Today, I’m sitting down with Su Zhu, my co-host and friend, to talk about the state of the crypto market and what he’s seeing on the horizon.
Su is the CEO as well as Chief Investment Officer at Three Arrows Capital, one of the largest trading firms in crypto.
Hasu is an independent investor, researcher, and writer.
Just one thing before we dive in, we make no money from this podcast we do want to grow our audience. so please take the time to leave us a rating and review on Apple Podcasts. It really helps a lot with discovery. Thank you.
Thanks so much for Alex Wice for creating this transcript!
In this episode:
We explore the state of the crypto market incl.
- why Defi crashed
- understanding risks for AMM LPs and how to hedge them
- how large firms think about industrial-scale farming
- Su’s outlook on Bitcoin dominance
Listen and subscribe here
H [Hasu]: Today I’m sitting down with Su Zhu, my friend and co host, to talk about the state of the crypto market and what he’s seeing on the horizon. Su is the CEO as well as Chief Investment Officer of Three Arrows Capital, one of the largest trading firms in crypto.
Just one thing before we dive in. We make no money from this podcast, but we do want to grow our audience. So please take the time to leave us a rating and review on every podcast. It really helps us with discovery. Thank you, and please enjoy the episode.
H: Su, welcome – I’m really excited to talk to you today.
S [Su]: Yeah – should be a fun one.
H: I want to talk to you about the market. So much is going on right now – I have questions about things that have happened or that are currently playing out, but I also want to use this as a segue to dive deeper into the fundamentals and the theory behind some of these things. So I hope we can tie them in a way that’s fun and formative for our listeners.
S: Sure.
H: To start out – so, the big story is probably the decline of Defi across the board. Many coins are down over 50 percent, some of them even down over 70%, such as YFI, yearn.finance. Why do you think Defi crashed, and did you see it coming?
S: I think the crash of Defi is, in any kind of big explosion of interest and community, you’ll have quite steep pullbacks, so I think this is no different in that sense – you can’t really have prices go up 10-20x without having 50%, 70% pullbacks. So I think part of that is the internal logic of markets cannot be fought against.
I think that for me personally, I think there were some warning signs a few weeks ago – where you had a lot of sort of, copy cat stuff popping up; a lot of forks thought they could just copy and get rich from that, there was a lot of me too kind of behaviors. So, I think that that’s usually symptomatic of blow off tops.
And there’s a few reasons for that – I think one is that there’s the… the local tops are always defined by a misallocation of resources – whether that’s in crypto or otherwise. So, during the dot-com boom, during the absolute tops, there were a lot of normal companies that were adding “dot com” to their name, and trying to refashion themselves as tech companies.
And similarly, during the top of defi here, you have random projects starting to say they are defi, that they can add an AMM, or add swaps. I had at least 3 or 4 discussions with top 200 coins where they wanted to figure out how to get into defi. So, I think things like this are always good signs, because they signal that the market … the market is creating much more novelty than innovation. We definitely had big corrections now, where people are realizing wait – the yields are not forever, someone has to buy governance tokens for me to get that yield. Who’s going to buy the governance token if there is no reason to own the governance token, or because the fully diluted market cap is extremely high – even at these levels.
So, I think that overall, people are just starting to ask the hard questions now – and nothing has really changed even that strongly fundamentally, but I do think that for people that are overexposed too, thinking that it’s just the new paradigm that “it never goes down”, I think that there also having their first taste of what markets are like, so I think it’s a mix of all those things.
H: Yeah, and we touched on this on our episode on YAM, that you tweeted a few days before that picture of the euthanasia rollercoaster – does this sum up what you’ve just described?
S: Absolutely, absolutely. The euthanasia – I think euthanasia, I first came across this motif in 2018 on speaking with another prominent trader in the markets, lightcrypto on Twitter. And he was discussing with me the idea of accumulation distribution – and what he means by that is when you have created a high in ALT/BTC pairs, once that high has been created and then nukes, the next bounce up is always a percentage away from that top, because people have seen that top, and there will be X% of people trying to sell to lock in some profits, and they know they have to sell at a lower price than the high that was printed.
