll say is like the lesson that I've learned from being in crypto for like the last Five years is that. It's almost always exactly what you think it is. It's just like you just keep like 2022 taught me that. Is that like.
- [50:24]
Tarun Chitra
It's razor.
- [50:26]
Haseeb Qureshi
I mean, it's just like. That's like the crypto razor.
- [50:29]
Tom Schmidt
Yeah, we did get the other Occam's razor. This is EB's Razor with James Wynn this week too, where there are all these conspiracies that like, oh, James Wynn is cz or like, actually it's Jeff and he's trading to like, you know, make the exchange more notable. And then he went on like a spaces and just some like random British dude who honestly sounds kind of stupid. You just complaining. I think he was complaining about getting ADL'd this time around.
- [50:51]
Tarun Chitra
Yeah, he sounded pretty drunk.
- [50:53]
Tom Schmidt
There you go. That's the simplest answer.
- [50:55]
Haseeb Qureshi
Okay, so you guys. Okay, so you guys are totally zero credence in the story.
- [51:00]
Tarun Chitra
Actually, I will say one thing that was funny about this address. How many conspiracy theories were they trying to dox this guy? No one was trying to dox his billion dollar depositor or at least I didn't see any like active effort. Whereas whoever. This is like, it's like a. An Internet crime documentary.
- [51:15]
Haseeb Qureshi
True. It's a scandal. Yeah. Because this is like insane, you know. Right.
- [51:18]
Robert Leshner
It'd be like one. You know.
- [51:20]
Haseeb Qureshi
Yeah, yeah. This would be like extremely, you know, national news if it was discovered that somebody was insider trading. Whether it's on hyper liquid or X us.
- [51:31]
Robert Leshner
Like if there was some spy that got this info. That's just as.
- [51:35]
Doug Colkitt
It's like intelligence.
- [51:37]
Haseeb Qureshi
True, true.
- [51:38]
Tarun Chitra
Imagine this Watergate. Except. Except the CIA breaks into an apartment and it's like a mattress on the floor with no bed and no other furniture.
- [51:45]
Robert Leshner
James Wynn, drunk out of England, trading.
- [51:49]
Haseeb Qureshi
Just crazy. Hold on, Doug. Doug was about to put in a word where I was gonna say.
- [51:54]
Doug Colkitt
I was just gonna say. At least it answers the question of who's. Who was up and you know, who's celebrating. Celebrating the casino is Barron Trump. I guess we'll be okay.
- [52:03]
Haseeb Qureshi
Yeah. That's why we. That's why we're not hearing about this about the winners. Because the winners all Baron Trump. That's the answer. He has all the piano on the other side. Yeah. Okay. Well, I think we can now say that we're mostly at the end of it. The dust is settling. Obviously there was a lot of snafus and a lot of learnings on the infrastructure side. I'm curious for. For. For you, Doug. For Tarun. What are you Guys seeing from. From infra providers. Right. So there was a lot of failures and a lot of stuff to just totally fell over amidst the market mayhem. Any learnings, reflections about like how things are going to be hardened going forward?
- [52:38]
Doug Colkitt
Yeah, I mean it's always, it's always tough, right? Like it's just crypto, the level of traffic, like you're on a dex or run an exchange, right. It's just the level of traffic spikes that you get on this stuff. It's just power law. So it's always, always tough to like predict this like how much traffic you're going to get until it actually, until it actually hits. But you know, at least, at least you have some data points there. I don't know. It's crypto. Crypto infra is always. Should be fixed and it's always like, oh, the exchanges are always falling over and they've been falling over for 10 years. And my guess is that most exchanges will be falling over when. Shit.
- [53:13]
Haseeb Qureshi
Well, so I, you know, somebody was asking me this today of like, why can't you just load test this stuff? And I was like, well, it's not that easy. Load tests don't always reflect reality. Da da da da. But I was like, I don't, I don't actually know why you can't load. You know, why, why, why does it keep falling?
- [53:30]
Tarun Chitra
People do, do their best to do that. I think it's just like always some unforeseen fucking thing like, oh, like in, you know, like Ethereum gas went up too much in my, like my L2 transaction. That's not getting in like, like there's so many.
- [53:47]
Haseeb Qureshi
Like that's one of the things that you should be load testing.
- [53:49]
Tarun Chitra
I don't know. It's not that they don't, it's not, I don't, it's not that people don't.
- [53:54]
Haseeb Qureshi
I'm starting to think that they actually don't.
- [53:56]
Tarun Chitra
Like, I don't, I don't think people definitely do a lot of this type of stuff.
