The latest episode of Odd Lots' Podcast talks about the crash process of Luna: https://castbox.fm/episode/This-Is-How-the-Terra-Stablecoin-Actually-Imploded-id4460195-id495103603

《This Is How the Terra Stablecoin Actually Imploded》

The collapse of the Terra ecosystem, and the tokens Luna and UST, will go down as one of the most painful and devastating chapters in crypto history. Over $60 billion market value has evaporated, and numerous retail investors are nursing major losses. What's particularly bad is that this was a big project, championed by some of the most notable names in crypto. But some people obviously saw it coming, and understood it to be a disaster in the making. On this episode we speak with Kevin Zhou, the founder of the crypto hedge fund Galois Capital. He began warning about Terra publicly earlier this year, and was short Luna starting in early May. He explains the exact mechanics of the coin's implosion. 

The collapse of the Terra ecosystem and its tokens, Luna and UST, will go down as one of the most painful and destructive chapters in crypto history. Over $60 billion in market value has evaporated, leaving many retail investors with major losses. Particularly bad is that this was a big project championed by some of the most prominent figures in crypto. But some clearly saw it coming and understood it to be an unfolding disaster. In this episode, we interview Kevin Zhou, founder of crypto hedge fund Galois Capital. He began publicly warning about Terra earlier this year and started shorting Luna at the beginning of May. He explains the exact mechanism behind the coin's collapse.

This episode was recorded May 12, 2022.

This episode was recorded on May 12, 2022.

Points of interest in the pod/Podcast theme time points:

• The difference between types of stablecoin / The difference between types of stablecoins - 07:36

• Terra as a perpetual motion machine or Rube Goldberg / Is Terra a perpetual motion machine or a Rube Goldberg device - 09:32

• Where do Terra’s yields come from? / Where do Terra's returns come from - 11:21

• How the Terra/Luna arbitrage mechanism works / How does the arbitrage mechanism of Terra and Luna work - 13:11

• Why did Terra have Bitcoin reserves? / Why does Terra have a Bitcoin reserve?- 18:46

• How did Terra collapse? What was the trigger? / How did Terra collapse?What are the triggers?- 25:17

• The role of the 3Pool/4Pool migration / The role of 3Pool/4Pool migration - 29:22

• Empty

• On reflexivity and Terra/Luna as the ultimate momentum asset /On reflexivity and Terra/Luna as an ultimate momentum asset - 40:50

• On financial contagion in crypto / On financial contagion in crypto - 44:03

• What happens to other stablecoins after Terra? / What happens to other stablecoins after Terra?- 45:24

• Why did big investors get involved with Terra?  / Why big investors got involved in Terra?- 48:51

• Terra and hyperinlation of Luna / Terra and Luna's hyperinflation - 53:53

Joe Weisenthal: (00:00)

Hello, and welcome to another episode of the Odd Lots podcast. I'm Joe Weisenthal

Hi, welcome to another episode of the Odd Lots podcast. I'm Joe Weisenthal.

Tracy Alloway: (00:06)

And I'm Tracy Alloway

I'm Tracy Alloway.

Joe: (00:07)

Tracy. Luna. Terra. Stablecoins. Crypto. Crashing. Stuff.

Tracy, Luna, Terra, stablecoins, Crypto, crashes, all sorts of topics.

Tracy: (00:13)

I'm going to throw something else out there. Roman history. Let's talk about classic history.

I also want to throw something else out there. Roman history. Let's talk about classical history.

Joe: (00:18)

Wait, what? What's the Roman history part?

Wait, what? What’s the Roman history part?

Tracy: (00:20)

It's Rome sacking Carthage. Have you not heard about this?

It's Rome sacking Carthage. Haven't you heard of this?

Joe: (00:23)

Oh, the memes. Is that what's going on? Is Carthage being sacked right now?

Oh, that meme. Is it like this now? Is Carthage being sacked now?

Tracy: (00:28)

Well, that's actually, yeah. I mean, that's kind of what we're going to talk about. So we're going to be speaking with someone who has framed himself as Rome in this historic parallel going after Carthage, which would be Terra and Luna, the algorithmic tablecoin/cryptocurrency.

Well, this actually is, yes. I mean, this is what we're going to talk about. So, we're going to have a conversation with someone who positions themselves as Rome in this historic parallel timeline, going after Carthage, which will be Terra and Luna, the algorithmic stablecoin/cryptocurrency.

Joe: (00:43)

This is such a wild story. I mean, I think, you know, look, crashes, hacks, ponzi, rug pulls. They happen all the time in crypto, right? Like it's almost boring and most of the time, they're not even worth reporting on or talking about because they're like a daily occurrence, but the Terra ecosystem, the UST stablecoin, the Luna token had gotten so big, so valuable. And so many major backers of crypto were invested in it or had invested in it in some way that this is not just like another coin that crashed.

It's such a crazy story: the crash, the hack, the Ponzi scheme, the rug pull. They happen all the time in crypto, right? Like it's almost boring most of the time; they're not even worth reporting or talking about because they're just like everyday occurrences. But the Terra ecosystem, the UST stablecoin, the Luna token had grown so large, so valuable. So many major backers in the crypto world were invested in it or somehow tied to it, so this wasn't just another coin collapse.

Tracy: (01:22)

I think that's absolutely right. So two things here. One: top 10 coin, right? It's not every day that you see something like this happen to a top 10 coin. Secondly, the marketing. So everyone knew it was an unusual experiment, an algorithmic stablecoin and we'll get into exactly how it's supposed to work. But even with that said, it was pitched as a stablecoin, right? As something that is supposed to maintain a peg one to one with the dollar and people are supposed to be able to use it to get in and out of more volatile cryptocurrencies. And clearly that's not what's been happening because we've seen the peg has crashed. I think it got to as low as like 0.3. Is that right?

I think that's exactly right. So, there are two things here. One: top 10 coins, right? You don't see something like this happen every day with a top 10 coin. Two: marketing. So, everyone knows this was an unusual experiment, an algorithmic stablecoin, which we'll discuss how it was supposed to work. But even so, it was marketed as a stablecoin, right? It was supposed to maintain a one-to-one peg with the dollar, and people were supposed to be able to use it to move in and out of more volatile cryptocurrencies. Obviously, that's not what happened, as we've seen the peg break down. I think it has gone as low as 0.3, right?

Joe: (02:07)

Yeah. No, I mean, that's exactly right. So it’s a stablecoin, the interesting thing is like stablecoins. There are many different flavors of stablecoins and people have been skeptical about them for a while. Most of the discussion about vulnerable stablecoins was focused on Tether. But this was a so called algorithmic stablecoin. They had some collateral backing. It's sort of complicated. We're going to get into it. But the point is that yes, it is sort of this classic example of the thing that blows up is the thing that has to maintain a peg, which is a lesson that the TradFi world is understood for a long time, which is often the sorts of big sources of big risk are the things that on the surface appear to have the least volatility.

Yes. No, I mean, that's completely correct. So this is a stablecoin, and interestingly like stablecoins. There are many different flavors of stablecoins, and people have been skeptical of them for a while. Most of the discussion about fragile stablecoins has focused on Tether. But this is so-called algorithmic stablecoin. They have some collateral backing it. It's a bit complicated. We're going to dig into it. But the question is, yes, this is a typical example where the thing that explodes is the thing that's supposed to maintain its peg, which is a lesson that the traditional financial world has understood for a long time: that often the biggest sources of risk are the things that appear to be the least volatile.

Tracy: (02:47)

Here's the other thing that makes this story so interesting. It's the people involved. You know, you mentioned some of the big funds like Galaxy/Novogratz who were invested in Terra/Luna.

Here's another thing that makes this story so interesting. It's the people involved. You mentioned some big funds like Galaxy/Novogratz, they were all invested in Terra/Luna.

Joe: (02:59)

Novogratz has a Luna tattoo.

Novogratz also has a Luna tattoo.

Tracy: (03:01)

Yeah. Which I thought was fake when I first saw that photo, but it is in fact real. But the other thing is there are strong personalities attached on both the pro-Luna Terra side and the sort of against or warning about Terra/Luna's. And that's where that Rome versus Carthage analogy comes in. You had the founder, Do Kwon, and I don't know if anyone follows him on Twitter, but he is incredibly, I guess outspoken would be … ?