And the incremental buyer is suddenly gone, because… or the incremental buyer is still there, possibly on the first bounce, but the sellers… the overwhelming size of the sellers is too big for the buyers to absorb. So you have… and as soon as you make a second lower high in something like this, then you will have this thing go again and again, toward this kind of vortex as we’re seeing now in some of these markets.
And I think that structure is very common in ALT/BTC, it’s very common in these hype cycles, because fundamentally its a money game – you’re betting that more money will go in than flow out; but if we are making lower highs, then that’s strong evidence to the contrary. So those who have unrealized profits – y’know, they thought they were up 5x, 10x, 8x, whatever it is, their now suddenly saying I want to make sure to book some profits so I’m not one of the losers at this game. Suddenly it becomes very apparent that you need new money to come in, and you cashing out strictly takes away some of that liquidity.
So I think that motif is a very good way to understand crypto markets, it’s helped me a lot and I strongly recommend … atleast when you see signs of things happening that are very euphoric, you ask yourself would it make sense for this to be the top of a euthanasia, and that’s in general a good way to think about it, and it will also help you not buy shallow dips right? Like it’s very scary, its very dangerous to buy a shallow dip at the beginning of the break, because fundamentally for the value buyers, they may still not be bidding even if it goes down 50 percent, right? So if you’ve been on a shallow dip, let’s say 10 percent, you’re betting that its not a euthanasia, that its just another dip to be bought. So you just have to be aware that if you’re wrong, you can be very wrong in that kind of scenario.
H: That’s a great point.
We’ve talked so much about scarcity, in the context of cryptocurrency, and each of these tokens is seen on their own as scarce – has a limited supply of tokens. And we’ve saw that with YFI, it had this kind of scarcity premium in the beginning.
But then what happens is, that we see a massive supply-side expansion for these kind of projects, because of all the different copycats. So first, there was just YFI, just yearn – and now you have at least 20 different yield farming projects, and many of them – like some of the bigger ones, like harvest.finance, or pickle.finance, attract over $200 million dollars of AUM, right? So would you say this contributes to the coaster? So if you have some projects that are genuinely innovative, and generally lead on the way up – but then with some delay, you have all these copycats coming online, and basically its such a large supply-side increase that you can’t be enough buyers to keep it going for much longer.
S: Sure, I think that’s definitely true – but what I would say is that it also depends on what part of the cycle we’re in. In the beginning of the cycle, I think the forks for the copycats … the first copycats, they actually help a lot. Like YAM, you could say was kind of a copycat … the ultimate copycat, right? Because they copied YFI elements, Ampleforth elements, Compound elements, and that was hugely popular at that time, even with these issues, and I think it led to the defi coins that you could stake for them going up shit tons, for a couple days, for a few days. So I think copycats at the beginning of the cycle are very good for the markets, price wise.
But I think that, like you said, as you get later in the cycle and people start asking questions about moats, and what are the actual cash flows that are going to be achieved, y’know, what’s going to actually happen? I do think that at that stage, you tend to have people saying okay – if there’s now Harvest that’s valued at this price, this market cap, and you have YFI valued at this price, what is going to be the long term reason why people buy this and not YFI, how are they going to capture the long term cash flows?
And I think that forks are actually really good at pointing out these kinds of things. Like, same with sushi and uni right? People might want to say UNI is worth all this because it has the brand behind it and has the [garbled] behind it, but at the end of the day, its open source code, and SUSHI really does the same thing – they arguably have a better coin distribution in some sense, because they don’t have a premine.
I think in defi the existence of those kinds of forks does start to have people ask questions: okay, can we actually put in a fee? Can we actually make this token capture value? So that has been what has been bearish for the markets. Like, Curve putting in fees, putting an admin fee of 2 basis points, and then having that go towards a buy and burn, and then having the price absolutely nuke right afterward, I think it was kind of the market saying that okay, maybe you have some cash flows, but it’s not a lot. And your inflation is still much higher than your cash flows.