- [54:00]
Doug Colkitt
I mean, I guess one thing, you don't necessarily get punished in the market, right? Binance went down for two hours. Or are they going to lose? Are they going to get punished?
- [54:06]
Haseeb Qureshi
They lost 283 million.
- [54:08]
Doug Colkitt
Yeah, that's true. Yeah, that is true. Yeah.
- [54:11]
Tarun Chitra
But I actually would give some, some credence to the, the Solana L1 kind of version of the world, which Doug is, is sort of also favorable, which is like, it forces you to test this more aggressively than say the sequencer model. I think like I do agree that people running L2s tend to have a little bit like, weaker testing because they kind of like assume that like they have some invariant about the sequencer. Like it's all in this one data center or whatever.
- [54:40]
Doug Colkitt
Yeah.
- [54:40]
Tarun Chitra
I mean, people in L1 can't assume that.
- [54:43]
Doug Colkitt
Yeah. And like, you did have stuff like, okay, there's an L2 and we have like force inclusion if the sequencer goes down. And like. Right. We did have a case of the sequencer going down. And I don't know if anybody knew how to do a force inclusion transaction. So it's like a lot of this stuff is like theoretical. Until it's not theoretical. And then people are like, oh, shit, where is this in the docs? This isn't well documented. This isn't.
- [55:04]
Haseeb Qureshi
So the docs are going down. That's the, that's how, you know, I mean, it's like Gitbook is a public good.
- [55:10]
Tarun Chitra
We need to like have it nationalized.
- [55:13]
Haseeb Qureshi
Like the blockchains. The blockchains work. The blockchains worked. Everything else on top of it. Tvd. Right. It's kind of like, okay, yeah, every blockchain was technically doing what it was supposed to and the gas limits were going up. I mean, does that count as a failure when Ethereum is $400 a transaction? I mean, Mert was on Twitter being like, oh, well, obviously this means that Ethereum is broken. And I think that argument is very compelling to people is that, yeah, if it costs $400 per transaction, it's broken. That's not how it's supposed to work. I think for a hardcore purist, it's like, well, no, that's the market clearing price for an Ethereum block space. And that's how it works. But I think for users, they're like, no, absolutely no, don't agree.
- [55:58]
Tarun Chitra
Actually, Haseeb, can I reformulate your question slightly?
- [56:01]
Haseeb Qureshi
Yeah.
- [56:02]
Tarun Chitra
Which is I'm going to ask Doug what things are you changing in the perp stacks you are designing based on this? Because I feel like that is a more like, what are things that you will change?
- [56:12]
Doug Colkitt
Yeah. So I mean, I think one is right. Like we run. We run like on chain. So like to. To Haseeb's point, right? The chains didn't really go down. So like a lot of perps, Dexes aren't necessarily on chain. There's on either on a sequencer or, you know, something that's more opaque. So like we're trying to put pretty much every. Everything on chain itself for like all the clearing Two, I think like the insurance stuff, I think you kind of definitely need an insurance fund in. In these systems. And like that goes to your point to run. It's like, it's not great optically when the vault makes 40 million and there's nothing like no insurance fund loss. Yeah, I mean, I'd say those, like those two things are like pretty critical. You know, you don't want downtime and you don't want kind of like unfair. Unfair gains from your ADL and liquidation systems.
- [56:58]
Haseeb Qureshi
Okay, well, we are up on time, Doug. We really appreciate you coming on and educating us about all the microstructure stuff. Very helpful to have you in a moment of crisis. Anything you want to plug or direct people to.
- [57:11]
Doug Colkitt
Yeah, yeah. So we are. We're building. We're building a perps Dex and hopefully next time this happens, it'll be a good place to. To be trading at. It's Ambient Finance. That's on Twitter at Ambient Underscore Finance. And also we're co building with Fogo, which is a new L1 kind of designed to handle this throughput. And that's at Fogochain. Twitter.
- [57:30]
Haseeb Qureshi
Fantastic. Well, we appreciate all the enlightenment that you have brought our way and we hope we. Anybody who is ADL'd, anybody who made money, anybody who lost money. By the way, if you made money, you should tag us in this tweet because I want to know, I want to hear from the half of you. Yes. Don't DM me if you're. Yeah. If you weren't looking for refunds.
- [57:51]
Tarun Chitra
Yeah, exactly. Haseeb is processing Binance refunds for the next two weeks.
- [57:57]
Haseeb Qureshi
Yeah, actually, any exchange, anywhere you need refunds, just hit me up. I'm. I am the reporter. First call. Thank you, everybody. Hope you're all safe and we'll see you all next week. Sam.