Yes. When I first saw that picture, I thought it was fake, but it turns out it's real. But another thing is, there are strong personalities on both sides of the equation—those supporting Luna Terra and those opposing or warning against Terra/Luna. This is where the Rome vs. Carthage metaphor comes in. You have a founder feud, and I don't know if anyone followed him on Twitter, but he was incredible, and I'll just say... outspoken?

Joe: (03:29)

Confident? Cocky might be the word…

Confident? Arrogant might be the word...

Tracy: (03:31)

Yeah, this is putting it all very, very politely, but people would come out with criticisms of the entire algorithmic stablecoin idea and people would highlight vulnerabilities in the system and he would just bat them away.

Yes, it's an extremely polite way of saying it, but people will criticize the whole idea of algorithmic stablecoins, they'll point out flaws in the system, and he just brushes them off.

Joe: (03:45)

Or he would just say, you're poor.

Or he'll just say, you're poor.

Tracy: (03:47)

Yeah you’re poor, I don't need to talk to you.

Yes, you're poor, I don't need to talk to you.

Joe: (03:49)

So what is going on? We are going to be speaking with a perfect guest. When everyone was like Odd Lots has to do an episode on Luna, obviously, and everyone, my DMS flooded with this is the guest you have to have on, because he's been warning about it for a while, we're going to be speaking with Kevin Zhou. He is the co-founder of Galois Capital, which is a crypto hedge fund that was launched in 2018. So that's old by crypto hedge fund longevity, but actually he's been in the space for about a decade, which is truly extraordinary. He knows it well, he had been warning about Terra and Luna for a while. So Kevin, thank you so much for coming on Odd Lots.

So what's going on here? We're going to talk to a perfect guest. When everyone was clamoring for Odd Lots to do an episode about Luna, obviously, everyone, my inbox was flooded with this is the guest you must have, because he has been warning about this for a while now. We're going to talk to Kevin Zhou. He's the co-founder of Galois Capital, which is a crypto hedge fund that launched in 2018. So by crypto hedge fund lifespans, that's old, but actually, he's been around in this space for about ten years, which is pretty unusual. He knows it well, and his warnings about Terra and Luna have been going on for a while. So, Kevin, thank you so much for coming on Odd Lots.

Kevin: (04:29)

Yeah, absolutely. Thanks for having me.

Yes, of course. Thank you for having me.

Joe: (04:31)

Kevin, you started a crypto hedge fund in 2018. Now  everyone has a crypto hedge fund obviously, but that's legit veteran in the space. What’s your background? How did you get into it so much earlier than most people?

Kevin, you started a crypto hedge fund in 2018. Now everyone has a crypto hedge fund, so clearly, you're one of the legitimate old guards in this space. What's your background? How did you get into this earlier than most people?

Kevin: (04:49)

Yeah, so, you know, I think my background, you know, I've been in crypto for a while now, really got started in 2011. Hadn't joined the industry at the time. I was just mostly trading my own PA. And then in 2013 I joined the industry, joined a small Bitcoin startup. It was an exchange, it was called Buttercoin. It was the second YC company that was dealing with crypto, the first being Coinbase, things didn't go so well over there. And in the winter of 2013, we shut down. And then, you know, afterwards I joined Kraken -- ran their trading desk for two years. And then afterwards, I guess early 2017, decided to leave and start Galois Capital. And we launched in January of 2018.

Yeah, I think my background, I've been in crypto for a while now. I really got started in 2011. At that time, I wasn't working in the industry yet. I was just mainly trading on my own PA. Then in 2013, I joined the industry and joined a small Bitcoin startup. It was an exchange called Buttercoin. It was the second YC company to deal with cryptocurrency; the first one was Coinbase, but things didn't go well there. In the winter of 2013, we shut down. Then I joined Kraken and managed their trading desk for two years. After that, I decided to leave and start Galois Capital around early 2017. We launched in January 2018.

Tracy: (05:39)

So as we mentioned in the intro, there's been a lot of criticism of Terra and Luna and you’re, I think probably the biggest voice in that space. What piqued your interest in this particular coin or, you know, system? When did you first start getting interested and start looking into it and why?

So as we mentioned in the introduction, there has been a lot of criticism about Terra and Luna, and you, I think, are probably the loudest voice in this field. What sparked your interest in this particular coin or system? When did you start paying attention to it and why?

Kevin: (06:00)

Yeah so, you know, when I first started looking into it, I'd say it was around mid or late last year and, you know, at the time I didn't think too much of it. I thought, oh, you know, this is just one of those algo stablecoins. It probably won't work out. You know, it'll just implode pretty soon. That turned out not to be the case. It just kept going, you know, Luna kept going up, UST kept maintaining stability. It grew bigger and bigger. And then I revisited it in January of this year. And at this point, you know, I was kind of surprised at just how big it got, you know, all of a sudden, you know, it's now like a top 10 coin or something like that. So now I was starting to get worried because I'd never seen one of these algo stablecoins get so big. Usually they just collapsed before that. So, you know, then I thought, well, maybe this is first, one, maybe a great shorting opportunity and two, maybe this poses some systemic risk to the entire space and maybe it's time to sound the alarm. You know, not, I mean obviously for my own benefit, I would definitely like to make money on this short, but also I think just as a public service to also let everybody else know too.

When I first started paying attention to it, I would say it was around mid to late last year, at that time I didn't think too much about it. I thought it was just one of those algo stablecoins. It might not succeed; it could implode soon. As it turned out, that wasn't the case. It kept growing, Luna kept rising, UST remained stable. It grew bigger and bigger. Then I revisited it in January this year. At this point, I was a bit surprised by how big it had become, suddenly it was like a top 10 coin or something similar. So now I started to worry because I had never seen any of these algo stablecoins get this big. Usually, they collapse before this. Then I thought, well, maybe this is the first time: one, it might be a great shorting opportunity, two, it might pose some systemic risks to the entire space, maybe it's time to sound the alarm. Obviously, for my own benefit, I definitely want to make money on this short, but I also think just as a public service, others should know as well.

Joe: (07:09)

So can you actually back up for listeners and we hear stablecoins and we sort of mentioned it in the intro, which is that there are various flavors of stablecoins. Tether is one model, USDC is a similar model. There's also Maker and Dai, but what is an algo stablecoin, this project they keep trying and they do have a long history of continuing to blow up, how is an algo stablecoin different than other stablecoins?

So can you enlighten our audience? We hear about stablecoins, and we mentioned in the introduction that stablecoins come in various styles. Tether is one model, USDC is a similar model. There's also Maker and Dai, but what is an algo stablecoin, the kind of project they've been trying, and they do have a long history of repeated liquidations. How are algo stablecoins different from other stablecoins?

Kevin: (07:36)

Yeah, so I think we should first separate out these so called decentralized stablecoins, with the centralized ones. So, you know, first on the centralized side you have folks like USDT, which is Tether, and USDC, which is, you know, Circle's dollar. And basically what's happening here is that there's a dollar in some bank account and for every dollar there, they issue out one coin. Anytime you return a coin to them, you can redeem out a physical dollar, you know, by wire transfer or whatnot. So, you know, it's very much kind of tethered together between the real world and the virtual world. You know, on the sort of decentralized side, you really have variations of two models, right?

Yes, so I think we should first separate these so-called decentralized stablecoins from centralized stablecoins. First, in the centralized aspect, there are ones like USDT, which is Tether, and USDC, which is Circle's dollar. Basically, what happens here is that for every dollar in some bank account, they issue a coin. Anytime you return the coin to them, you can redeem a physical dollar, via wire transfer or other means. So, it largely ties the real world with the virtual world. In the decentralized aspect, you really have two modes of change, right?

So you have sort of like the collateralized model, which is like Make or Dai. What that is, is basically you can post some amount of some other type of asset and in return, you're going to get some stablecoin which is maybe, you know, like if you overcollateralize it by, you know, 125%, then you're going to get a hundred percent. So if you put $1.25 worth of value, you get back a dollar. And then at some point the price of the underlying collateral will fluctuate to the point at which you reach like a point where there's a margin call, and then, you know, either your collateral gets seized or you top it up, or you return the borrow, right? So it's basically collateralized lending. That sort of model. And then finally, and this is where, you know, we start talking about Luna and UST, you know, finally there is this pure class of algo stablecoins, which do not have collateral or are severely undercollateralized. And they have some kind of stabilization mechanism, long story short, it's basically like a perpetual motion machine .