So there’s ultimately not an easy way to say for new money that comes in and sees this space, why are these good investments? If you can’t articulate, how this token has a moat, how this community has a moat, what exactly is the value accrual? So that’s why I think that … I’m not as dovish as some people on Twitter – I mean, I’m a big long term believer in Defi, I’m not as dovish as some people are. I think there’s going to be a fairly long rebuilding period. Especially as we’re in a backdrop of what I think is going to be a high bitcoin dominance regime in the next few months – possibly a couple of years, even. I think that the builders in defi are ready to build, they are ready to truly innovate and create stuff. I do think that the speculators need to be braced for more pain, if … especially if they are in there short term, because there is still a lot of froth in the markets, there is still a lot of stuff that’s not that’s rational that’s happening, and I think the max pain point … from my point of view, from what I can see, it’s not really close yet.
H: I see. We’ve touched on this earlier, but does this crash invalidate any fundamental assumptions that you and other people have had about Defi, and perhaps even more interestingly, does it validate any assumptions as well – maybe things that continue to work, even though we’ve seen such a big drop?
S: I think, not a lot is invalidated per se, aside from the idea of … I think governance tokens people have been a little bit too bullish on as a concept. Like you’ve mentioned to me before, defi should be governance-minimalized, not maximalized. While I think that this whole wave of defi, especially the newcomers, you’ve had this big push into governance maximialization. The idea [is] that, this token can literally change anything at any point, and therefore it’s really valuable because you don’t know what they are going to pump and dump next. They may decide to vote in some crazy pool, and then those guys will want to pump that. Well you know, this logic only really works in a bull market. As soon as you turn into a bear market, it doesn’t really matter what you do, it’s still going to nuke.
So I had this general heuristic that a good sign that things will take a next leg down, is when projects try to vote in tokenomics that try to pump the coin, and it doesn’t even pump for like 4 hours, it just pumps for like, half an hour and then nukes. Or, when people try to make the rewards vesting, that’s like ultra-bearish, because it doesn’t actually accomplish anything, and it’s like the last card that any of these can play, so once you play that, TVL goes lower – like from the smaller players, but the bigger players can hedge anyway, and they don’t care, they’re very mercenary about it, so you’ll end up just having even less of a community with vesting because you’ll just have whales farming the entire thing, because they can handle the vesting, and they can hedge it. So that’s also a good sign that it’s not going to go great.
So I think that there’s … these are some of the things that you’ll tend to see, where I think tokenomics, I think people are putting a little bit too much focus on tokenomics, the ability of tokenomics to ultimately fix adoption I think has been overstated. I do think that’s a good thing, because that means in this bear market, people will go back and think about what innovation they’re really bringing, as opposed to trying to take something that’s already worth a lot and then say “I’ve created something else that’s a subset of this with one thing changed, can I be worth that?” So I think that’s going to be the direction that it goes towards.
I think in terms of validation, I do think that AMM’s have been quite validated overall as a concept. I think that you’ve seen a bunch of token sales happen on AMM’s, you’ve seen projects really bootstrap their liquidity through AMMs, Uniswap as well as Balancer, as well as others. So I think that is very healthy to see, and that’s happening even now, you know, Uniswap volume is still decent, Balancer volume is still okay. So I think that shows that, even if people are losing money, they still find it convenient to trade on a DEX, on a web wallet. So I think those are good signs and those do validate the idea that something has been created here of lasting value, and people can build on that.
H: Yeah, so since you brought up AMMs, one project that basically marked the top of the Defi hype cycle, is Uniswap with the launch of UNI token. You can farm UNI by being an LP in one of 4 Uniswap trading pairs, and then stake the LP shares in the Uniswap DAO, and by doing that you earn UNI rewards. So, I am farming this and so it’s very interesting to me. And so I would like to use this as a segue to talk about market making in general.