So you have a mode like collateralized, which is like Maker or Dai. This is where basically you can post a certain amount of some other type of assets, and in return, you get some stablecoins. For example, if you over-collateralize it, say 125%, then you get 100% back. Therefore, if you put in $1.25 worth, you get $1 back. Then at some point, the price of the related collateral will fluctuate to a point where there is a margin call, either your collateral gets liquidated, or you top it up, or you pay back the loan, right? So it's basically a collateralized loan. This kind of model. Finally, this is where we start discussing Luna and UST, finally there is a class of purely algo stablecoins, which have no collateral or are severely under-collateralized. They have some stabilization mechanism, to cut a long story short, it's basically like a perpetual motion machine.

Tracy: (09:25)

That’s not the first time that exact phrase has been used on this podcast in reference to Terra/Luna.

This is not the first time this exact phrase has been used when mentioning Terra/Luna on this podcast.

Kevin: (09:32)

Yeah, definitely. I would say either perpetual motion machine or giant Rube Goldberg machine. Right? So some of these are just really elaborate contraptions and, you know, you don't know where the hand crank is, but someone's turning a hand crank to keep the system going. And it's not actually a perpetual motion machine, which we know is impossible. So that's the analogy, but basically the idea is that you have some kind of mechanism, which, in some way, you know, indirectly or directly expands the supply of that stablecoin, when the price is too high in order to push it back down. And on the other side contracts the supply of that stablecoin through some mechanism when the price is too low, hence pushing the price back up. Right? So these are kind of like these, you know, there are these feedback mechanisms, regardless of how you design it within this class of stablecoin, there are all these feedback mechanisms, which eventually caused something like this to happen. Which is why in the end. I think they're all pretty much the same, but you know, just through different methods, I mean, they all look different, but when you really start to really break it down, it's all just about supply contraction and expansion.

Yes, certainly. I would say, either it's a perpetual motion machine, or it's a huge Rube Goldberg machine. Right? So, some of these are just very complicated contraptions where you don't know where the crank is, but someone is turning the crank to keep the system running. And it's actually not a perpetual motion machine, we know that's impossible. So, it's a metaphor, but the basic idea is that you have some mechanism, somehow, indirectly or directly expanding the supply of the stablecoin when the price is too high, to push it back down. On the other hand, contracting the supply of that stablecoin when the price is too low, through some mechanism, to push the price back up. Right? So these are kind of like these, having these feedback mechanisms, regardless of how you design within this category of stablecoins, these feedback mechanisms eventually lead to such things happening. That's why in the end, I think they're all pretty much the same, just through different methods, they look different, but when you really start breaking it down, it's all just about supply contraction and expansion.

Tracy: (10:46)

Two things here. One, could you give us a little bit more detail on that arbitrage mechanism between Terra and Luna, like walk us through exactly how it works. If the price of Luna goes up or the price of Luna goes down and, you know, vice versa with Terra. And then secondly, you mentioned a hand kind of turning the crank. And one of my understandings here is that the yield on offer from Terra/Luna were exceptionally high. I think something like 19% or 20%, and this is another question that we ask a lot in the crypto or DeFi space more broadly, but where do those yields actually come from? Like how does this entire incentivization mechanism work?

There are two things here. First, could you give us more details about the arbitrage mechanism between Terra and Luna, like let's see exactly how it works. If the price of Luna rises or falls, so does Terra. Second, you mentioned there is a hand turning the crank. One of my understandings is that the yield offered by Terra/Luna is particularly high. I think it was around 19% or 20%, which also leads to another question we ask more broadly in the cryptocurrency or DeFi space, but where do these yields come from? Like how does this entire incentive mechanism work?

Kevin: (11:21)

Yeah, definitely. And I completely agree with you that I think this component of it, which is the anchor yields on deposits for UST, this 19.5%, which then eventually dropped down to 18%. That's exactly the hand crank in this perpetual motion machine that makes it not perpetual. And it's a great question, always, I think, to ask where do the yields come from? I think a lot of times, you know, it just seems like there's free money, but it turns out you're just subsuming all of this kind of indirect or hidden risk that you're not aware of. A lot of this kind of like tail risk, which, you know, only manifests once in blue moon, but when it does it completely wipes you out. So it's just really kind of like very dangerous kind of invisible risk.

Yes, sure. I totally agree with you. I think this part of it, the deposit anchor yield for UST, which was previously 19.5% and eventually dropped to 18%. This is exactly the hand crank in this perpetual motion machine that makes it not perpetual. It's a great question, I think, always to ask where does the yield come from? I think a lot of times, it just seems like there's free money, but it turns out you're just loading up on all these indirect or hidden risks that you don't realize. A lot of this tail risk kind of stuff only happens once in a blue moon, but when it does, it completely wipes you out. So it's really a very dangerous hidden risk.

What I always like to say is that most of the time, you know, if you can't find where the yield is coming from, then effectively it's coming from future bag holders. So it's like this idea that, you know, the sort of true belief and sort of the cultish behavior of true believers will eventually produce some value, which gets extracted from them and is given to you now in the present. So it's extracted in the future and given to you in the present right now. And I think that's basically exactly what's going on  with this Luna model. I know that's a little bit abstract, but, you know, that's kind of I see it. And then maybe just returning to your first question, could you remind me again what the first question?

What I always want to say is that most of the time, if you can't find where the yield comes from, then actually it comes from future bag holders. So it's like this idea, this true faith and true believer kind of cult behavior will eventually generate some value, which is extracted from them and given to you now. Therefore, it is extracted in the future and given to you now. I think this is basically exactly what happened in the Luna model. I know it's a bit abstract, but that's the situation I see. Then maybe just go back to your first question, could you remind me again what the first question was?

Tracy: (12:51)

Oh yeah. Can you explain exactly how the arbitrage mechanism between Luna and Terra or UST actually works? Like walk USThrough what is happening as the prices of either of these things move, like how they balance out. Or how they're supposed to balance out, I should say.

Can you explain how the arbitrage mechanism between Luna and Terra or UST works exactly? Like walk us through UST, what happens when the prices of these things move, like how they are balanced. Or how they should be balanced, I should say.

Kevin: (13:11)

Yeah. So I'll first sort of just describe the simple model and then I'll add some caveats on some of the intricacies there. So the simple model is that at any point you can always redeem one UST for $1 worth of Luna, right? So like if Luna was at a hundred dollars, then one UST would get you one Luna penny. Point zero one Luna, which is equivalent to $1. And if Luna is $1, then you know, one UST is going to give you one Luna. But in any case, you're always able to redeem UST for the equivalent amount of dollars worth of Luna. And the amount of Luna you get is based on where the current market value for Luna is, and the opposite is true too. You can always destroy or, or burn one dollar's worth of Luna to create one UST, right?

I'll first describe a simple model, then I'll add some notes about some complications. So the simple mode is that at any time, you can always exchange one UST for one dollar worth of Luna. So if the price of Luna is $100, then one UST can get you one cent worth of Luna. If Luna is $1, then one UST will give you one Luna. But in any case, you can always exchange UST for an equivalent amount of Luna dollars. And the amount of Luna you get is based on Luna's current market value, and vice versa. You can always burn or destroy one dollar worth of Luna to create one UST.

So the idea behind this feedback mechanism is that if UST is trading below $1, what you can do is you can buy that on the open market. And then convert that to one dollar's worth of Luna and then sell that $1 of Luna. Right? So that's kind of how that arbitrage mechanism works in order to maintain this peg. So that's basically the simple model. Now, it gets a little bit more complicated because there's a lot of dials and knobs that you can adjust, and, you know, bells and whistles around this. So the first thing is that it's not technically true, that you get exactly $1 worth of Luna for burning one UST. There is basically something called an automated market maker. This is something that has been popularized by, you know, UniSwap, SushiSwap, Curve, Balancer, a lot of these, you know, AMM protocols, these sort of decentralized liquidity pools, and it's really this automated market maker or bonding Curve, which governs sort of the slippage or the costs of converting between Luna and UST.

So the idea behind this feedback mechanism is that if UST trades below a dollar, what you can do is buy it on the open market. Then convert it into one dollar worth of Luna, and then sell that one dollar worth of Luna. Right? So that's how this arbitrage mechanism works to maintain the peg. So that's basically the simple model. Now, it gets a little complicated because there are many dials and knobs you can adjust, and bells and whistles around this. So the first thing is, technically speaking, it's not true that burning one UST gives you exactly one dollar worth of Luna. There's basically something called Automated Market Maker. This is popularized by AMM protocols like UniSwap, SushiSwap, Curve, Balancer, etc., which indeed controls the slippage or conversion cost between Luna and UST.

So the first component of this mechanism is that the greater the size that you do, the worse of a price you're going to get, right? So this is like equivalent to market slippage. You know, if you want to buy, you know, a million dollars’ worth of something, you know, you're going to pay a little above market price. If you want to buy a billion dollars’ worth of something, you're going to pay well above market price. So it just, you know, it kind of depends on liquidity in the market, but really it's governed by this bonding Curve and, you know, the greater the size that you do, the more slippage you incur. The second part of it, that's a little bit nuanced, is that there is some fee that's collected in the middle, just like all these other bonding Curves, all these other AMMS, there is a fee for converting, it's not very high, but, you know, there is still some fee which adds onto basically the transactional costs, which are the slippage, you know, plus the fee itself.

So the first component of this mechanism is that the larger the scale of what you do, the worse the price you will get, so it's like slippage in the market. If you want to buy something worth a million dollars, you'll pay a little more than the market price. If you want to buy something worth a billion dollars, you'll pay much higher than the market price. So it kind of depends on the liquidity of the market, but in fact it is controlled by this bond curve; the larger the scale of what you do, the more slippage you generate. The second part, there's a subtle difference, which is taking some fees in the middle, just like all these other bonded curves, all these other automated market makers, there's a conversion fee, not very high, but still some fees, which basically add to the trading cost. This is slippage plus the fee itself.

And then the last part that's really interesting about this mechanism is that, you know, before the full collapse, you know, in the past couple of days, there was basically a gating mechanism, which governed exactly how much UST could be created or destroyed per day. So being destroyed being, you know, moving back into Luna. And I think that that number was around like 250 million dollars’ worth per day. So anything beyond that, you just, you're just going to have to wait until the next day. And then, you know, there's all also all sorts of like, you know, much finer details, like pool recovery periods and stuff like that. But, you know, none of that is really that important, I think, to what we're going to be talking about next. So I think just based on that explanation, I think we should be able to describe the phenomenon that we've seen in the past few days.

Finally, the very interesting part of this mechanism is that, before the total collapse, there was basically a threshold mechanism in the past few days that exactly controlled how much UST could be created or destroyed each day. So, what was being destroyed was moving back to Luna. I think this number was about $250 million per day. Therefore, anything beyond this figure, you just had to wait until the next day. Then there were various finer details, like pool recovery periods and similar things. But these weren't really that important, I think, for the issues we're going to discuss next. So I think just based on this explanation, I think we should be able to describe the phenomena we've seen in the past few days.

Joe: (16:57)

So before we get into sort of like what happened over the last few days, because I think you've said some important things. The fact that there's a limit on how much can be converted in a given day is really important to understanding the action. The fact that there's a fee, obviously for the more you trade, which if everyone's trading, fees go up. The one other element I'm curious about in understanding its role. Actually two, I want to understand a little bit further the Luna Foundation bought a ton of Bitcoin, I think earlier in the year. And the basic idea was like, okay, you know, just in case, I guess, the perpetual motion machine starts to wobble a little bit, we can defend the peg with this big Bitcoin reserve -- sort of like classic EM-style currency peg.

So before we dive into what happened in the past few days, because I think you've already mentioned something important. The fact that there is a limit to how much can be converted in a given day is crucial for understanding the dynamics. And the fact that there is a fee, obviously the more you trade, if everyone is trading, the fee will increase. I'm curious about another factor in understanding its role. Actually, there are two factors I want to further understand: the Luna Foundation bought a lot of Bitcoin, I believe it was earlier this year. The basic idea was, well, in case the perpetual motion machine starts to wobble a bit, we can use this large Bitcoin reserve to defend the peg -- a bit like a classic EM-style currency peg.

Tracy: (17:44)

I found this move really weird. Because either you're sort of a fully backed stablecoin with traditional reserves or you're an algorithmic stablecoin. Because the whole, the whole problem they were trying to solve was that they wanted to get away from the traditional financial system, without having to build the reserves weird. It's sort of like being a little bit pregnant or something.

I found this move really strange. Because either you're a fully-backed stablecoin with traditional reserves, or you're an algorithmic stablecoin. The whole point is, they wanted to solve the problem of getting rid of the traditional financial system without needing the weirdness of building reserves. It's a bit like being somewhat pregnant.

Joe: (18:07)

A little bit unbacked.

It's a bit unreliable.

Tracy: (18:11)

Yeah, like we're sort of reserved-back, but not really.

Yes, it's like we kind of have a reserve for backup, but not really.

Joe: (18:13)

So what was the role? Well, A) I want to know, what is the role of the Bitcoin stabilization fund? And then also, I know you touched on it, but could you just explain a little bit and Tracy asked, but I think it's crucial -- that 20% yield that was being given to induce people to hold UST -- the sort of big reason to hold the stablecoin and actually make money holding a stable, where was that coming from? Like, what was the source of funds specifically, other than just sort of theoretical future bag holders? How was that paid out?

So what about the role? Hmm, A) I want to know, what is the function of the Bitcoin Stability Fund? Then, I know you mentioned this issue, but can you explain Tracy's question, but I think it's crucial -- that 20% yield is used to induce people to hold UST -- the important reason for holding stablecoins and actually making money from holding them, where does that come from? For example, apart from theoretical future bag holders, what is the specific source of funds? How is it paid?

Kevin: (18:46)

So basically, you know, when Luna, or the Terra ecosystem first got started, there were some funds that were set aside for the company itself. Right. Right. And the main company is Terraform labs, TFL, and they have this huge stash of Luna, which unlocks over a certain investing schedule. So even for them, you know, slowly unlocks over time, so what they would do in order to finance their operations and to also finance the anchor yield reserve, is they would sell large clips of this to, you know, willing investors at some kind of discount that also has a one year cliff or some kind of vesting schedule, something like that. And then they would use that for operations. And they would also use that to keep basically topping up the anchor protocol on their yield reserve. Because they were paying more interest to depositors than they were collecting from borrowers. And, you know, I think in the end stages of Luna in its final days, you could see that the, you know, the deposit amount was way, way higher than the borrowed amount. So, you know, they, they were bleeding. I mean, I think at some point…

So basically, when Luna or the Terra ecosystem first started, there were some funds reserved for the company itself. The main company is Terraform Labs, TFL, they had a large amount of Luna reserves that would unlock on a certain investment schedule. Therefore, even for them, it was about unlocking slowly over time. What they would do to fund their operations and finance the anchor yield reserve is that they would sell these things at a discount to willing investors, also with a one-year cliff or some sort of vesting schedule, something like that. Then they would use that money for operations. They would also use this money to maintain the yield reserve of their anchor protocol because they pay more interest to depositors than they charge borrowers. And, I think in the final stages of Luna, in its last days, you could see that the deposit amounts far exceeded the loan amounts (insolvency). So, they were bleeding, especially at certain times...

Joe: (19:51)

So everyone buys UST in order to collect that 20%, but that basically has the effect, if you think it through, of sapping those reserves fund faster, they're sort of like set aside to sort of bootstrap the whole thing and incentivize the whole thing. But essentially if everyone is chasing for it at once that starts to get depleted, or you start to like strain your ability to pay that up.

Therefore, everyone buys UST to collect that 20% interest, but this essentially has an effect, if you think about it carefully, of consuming these reserve funds faster, which were somewhat reserved to bootstrap the whole thing, incentivize the whole thing. But essentially, if everyone chases it all at once, it will start to be depleted, or you start to feel the strain on your ability to pay these.