Last week, you wrote an article that was very well received, for Deribit, called “Toxic Flow: Its Sources and Counter-Strategies”. Maybe let’s start with that: what are the main insights from that article?
S: So I think AMM’s are revolutionary in the sense that anyone can be a marketmaker 24/7. Their liquidity is posted on the internet, and anyone can buy or sell from it. So I think that’s important to understand first and foremost. Now, with that, you have takers as well who are trading a lot of volume. So when you see very high Uniswap volumes, the questions might be … ok well, those volumes are so high, but what are they doing? Are all those people clicking on web wallets? Are all those people just fascinated by trading on MetaMask?
The answer, if you look on chain, is definitely not, right? What’s actually happening is that people are arbitraging between the price on the AMM, and the price on a centralized exchange. So, for pairs like tokens – like smaller tokens – you may find that the price discovery is actually occurring on Uniswap. But for a lot of other tokens – say Ether itself – it’s definitely not occurring mainly on Uniswap, it’s occurring on centralized exchanges, right? Like Coinbase, Bitfinex, and others.
And so what that means is that every time the ETH price moves, people are competing to do the trade on Uniswap, and then hedge it with a centralized exchange. So, that whole process of arbing the DEX is what’s really generating a ton of this volume. And so you might ask the question, these guys are obviously all making money, because they all will obviously only trade if they’re making money, but the AMMs all seem happy because they have their money in there, so they seem to be making money. So who’s losing money, or have we just created an infinite money machine?
And the answer is no, you have not created an infinite money machine. What you have created is two things, one is that the belief that the assets in this pool that the up and down movements are relatively safe to be indifferent between. So what that means is that one of the two assets in your Uniswap pool is not going to go to zero. If that goes to zero, that’s what the kids call a rugpull – that’s when you have a token in the pool, and ETH in the pool, and one day you wake up and the token is worth zero. And by being in the pool, you ended up now owning zero Ether and a lot of tokens. That’s impermanence loss in a nutshell – where you basically get stuffed with the one that’s worse, and you don’t have the one that’s better.
So I think that [at] that price right now, people don’t really … I mean it’s already very mathematical I think, and I think what essentially it being … an AMM LP is, is buying an unlicensed, structured product.
H: A what?
S: An unlicensed, structured product.
So, in Korea, a very common structured product is called a “worst of put.” So a normal put is when you sell a put, and that means that you agree to buy something at a lower price if it’s out of the money, you agree to buy it if it gets there. So if the price blows through there, then you … you’re losing money, because you sold that put. And you get assigned. The person who bought the put from you, they say okay I’m going to sell it to you at the put-strike, here you go.
So, a “worst of put” was invented by investment bank sales, devious as they are, and they said: okay, why don’t we put 3 assets into a basket, and we say to the guy, we say to the client: how about you’re selling a put not on one stock, but the worst of 3 stocks?
And the guy’s an idiot, so he doesn’t really care – so he’ll go and say okay, well, what’s the yield? Because you’re getting a premium when you sell a put. And obviously the yield is much higher the more stocks that you put into the basket, right? Because given sufficiently many stocks into a basket, the certainty that one will go to zero is 100 percent – even if its large caps. So that whole process became a very fun game for investment banks for a few years, and then regulators started saying “what’s going on here? How do you do this?”, and so now it’s generally capped like they only put 3 stocks in, and they are very blue-chip ones usually.
But, it’s something similar to that, you are going to get the asset that is the worst. And the question is, are you being paid enough in the meantime that it was all worth it? So, noone is really doing that math that closely aside from the big industrial scale farmers of UNI, and SUSHI as well – we all know that there are quite a few industrial scale farmers of SUSHI as well. And for those guys, what they’ll be thinking about is how do I hedge my risk? How do I use options on Deribit – because that is the only place where that’s liquid enough to buy huge size – you have to go there and basically buy, buy out of the money calls, out of the money puts, and hedge your risk that indeed, one of the assets you wake up one day, and one of the assets goes down 50 percent, or goes up 50 percent. And if you backtest crypto, that happens all the time.
So, I think right now that it’s quite fascinating to me that Defi has gone up in a period of Bitcoin going very flat, it’s basically been very range bound in [the] 10k to 12k range, the past few weeks. And ETH itself has also been relatively rangebound, between 310, 320 and 480 [sic]. So I think that Defi markets all, a lot of these primitives have been generally in that kind of environment where it hasn’t really been tested how people will behave if you start to get some serious impermanence loss.
So I think if you are not hedging, it’s probably gonna go bad at some point, as a guarantee, because you’re basically selling volatility. And that’s inescapable because you know that your hedge is to buy strangles, its to buy straddles, to buy strangles. So if you don’t buy them, then you’re not flat, you’re actually short.
That’s what’s paying for everything, right? It’s the fact that if one day people wake up and you know, Donald Trump says he’s buying all the Bitcoin he can, and Bitcoin doubles, that if you’re in the ETH/WBTC pool, that you see that you actually don’t have much WBTC, that you actually have all ETH, that’s unfortunate right. That hasn’t happened yet, because of the relatively low volatility environment that we are in, historically low volatility environment, but I think that’s a huge mistake to think that that will last forever.
H: So, you mentioned buying straddles and strangles – what does it take to buy those? Are those individual products on Deribit, for example?
S: Yeah, you just buy an out of the money call, out of the money put, and then at least if the prices goes up, you’re covered for that, and the UNI rewards you’re getting and the fees you’re getting, it pays for the premium for your options. All the big industrial scale players are doing this, because one is their size is big, right? So – if people are playing with around with a few hundred k, or tens of thousands, you’re hedging… you’re ultimately trying to save yourself from losing tens of thousands, which is not a lot of money. But if you are farming with 50 mill, 100 mill, then you start to say, okay, I don’t want to do it as a game, I want to do it like ASIC farming, ASIC mining itself.
So, there have been a lot of these products that have been launched on some.. in both China and well as the US. I think Coinlist has one, and there’s a few in China that I can’t name that have them as well. But, many of those will be hedging in some intervals with different strategies – I helped advise some.
Those guys will be fine, because they understand the math very well, and they’ll even pull out if the math goes bad. But, I do think that the math is capable of going very bad for people that don’t hedge, because they won’t have the gains on their hedge legs to offset when the market starts moving again.
H: So, you hedge your LP shares with options, and you mentioned you also hedge your reward tokens I suspect?
S: Hedging reward token is an interesting question, I can’t comment on what we do, but I will say that a lot of people do hedge rewards, similar as they might hedge Bitcoin mining rewards – if they think they will farm this amount of Bitcoin, they’ll definitely overhedge it in a neutral environment, because they want to lock in their profits. So you definitely see some of that going on in the farmable coins. And I think that’s just prudent, because if the price goes up, you just become richer, and if the price goes down, well, you’re glad to have been hedged.
And so I think that, in a way, that kind of shows the ultimate irony of yield farming. Because the yield farming, or generalized liquidity mining, the initial idea was that you were going to distribute the coins of your protocol to people that are using it the most, and by this process, it would be the fairest distribution and achieve the most motivated participants. But in practice, what has happened is that the largest players in crypto itself just use it as a money market. They put capital in, they farm it, they hedge everything out, its pure arbitrage, and they walk away with lots of money.
And in a way, it’s similar to mining, where – with GPU mining lets say – the GPU miner doesn’t need to care about any specific coin. He just wants to mine the stuff that people want to buy. So if there are people buying some random GPU coin, then he’ll obviously mine it and dump on them and hedge it out, and when the coin dies, he’ll move on to another coin.
So, money is sort of the ultimate GPU or CPU, because you can use money in anything. There’s no other machine needed, there’s no other proof of work needed. Hence, why you see a lot of heavy dumping – because the participants who have the money … you’ve created a dollar auction whenever you have a yield farming scheme, right? You’ve created a dollar auction for people to come into your network and use liquidity for that dollar. And there is no promise that they will stay around and play house after they … if the ROI is not high enough. There’s no such social contract, and there’s no such promise either.