Kevin: (20:14)

Yeah, that's exactly right. I think at the peak, they were burning maybe about 7 million dollars a day worth of their yield reserve. And, you know, originally I think it was something like 50 mil or 80 mil or something like that. And then they had to do a top up of 450 mil. And then, you know, very quickly soon after that soon was almost depleted and they were thinking about how much to do another top up, whether, you know, they were thinking about whether to do it or not. Everybody was, you know, lobbying Do Kwon to do it. You know, he was, running some rumors that, oh, maybe it's going to be over a billion this time, this and that. So, you know, the whole thing was very expensive now.

Yes, that is absolutely correct. I think at the peak, they might have been burning about $7 million worth of yield reserves per day. And, initially, I think it was around 50 million or 80 million or something like that. Then they had to do a $450 million recharge. Then, it almost ran out quickly, and they were considering how much more to do, whether they should do it. Everyone was lobbying Do Kwon (Luna founder) to do this. He was, spreading some rumors, maybe this time it would exceed $1 billion, this and that. So, the whole thing is now very expensive.

I think what they were thinking -- because I don't want to strawman them either -- I think what they were thinking is that, you know, they just want to, you know, they think about this as a marketing expense, right. So they just want to get to get everybody talking about Luna, everybody using US, you know, they're just bleeding 7 million in a day, but they're getting a lot of people, you know, talking about Anchor, using Anchor, you know, using their ecosystem, putting their money into it, into this borrow lending protocol. So, you know, I personally don't think it was worth it. I think they, you know, probably even if they were right, which I don't think they were, I think they're absolutely wrong, but even if they were right, they probably could have gotten away with a little bit lower than 20% yields. I mean, probably like even like 18%, 17%, probably would've been fine. Now, it's not going to save them that much more money, but I do think they overpaid for that. You know, just a random thought there.

I think what they were thinking is (because I don't want to be too harsh on them either) that they're treating this as marketing expense, they just wanted to get everyone talking about Luna, everyone using UST, and they're just bleeding seven million a day, but they're getting a lot of people talking about Anchor, using Anchor, using their ecosystem, putting their money into it, into this lending protocol. So personally, I don't think it's worth it. I think even if they are right, and I don't think they're right, I think they're absolutely wrong, but even if they're right, they probably got a little bit less than the 20% yield. I mean, probably like even like 18%, 17%, might have been pretty good. Now, this wouldn't have saved them more money, but I do think they paid too much for it. This is just a random thought.

Tracy: (21:45)

And what about, what do you think was the rationale for the Bitcoin reserves? Because I think before this week happened, I think they ended up with something like $3 billion worth of Bitcoin that they'd accumulated, I think that's right.

So what do you think the rationale was behind the Bitcoin reserve? Because I think before this week happened, I think they ended up with three billion dollars worth of Bitcoin, I think that's right.

Kevin: (21:59)

Yeah. Yeah, that's about right. So, you know, I think this , this whole purchasing of the Bitcoin was in my opinion, a great move in some ways, and in some ways a really bad move, right. So it was a really good move, I think, in terms of trying to make the system solvent eventually. They probably needed about actually about $10 billion worth of collateral.

Yeah. That's right. So I think the whole buying Bitcoin thing, in my opinion, was in some ways a great move, in some ways a really bad move; overall, I think it was a very good move in trying to make the system eventually solvent. They probably needed about ten billion dollars worth of collateral.

Tracy: (22:26)

Only $7 billion, more.

Only seven billion, more.

Kevin: (22:28)

Only $7 billion more, you know, and it's, I mean, easier said than done definitely. But you know, at least it was, I think, a step in the right direction in terms of creating solvency for the system. I think the system was way in the past, it was already insolvent, you know, it's just that nobody realized because they had created such a strong supply sink in Anchor for this UST, you know, if that disappeared overnight, or even gradually, the entire system was insolvent. I mean, even at a price, I would say of over a hundred dollars a Luna, this whole thing was already insolvent, it's just, nobody realized it. But anyway, returning back to the Bitcoin, so I think in that sense, it was really good that they bought the Bitcoin. But in some ways it was a little bit bad because it kind of destroyed the narrative a little bit. Right.

Only seven billion, easier said than done for sure. But at least it was a step in the right direction in terms of creating solvency for the system. I think the way the system was going before, it was already insolvent, it's just that no one realized it because they were creating such a strong supply sink for this UST through the anchor, that if it went away overnight, or even gradually, the whole system would be insolvent. Even at a price, I would say a hundred dollar plus Luna, this whole thing was broke, it was just no one realized it. But anyway, back to Bitcoin, so I think in that sense, them buying Bitcoin was actually pretty good. But in some ways, it was kind of bad because it kind of breaks the narrative.

Joe: (23:11)

As Tracy was saying.

As Tracy said.

Kevin: (23:14)

Now it's just like a kind of half-backed, a little bit half*ss-backed, you know, kind of a, I don't know if I could say that on the podcast by the way.

Now it's just like half-supported, half-not-supported, I don't know if I can say this in the podcast.

Tracy: (23:21)

It's fine.

That's good.

Kevin: (23:21)

But, you know, now it's sort of destroying their own narrative because they basically said that, oh, we, we finally constructed this perpetual motion machine, you know, behold, everybody, we finally did it, you know, and it's working and it's amazing, but actually, you know, just in case it doesn't work, let's just get some insurance, you know? So like before, the narrative was very strong, it was just like, there was just saying, oh yeah, for sure this thing works. Now obviously it doesn't, but you know, they were saying, and a lot of people were believing them, but you know, it's almost like they were capitulating a little bit on the narrative and making certain concessions by even buying the Bitcoin in the first place, because now it begets the question. Now people were thinking, well, wait a second, if this thing is just always, it was going to work all along, why do we even need that? And you're telling me that, you know, that's not really a vote of confidence there, you know, that you need, you even feel the need to get some insurance there. But you know, all that being said still happy that they did. I mean, it would've been even worse if they didn’t.

But now it kind of undermines their own narrative a bit because they basically said, we finally built this perpetual motion machine, look guys, we finally did it, it's working, it's amazing, but just in case it doesn't work, let's just get some insurance. So like before, the narrative was very strong, like, some people were just saying, oh yeah, for sure this thing works. Now obviously it's not that way, a lot of people believed them, it's almost like there's a bit of surrender in their narrative, even making certain concessions by first buying Bitcoin because now it raises questions. Now people are thinking, well, wait a minute, if this thing has always been this way and it's always going to work, why do we need this? You tell me, this isn't really a confidence vote, you need, you even feel the need to get some insurance there. But all these things still, they're happy they did it. I mean, if they didn't, it would be worse.

Tracy: (24:30)

So walk us through what you think just happened. Because I've seen some wild theories out there. I mean, a lot of the Luna bulls, I guess they sometimes call themselves Lunatics, but a lot of them are talking about this idea of a concerted attack from people like Citadel and even BlackRock, it just sounds crazy to me. But also here's the thing, even if it was some sort of concerted attack, a bunch of people shorting all at once, that still seems like a fundamental vulnerability in this machine that you've designed to be stable. So I guess walk us through what you think happened and what the exact trigger was for the chaos that we've seen this week. The depegging.

So let's see what you think just happened because I saw some crazy theories out there. I mean, a lot of Luna bulls, I guess they sometimes call themselves Lunatics, but a lot of them were talking about this idea of a coordinated attack from people like Citadel or even BlackRock, which sounds crazy to me. But the question is, even if it was some sort of coordinated attack, a group of people shorting at the same time, it still seems like a fundamental flaw in the machine that you designed, right? It's supposed to be stable. So I'd like you to explain to us what you think happened and the exact trigger for the chaos we saw this week.

Kevin: (25:17)

Yeah. You know, I think we're really entering this phase in the market cycle, which I I've been calling on Twitter, the finger pointing phase. You know, now that this thing is clearly failed, who should take the blame for this, right? Is it A) some external party in a foreign world, the TradFi world, right? The boomers and the suits that nobody likes in crypto. Yeah. Easy scapegoat, right. Probably was Citadel or BlackRock, right? Turned out to be just a post on 4Chan, you know, like, I mean…

I think we're really in this phase of the market cycle that I've been calling on Twitter the blame phase. Now that this has obviously failed, who gets blamed for it, right? Is it A) some outside party to the Crypto world, the TradFi world, right? Those guys in suits that nobody in crypto liked during the boom; they're an easy scapegoat, maybe Citadel or BlackRock, right? It turns out just a post on 4Chan.

Joe: (25:49)

We're in historic times that BlackRock and Citadel now I have to issue rebuttals to a 4Chan post, but this is 2022.

We are in historic times where now BlackRock and Citadel have to issue rebuttals to a 4Chan post, but it's 2022.

Kevin: (25:57)

This is absolutely ridiculous. I mean, it just goes to show how quickly memes spread and how quickly narrative spreads, whether it's right or wrong we’re in the age of social media and word spread so fast that literally a 4Chan post can trigger a response from Citadel and BlackRock. Iit's absolutely ridiculous. I mean, people coming up with these, you know, conspiracies. So is it A) them, right? These outsiders, these boomer TradFi guys. Is it B) somebody internal to crypto, you know, some big shorters like, you know, like Galois Capital or like some of the other noted, you know, detractos of Luna. Was it Sam, you know, was it Wintermute, who knows? Maybe it was one of our own people. Or is it C) is it Do Kwon’s fault? You know because he built this project, the mechanism was clearly unsound. Maybe he's a grifter. Maybe he was a scammer all along. Maybe it's his fault or is it D) is it like people's own fault in a way, right? Is it that they themselves got a bit greedy? They wanted so badly to believe that there was free money raining from the sky that they turned off all their reason and logic because they, you know, they had some hope of changing the circumstance of their life. You know, it's a bit sad to say. But I think, you know, there is some bit of that too, you know, so what I would say is, really hard to say what, you know, exactly triggered it, but I think it would, you know, if this is any lesson for us and I, I think, you know, the crypto space, we've always been kind of like we want to be a self-regulating industry.

This is absolutely absurd. It just shows how fast the memo spreads, how quickly narratives spread, whether they're right or wrong. We're in the social media age, and messages spread so fast that a literal 4Chan post can prompt responses from Citadel and BlackRock. This is absolutely absurd. People come up with these conspiracies. So:

• Citadel and BlackRock, right? These outsiders, these TradFi people.

• Someone inside crypto, some big shorters, like Galois Capital or like some other famous ones, Luna detractors. Is it Sam, is it Wintermute, who knows? Maybe it's one of our own.

• Is it Do Kwon's fault? Because he built this project, and the mechanism was clearly unsound. Maybe he was a scammer. Maybe he has always been a scammer. Maybe it's his fault,

• in some ways it's the people's own fault, right? Was it that they were a little greedy themselves? They really wanted to believe in free money falling from the sky, so they turned off all their rationality and logic because they had some hope of changing their living environment.

It sounds a bit sad. But I think it's really hard to say what exactly triggered it, but I think if there is any lesson for us, in the cryptocurrency space, we have always been a bit like we want to be a self-regulating industry.

We don't really need regulators to step in, you know, protect the little guy, protect retail from themselves. We can handle things on our own. And if that is the case, then every time something like this happens, we have to take some very good lessons. We have to have to, you know, take a hard look at ourselves on why this happened. You know, how did people get so greedy and not just retail themselves too. I think it is good for them to reflect on themselves too, but also, you know, how did the VCs, how did the investors get so greedy about this stuff? How did the founders get so greedy about this stuff? You know, the exchanges, I mean, to some extent, I mean, they definitely benefit from all the trading volume, right. So, you know, maybe it's not that, you know, they're particularly greedy, but maybe they look the other way because they didn't really care as long as you could trade a coin back and forth and collect the fees.  Maybe it was still good for them, right?

We really don't need regulators to intervene, to protect the little guy, to protect retail from itself. We can handle things ourselves. If this is the case, then every time something like this happens, we must learn some very good lessons. We must seriously examine ourselves as to why this happened. How did people become so greedy, not just retail. I think it's good for them to reflect on themselves, and how could venture capital firms, investors be so greedy about these things? How could founders be so greedy about these things? Exchanges surely benefited from all the trading volume. They were particularly greedy, but maybe they looked at it another way because they didn't really care, as long as you can trade a coin back and forth and charge fees. Maybe it still benefits them, right?

So, you know, I think, it is important for us to reflect on ourselves before we just start pointing fingers. Now that being said, if we want to talk about how, you know, it actually went down. There's been, you know, a lot of speculation and some people have woven these very intricate narratives about, okay, on this time, this thing happened, this time, that thing happened. What I would say is that mostly it’s speculation. There are some things that are, I think, factual. So maybe we can just start with that, which is that when Luna first started to unwind, this was during a period, you know, when the first depegging happened, this was a couple days ago…

I think, before we start pointing fingers, we must reflect on ourselves. Since then, if we want to talk about how it actually went down. There's a lot of speculation, some people wove these very complex narratives, okay, at this time, this thing happened, at this time, that thing happened. What I'm saying is, most of it is speculation. There are some things, I think, that are factual. So maybe we can start from that point, which is when Luna first started to come undone, it was during a period when the first unraveling happened, it was a few days ago...

Joe: (28:57)

Before you keep moving, we're recording this Thursday, May 12th. And so it was really, I feel like last Saturday morning, I guess that was maybe the 7th or something. But that was sort of when it suddenly started to deviate, I think it was about roughly five or six days ago. But anyway, sorry, keep going. I just want to make sure listeners sort of understand the timeframe here.

We recorded this on Thursday, May 12th, before you continue. So it was really, I think last Saturday morning, I guess that might have been the 7th or something. But that’s when it started to kind of deviate, I think maybe like five or six days ago. Anyway, sorry, go ahead. I just wanted to make sure the listeners understand the timeline here.

Kevin: (29:22)

Yeah, exactly. So it was around that time and it was during a time when there was a migration of assets from the 3Pool on Curve to the new 4pool. Right. So maybe I'll explain that a little bit. So I guess starting with Curve. Curve is basically, it's a little bit like Uniswap, it's a bonding Curve, AMM, it's a bit flatter in terms of the bonding curve, meaning that you can trade greater size near ‘the peg,’ you know, and this is mostly for stablecoins, right? So, you know, for 3Pool, it's like, you have USDC, USDT, Dai, and those are the three coins in 3Pool and they're all stablecoins. They should all be pegged roughly to a dollar. So, you know, around a dollar, you can trade huge amounts of size, which is not quite doable in, you know, rounder bonding curve, like, you know, most things on Uniswap.

Yes, exactly. So it happened right around the time when assets were being moved from Curve's 3Pool to the new 4pool. Maybe I should explain a bit more. Let me start with Curve. Curve is basically like Uniswap but with a flatter bonding curve, meaning you can trade larger sizes near the 'peg'. It's mainly for stablecoins. For 3Pool, you have USDC, USDT, and Dai, which are the three coins in 3Pool, all of which are stablecoins. They should all be roughly pegged to one dollar. So around one dollar, you can trade huge amounts, which isn't as feasible with steeper bond curves, like most things on Uniswap.

Joe: (30:19)

So it's an AMM, it's a DeFi AMM, but it's particularly well optimized for trading stables between each other.

So it's an AMM, a DeFi AMM, but it's specifically optimized for stablecoin exchanges.

Kevin: (30:26)

Yes. I mean, there's some caveats there, but you know, I think the details don't matter too much. It is particularly optimized for trading stables for the most part. So that's 3Pool, 3Pool is, you know, Circle’s dollar, Tether and DAI, which is Maker's coin. And then 4Pool would've been two of the coins from 3Pool, being Tether and Circle dollar. But instead of having DAI, they were going to have UST and they were going to have USTT and they were going to have Frax’s dollar. So, you know, the whole idea there is that they kind of wanted to kill Dai and they wanted to have more of their own native liquidity. They didn't want to just have a meta pool, which is, you know, for example UST against three curve, right? So it's a pool between two assets, but one of the assets is pool itself.

There are some caveats there, but I think the details aren’t too important. In most cases, it's specially optimized for stablecoins. That's what 3Pool is—Circle dollars, Tether, and DAI (Maker's stablecoin). Then 4Pool will consist of two coins from 3Pool, namely Tether and Circle dollars. Instead of DAI, they’ll have UST, they’ll have FRAX (Frax's dollar). So the whole idea is that they kind of want to phase out Dai and create more native liquidity for themselves. They don’t just want to have a meta pool, like, say, UST against three pools on Curve, right? So it’s a pool between two assets, but one of those assets is itself a pool.