So, I think that in the beginning, these projects probably got some hype out of it, because the total value locked went up so much, right? On Compound, the TVL, the strategy was first USDT vs USDT, then it was BAT vs BAT, and now its DAI vs DAI, and then they realized DAI vs DAI industrial farming is the least bad option, because at least it makes DAI supply pump, which helps Maker, so there’s that kind of logic to it, where they decided that’s the most allowable form of yield farming.
But yeah, it’s ultimately tied to that idea that you are distributing the token to people that are just going to sell it for dollars. But how do you do it in a way that people buying it actually make money in the end? You have to let the people that are buying it make money – otherwise, there’s no real chance.
And so for UNI, that will be the test, right? If people are actually farming and holding UNI, the question is what do they expect to achieve, what do they expect that they’ll be able to get for their UNI? Because they are very much just paying for the people that are farming just pure arb, right?
H: So, given that all these large firms basically mine fully hedged, would you say that other market participants just underestimate, or just basically misvalue the combination of low return with a very high sharpe ratio? Because you can always use leverage to increase the returns.
S: You can’t really increase leverage – at least not yet. There’s a project coming out that we recently invested in called Alpha Finance, they actually do allow you to leverage farm on Uniswap. But you can’t really use leverage in the traditional markets – I mean in the farming markets, because you need actual capital. Whether it’s your capital, or you’re risking someone else’s capital, it’s still ultimately one unit of risk, its atomic risk basically. It’s not like a derivative where you can go 5x, or 10x, or 25x right?
So I think that generally gives a huge advantage to big players, because they have the money, right? And as I alluded to in a tweet, the main money in this space is not even from stablecoin whales, or altcoin whales, but from bitcoiners, right? Bitcoiners have all the money because their coins are worth the most.
So what that means is that, from what I know, Bitcoiners make the most money in Defi yield farming, because they have the most money to begin with. So that kind of has become a basic yield, its ultimately a BTC-denominated yield for a lot of folks.
So I think that is the ultimate irony of defi, where the populist narrative is that Bitcoiners don’t understand Defi, you know, Tony Vays, he doesn’t even get that Uniswap can’t be stopped. But in reality, there are a lot of quiet Bitcoiners, with 100 thousand … 50 to 100 thousand Bitcoin, large scale. And if they deploy even 10% of this into Defi, that’s what happens. Because ultimately, the base reserve of crypto is BTC – it’s got the market cap, and it’s got the liquidity. I think that reality means that all yield farming ends up being done by people who view it as a money market.
And that’s not to say that they are naive either – these guys read the code. They read the Solidity code. They hate Solidity, they hate Ethereum, but they read it. So it’s this funny thing where, they view it just like mining – they may mine Litecoin, they may mine ZCash, they may mine other coins, but ultimately they view it as a business. So that’s ultimately the reality in this space. All yield farming projects today are being farmed in that way for a very large percentage of their supply.
H: So yeah, I mean it’s easy to forget that Bitcoin has been 20x larger than USDT which would be the second reserve currency of the space. So, given that BTC-denominated farmers capture most of the rewards of all of these tokens, is that part of why you think that Bitcoin dominance is going to increase over time, as basically these yields flow back into their base currency?
S: Yeah, that’s one reason I see. The other reason I see is that WBTC is very bearish for Ether. There are some arguments that, because it hurts the ability of the Bitcoin blockchain to receive fees for transactions, that it will therefore … it will be more bullish for Ether. But I think, in the near-to-medium term, what it will mean is that if people come in buy Bitcoin up [to] 15-16k, now you can FOMO Bitcoin from the ease of your DEX, right? You can go from your Ether straight to your WBTC, you can go from your stablecoins straight to WBTC. So I think that that effect in the near-term will be much more pronounced than the theoretical effects that, you know, Bitcoin is on the Ethereum chain.
I think that is something that is very, very bearish for ETH. Because it also, if you think about the Uniswap pool, ETH/WBTC is one of the four pools. That means that anyone can go from a lot of ETH to a lot of WBTC very quickly. And, because of the way that market works, you probably don’t have a lot of people sitting on WBTC to begin with, as a natural user. The natural users are all Ethereum ETH holders, on the Ethereum chain. So, WBTC is not really held in that way – so, in a sense, the most likely flow is that ETH people sell that pool consistently for WBTC to diversify. So, I think that that concept is something that people need to be aware of as well.
I think separate from that, I think there are going to be some headwinds for Ethereum over the next few months. I think as Vitalik has pushed towards a more rollup-focused roadmap, I think that it’s fundamentally more sound, but I think it’s going to be a wakeup call for the community too, because it does mean that a lot of the ETH2 research is probably wasted, and it also means that the TPS may not get to anywhere that allows mass scaling, or anything that resembles mass-adoption any time soon. So I think that will be the challenge.
I do think that all applications will still continue to use Ethereum and probably hook other chains if they make sense, they’ll be very chain-agnostic but they’ll use Ethereum given that’s where the uses are now. But I think for ETH the asset, it will be a challenging environment for sure. Because ultimately that headwind of the long term roadmap being so ultimately not very scalable and not very clear – even at this stage several years later, we’re still at the same TPS – and optimistic [rollups] will only really get you to 200 TPS, and it will … it will force a lot of social coordination of the stack right? It just forces it into Layer 2, you don’t really solve that many issues, and it creates centralization in terms of community in some sense too. So I think those are all things to be aware of.
H: Right, and it’s not just WBTC, we also saw the recent launch of tBTC and at least in my little Bitcoin bubble, I’m seeing some people that would use tBTC but not WBTC, because tBTC is trustless bridge – the first trustless bridge for any cryptocurrency, so that’s interesting as well.
S: I get that, I get that. I think the difference between tBTC is between DAI and USDT. I think Keep is academically quite interesting – that it can be done – but the thing that academics need to remember is that not everything that can be done should be done; or not everything that can be done, is more than an academic experiment, right? Because for players like us, or Alameda, or Grapefruit, or any of the other big houses, you want to get in and out fast – like, you want to get in and out of 5k BTC, 3k BTC. To be able to do that, you ultimately need that 1:1 peg somewhere – and, I think that also has to be capital efficient. And that’s very hard to do with a lot of the other more decentralized constructions.
And then the question is, does the market care or not, or do they trust BitGo, or do they trust BitGo DAO? I think the answer is that they trust the BitGo DAO more than they trust the holders of whatever token is come up with to bootstrap whatever new, so-called decentralized Bitcoin is made up.
H: BitGo DAO?
S: So, WBTC is governed by a DAO – actually, this DAO is quite powerful, it can even vote – so if you send WBTC to a contract address, they can actually vote to retrieve it, so they have some interesting powers.
H: Oh yeah, okay. That’s something that happens sometimes to people, where they accidentally copy-paste. Like, one time I’ve seen this happen is when people try to edit a token to a wallet, like a custom token or to an exchange, and they accidentally copy-paste the address, the contract address into the address field and send it. Well, I was not aware that WBTC is governed by a DAO, so … interesting.
I actually want to jump back to Uniswap for one moment, because there’s one thing that has been on my mind for months actually, and its the question if [abrupt end]
So, in a central limit order book exchange, marketmakers tend to increase their spreads when volatility is high, basically to spread their inventory over a larger level of prices. But AMM market makers don’t do this – atleast yet. But, you can do this. Even in Uniswap, you can do this. So, do you actively manage inventory in LPs based on expected volatility? I guess, is my logic even sound that you should be doing this?