So they didn't want to just be attached as like a side car, you know, to 3Pool. They wanted to be, you know, have their own native pool. So these great plans for, you know, 4Pool and whatnot. So, you know, during this time in the depegging, you know, returning to that point in the story, they were basically the Luna guys -- TFL -- they were doing a migration from 3Pool to 4Pool. So they're pulling all the liquidity from 3Pool putting into 4Pool. And this is basically when somebody just took up all the liquidity left on UST by basically just, in the UST pools, just by dumping all their UST and taking out all the other assets. And this basically was what caused the first panic. Now was that one person, was that multiple people? I think it’s really hard to say, I think you'd have to take a look at the chain.

So they didn’t want to just be a subsidiary vehicle for 3Pool. They wanted their own native pool. Hence these grand plans for 4Pool and all that. So during this period of devaluation. Back to that point in the story, they were essentially the Luna guys - TFL - doing a migration from 3Pool to 4Pool. So they pulled all the liquidity from 3Pool into 4Pool. This is basically when someone just drained all the liquidity left in UST, basically just by dumping all their UST and taking out all the other assets in the UST pool. This is what caused the initial panic. Now, was it one person, multiple people? I think that’s really hard to say. You’d have to look at the chain.

Right now, the narrative is that it was just one attacker, but I mean, I think this is kind of like boogieman. I think we should actually take a look at the actual transactions that had happened on Curve at that time, and see if it came from multiple addresses or came from a single address. I mean, to be fair, even if it comes from multiple addresses, it could still be the same party. Maybe they just split up their wallets, you know, but at least we should verify, right before making these kinds of speculations. But in any case, basically all the UST was dumped. All the other assets were drained from these pools. And then that basically caused a little bit of a panic. Other people pulled money out of Anchor. People tried to find ways of getting rid of their UST. Luna started tanking. The entire markets were already tanking. You know, it was kind of like an alignment of the stars. The equity markets were tanking and crypto’s really correlated with the equity markets these days. So everything was dropping and on top of that, the migration was happening. So it was, you know, a cacophony of the perfect sequence of events and, you know, coincidences there.

Now, people are saying that it was just one attacker, but I mean, I think it's a bit like the boogieman. I think we should actually look at the actual transactions that happened on Curve at the time and see if they came from multiple addresses or from one address. I mean, to be fair, even if it came from multiple addresses, it could still be the same party. Maybe they just split up the wallet, but at least we should verify before making these types of guesses, right. Anyway, basically all UST was dumped. All other assets were drained from these pools. Then this basically caused a bit of panic. Other people pulled their money out of Anchor. People were trying to find ways to get rid of their UST. Luna started to decline. The entire market was already declining. It was kind of like the alignment of stars. The stock market was falling, and cryptocurrencies have really been correlated with the stock market these days. So everything was going down, and on top of that, the migration was happening. So, it was a perfect storm of coincidences there.

Joe: (32:59)

So the system was already very unstable and this was something that, you know, you recognized and other people, but you recognized going back to last year, it was inherently flawed, but then we get this sort of perfect storm because we have this big risk asset sell off. You know, stock market has obviously been tanking. Bitcoin, which on some level had of course been a contributor to UST stability, Bitcoin has obviously been tanking. So to the extent that that's a backstop that's dissolved or diminishing by the day. And then you have this migration and there was about to be a switch and someone dumps a lot and drains the liquidity from the existing pool. You mentioned, and I saw, I've seen a lot of people talk about this move from the 3Pool to the 4Pool. Was there something inherent about that migration that made it vulnerable? Was that inherently going to be a less liquid moment for UST?

So the system was already very unstable, which it was, but you recognized last year that it had inherent flaws. But then we got this perfect storm because we had this big risk asset sell-off. The stock market has obviously been falling. Bitcoin, which is certainly a contributor to the stability of US Treasuries, has obviously been falling. So in a way, this support point is disappearing or diminishing. Then you have this migration, there will be a switch, someone dumped a lot, draining liquidity from the existing pools. You mentioned that I've already seen a lot of people talking about this migration from 3Pool to 4Pool. Is there something inherent in this migration that makes it fragile? Was this supposed to be a moment of low liquidity for UST?

Kevin: (33:55)

Yeah. I mean, it would definitely be a less liquid moment, but I don't want to just say that definitively someone was trying to attack the protocol at that time, you know, because what it really could be is that somebody was just looking at the liquidity and didn't even know that the migration was happening that day. And then just saw all the liquidity evaporate, panicked themself. And then just dumped all their UST, draining the rest of liquidity. Or it might not have been one person. It could have been multiple people who all, you know, were sitting at their computers one day, not having read that they were doing the migration, everybody feeling the same panic, everybody dumping.

Yes, I mean, it would definitely be a moment of low liquidity, but I don't want to just say for sure that someone was trying to attack the protocol at that time, because the real possibility is that someone was just looking at the liquidity, not even knowing that the migration was happening that day. And then they saw all the liquidity evaporate and panicked themselves. Then they dumped all their UST, draining the remaining liquidity. Or, it may not be the work of one person. It could be multiple people, sitting at their computers one day, not seeing that a migration was happening, each feeling the same panic, and everyone selling off.

Tracy: (34:32)

So speaking of opportunistic moments, I guess, or triggers for these types of moves, you mentioned that you were short and I know you probably can't necessarily go into detail of exactly what that position looked like. But I imagine given the degree of criticism against Terra/Luna, that there were a number of people who would've liked to, or maybe have bet against this over time. And I'm wondering, you know, how did those trades theoretically work? And then is it possible that there were frictions within the Luna Terra ecosystem that made shorting it kind of difficult because this is a classic thing in markets…

So when it comes to opportunistic moments, I think, or the triggers for these types of actions, you mentioned that you were short, and I know you may not necessarily be able to specify exactly what that position looked like. But I imagine, given the level of criticism towards Terra/Luna, there were many people who wished, or over time, might have bet against it. I'm wondering how these trades theoretically went down? And then, could there have been any friction within the Luna Terra ecosystem that made shorting it a bit difficult, as this is a typical phenomenon in the market...

Joe: (35:20)

You can identify the bubble or you can identify the ponzi and then get killed as it goes up 

thousand percent.

You can identify bubbles, and you can identify Ponzi schemes, and then get killed as it rises to a thousandth.

Tracy: (35:27)

Right. Exactly. So I guess I'm wondering how painful it would've been to short this for a substantial amount of time?

Exactly. So I want to know, how painful is it going to be to short this thing for a pretty long time?

Kevin: (35:33)

Yeah, it definitely would be very painful. And this is why we didn't start shorting this thing until actually pretty recently. I want to say, I don't want to say exactly when, but it was sometime this month and it wasn't actually earlier than that, that we were short. And you know, I think that with something like this, being early is almost as bad as being wrong. And you know, partly one of the reasons that this thing was so difficult to short was because the funding rate on putting on shorts was extremely high and that's because the opportunity cost was getting yield on Anchor at 20%. So basically you're paying, it's about that. It'll be about that. If that yield was higher, it'd be even harder to short. If that yield was lower, it'd be easier to short. But basically that's the opportunity cost of not shorting.

Yes, it would definitely be very painful. That's why we only started shorting this recently. I don't want to say the exact timing, but sometime this month, not any earlier than that, we initiated our short position. I think with something like this, being early is almost as bad as being wrong. Part of the reason it's so difficult to short is because the rate to borrow the funds to short is very high, and that's because the opportunity cost is earning 20% on Anchor. So basically what you're paying out, that will be about it. If this yield were higher, it would be even harder to short. If this yield were lower, it would be easier to short. But basically, that's the opportunity cost of not shorting.

So that's why, you know, that's why this trade I think was particularly difficult to put on, just mechanically, from that standpoint. But also I think on top of that, this whole system and this whole design is so reflexive on both the up and the downside. So like we saw this thing go negative 99.9% down. Just in a couple of days. Right. But, you know, on the upside, it's also very violent. So, you know, you could be right. But you know, the market could still liquidate you, you know, for some reason, you know, the equity markets rally, everything rallies, Bitcoin rallies, Luna having beta to Bitcoin, it also rallies. And then on top of that, they make some kind of crazy announcement, whether it's true or not, and then, you know, you could completely get blown out. So, you know, I think that's why, you know, this thing was just particularly difficult, this short, you know, just structurally it's just very difficult, but we did get it in. And I think it was just roughly about the right timing. So very happy about that.

That's why I think this trade is particularly hard to execute, just from that perspective, mechanically. But also, beyond that, the entire system and design are so reflexive both on the way up and on the way down. So, like we saw this thing drop 99.9% in just a few days. Yes. But on the way up, it was also extremely aggressive. So, you might be right, but the market can still liquidate you. For some reason, if the stock market bounces and everything bounces, Bitcoin bounces, and since Luna has beta to Bitcoin, it bounces too. Then on top of that, they release some crazy announcement, whether it's true or false, and you could get completely wiped out. So, I think that's why this is particularly difficult, and the short is just structurally very challenging, but we did get it right. I think we just got the timing roughly correct, so I'm very happy about that.

Tracy: (37:23)

Yeah. I bet. I mean, it feels to me like one of the difficulties or one of the reasons this week has been so extreme is because there's no natural circuitbreaker on, well, either on the upside and this is why you saw, you know, some phenomenal valuations, but also on the downside.

Yes, I bet. In my view, one of the reasons or difficulties this week was so extreme because there were no natural circuit breakers, either on the upside, which is why you saw some incredible valuations, but also no natural circuit breakers on the downside.

Kevin: (37:41)

Yeah. You know, I think my opinion there is a little bit, you know, different than yours Tracy. Because you know, I think that the circuit breakers could have slowed things down, but couldn't have stopped them. And in some ways, by doing the circuit breakers, you know, like let's say the exchanges just, you know, limit up or down, they just hit the circuit breakers. Well, there's still the DeFi markets. And then there's still like all these other sources of true price discovery. And as long as there's some kind of outlet for it, you're basically just building up pressure on the CFI exchange once the circuit breaker is released. Right. So now, like, instead of it just exponentially crashing down, now it might just be straight up vertical line, you hit the limit. And then the moment the market reopens another straight vertical line down to the next limit and then just like staggered lines down.

Yes. I think there's a point to my view that differs slightly from your Tracy. Because, I think circuit breakers can slow things down, but they can't stop them. In some ways, by implementing circuit breakers, like exchanges just limiting the rise or fall, they just hit the circuit breakers. Well, there are still DeFi markets. And then there are all these other sources of real price discovery. As long as there's some outlet, once the circuit breakers are released, you're essentially just building pressure on centralized exchanges. So, instead of an exponential collapse, now it might just be a straight vertical line until you hit the limit. Then at the moment the market reopens, another straight vertical line drops to the next limit, and it descends like staggered lines.

Instead of like more of a curved line, almost vertical down. So, you know, I think at the end of the day, you can't really go against the market if it wants to go down, it's just going to go down. I mean, you know, any kind and especially for something so reflexive, most people are not trading, you know, people are trading emotionally, but there's what I'm saying is that there's more that's coming, right. It's not that, oh, you know, after a while people just want to go buy the dip, right. This asset is the complete opposite of the dip buying asset, right? Some assets they're more mean reverting when it goes down, maybe you buy some that looks cheap when it pops up. You know, maybe you sell down a little bit because it looks expensive. This is the complete opposite. The further it goes down, the easier it is to short, right? Because then first of all, like everybody else is unwinding, you know, that's happening. And then, second of all, the entire value of everything backing UST, which is circularly Luna itself, and Bitcoin, are both losing value. Right? So then at some point, you know, if you think the collateralization ratio was bad yesterday, then today it's even worse and the next day it's going to be even worse and you can see that trend deterministically playing out. Then at that point you might as well just dump it today. Right? So this thing is a purely reflexive asset. It's the purest of momentum assets.

Instead of being more curved, it's almost vertical downward. So, I think at the end of the day, if the market wants to fall, you can't really stop it; it just wants to fall. I mean, any kind of asset, especially something so reflexive, most people aren't trading rationally; they're trading emotionally. But what I'm saying is that there's more to come, right? It's not like after a while, people just want to buy the dip. This asset is completely opposite to buying the dip, right? Some assets are more vicious when they fall. Maybe you bought something that looked cheap, and when it bounces back, you might sell a little because it looks expensive. It's completely the opposite. The further it goes down, the easier it becomes to short. Because first, like others unwinding their positions, this is happening. Second, all the value supporting UST, which is tied to Luna itself and Bitcoin, is losing value. Right? So, at some point, if you thought the collateral ratio was bad yesterday, it's worse today, and it will be even worse tomorrow. You can see this trend unfolding clearly. At this point, you might as well dump it today. So this thing is a purely reflexive asset. It's the purest momentum asset.

Joe: (39:44)

So what you say makes total sense, but you know, I'm trying to understand actually how different fundamentally Luna Terra is from a lot of other crypto DeFi assets. And of course we recently did an episode with Sam Bankman-Fried and he was asked to describe yield farming. And he is like, oh, it's a box and you put money in. And then, you know, you get some governance tokens to incentivize more money in the box and then more money goes into the box. And then you make a lot of money if you're early in the box. You talk about the reflexivity of the Luna Terra box. And when it's going down, there's literally no reason there's no cash flow. There's no book value or anything that inherently stabilizes the price.

So what you're saying makes complete sense. I've been trying to understand how fundamentally different Luna Terra is from many other crypto DeFi assets. Of course, we recently did an episode with Sam Bankman-Fried where he was asked to describe yield farming. He said, it's like a box, you put money in. Then, you get some governance tokens to incentivize more money into the box, and then more money goes into the box. And if you were early in the box, you made a lot of money. You talked about the reflexivity of the Luna Terra box. And when it falls, there's literally no reason, no cash flow. No book value or anything inherently stabilizing the price.

Tracy: (40:27)

This is what I meant by a natural circuit breaker, by the way.

By the way, that's what I mean by a natural circuit breaker.

Joe: (40:30)

So then how different is it from a lot of other crypto things in terms of this reflexivity? And is this A) I mean, is it really different from the rest of this space? Because I don't see, like, it feels like a lot of crypto assets have the same reflexivity?

So, in terms of this reflexivity, how is it different from many other cryptocurrencies? Is it really that different from the rest of the space? Because I haven't seen it, and it feels like many crypto assets have the same reflexivity?

Kevin: (40:50)

Yeah. So I want to separate out the reflexivity from the garbageness of it. Okay. So I think on the reflexivity part, this is uniquely bad. Like Terra Luna is uniquely bad in its reflexivity. It's extremely, extremely high reflexivity. I think that's not the case for a lot of these other yield farming boxes. Now that being said, on the other side, there are some similarities in the sense that, there was a box also for Luna and it was called Anchor and you put money in that box. And seemingly you got money out of it. Now what it turned out to be is that, you know, and I want to maybe just go off of Sam's metaphor here, which I found hilarious, which is that, you know, in a lot of these cases, right, you put money in the box and then money comes out of it. Especially, you know, if you're early. I would say that in this case you put money in the box, seemingly money comes out of it. But really the true transfer of value here is that you put your money in the box, that money somehow through many, you know, different pipes goes to investors and founders of the project and then your money disappears and you have nothing. So it's more like that, I would say.

Yes. So I want to separate reflexivity from its junkiness. I think in terms of reflexivity, this is uniquely bad. Like Terra Luna is uniquely terrible in its reflexivity. Its reflexivity is extremely, extremely high. I think a lot of other yield farming isn't like this. That said, there are also some similarities from another perspective, which is that Luna also has a box called Anchor, where you put your money in that box. And it seems like you're getting money out of it. Now, the result is just according to Sam's analogy, which I find funny, in many such cases, you put money in the box, and then money comes out. Specifically, if you're early, in this case, you put money in the box, and it seems like money comes out. But the real value transfer is that you put your money in the box, and that money flows through many different channels to investors and the founders of the project, and then your money disappears, and you have nothing left. So, what I'm saying is, it's more like that.