S: Yeah, if you’re a market maker that needs to make money on trades, then yes, you should be doing this. If you don’t need to make money on trades … so, people who LP in Uniswap don’t need to make money, right? They may prefer to make money but they don’t need to, and that’s a key difference. I think Tiantian, at Amber in the article he wrote, he makes a very good point, where he said: for a holder of an LP token, he simply needs to personally be indifferent to these two coins somewhat, and he needs to believe that roughly over the long run, that these will outperform buying and holding the two of them.
So, as long as he believes this, then everything is fine. And even if he loses money, it’s the same as the Korean retail that buys structured product — he also thinks it’s okay because most of the time he gets the coupon, but then once every four years you have a big crash, and then he ends up being long the worst of 5 stocks.
So, you know I, I think that there’s a lot of things that in theory shouldn’t work because of efficient markets and because people are not dumb, but in practice they are working in Defi because, even if its dumb, its still less dumb than other things they may do. Like, that guy who puts it in the AMM pool, he may have been doing something even dumber with his money. He may have been trying to buy and sell levels, he may have been trying to get liquidated on BitMEX. Right? You have no idea what he was doing. So, it’s not a perfect – there’s no perfect solution – not everyone can arb everything all the time. Some people will have to just own things, we will just have to try to, compete … not really with the market but just with themselves, just try to make a little bit more money every year, right?
So, I think that in that framework, then you quickly see why spreads aren’t going to widen for announcements, probably the simplest solutions will do okay, better than people think. But yeah, there’s definitely a lot of room for AMM-innovations: we’ve backed a number of AMM-competitors that have what we think are slightly smarter AMM’s. But yeah, the market will decide. Because, if there’s an announcement and then [the] price goes way up then way down, then Uniswap-LP’s are happy, and if you were an LP in a smartpool where you widened, you wouldn’t get that volume. So maybe your LP’s will be results oriented and say – well okay, I wish I was in the other pool. You never really know how other people will think about it.
“If you try to always make money, you won’t do any trades”, that’s also a good way to think about it. “If you always want to make money, then you should never be a maker” actually, because makers by philosophy can never guarantee that they make money.
H: A lot of these AMM-models get constructed on paper, in an academic context where you assume strong rationality. Then they meet the real world, where the market acts with bounded rationality, and suddenly some model becomes dominant. That kind of relies on people not being rational, right? People having bounded rationality, such as Uniswap – and it becomes a new Schelling point for liquidity.
One story that bubbled up yesterday was… so, on-chain analysts found that someone borrowed a lot of YFI on — I don’t know if it was CREAM or AAVE, actually — and then sent it to Binance, which makes it pretty likely that they were a short seller, right? Because for what other trade would you do that [for] – maybe market make the coin on Binance, right? The address was I guess associated with Alameda, which mostly, probably makes [on] their own exchange, and not on Binance.
So, there was a bit of shitstorm against them, in response on social media, telegram groups and so on. What do you think of that event?
S: I think it showed how early this whole space is, where people don’t really understand trading, and they don’t understand what the goal of Defi should be.
I think first of all, whether Alameda was short or not, or they were short for a client, or they were short for an inflow, or they were hedging, it’s completely irrelevant. Because, for every trade in the market, there needs to be a buyer and seller – we can’t all be bulltards and no one does any trades and it goes to infinity. It has to trade at every price up, someone has to sell for every buy, even if its going up. So, to say that because he’s short selling – it kind of just tells you the mood of the market, right? Where it’s getting into an anger phase of the market, and they’re looking for excuses or scapegoats.
I think that, what’s very likely that they are doing and all [swod? – garbled] people do, is that they’re just arbing the centralized exchange versus the DEX. If you have most of YFI sitting – in terms of liquidity – if you have it sitting in the Uniswap pool, in the YFI/ETH pool, if there’s huge sellers on centralized exchanges, then you want to buy it from those guys, then sell it to the Uniswap pool. But then you have to wait for your withdrawal from Binance or wherever, which might take quite some time, might take half an hour.
So you might want to borrow some YFI, and that way you have it in your account, and as soon as the market nukes, you can hit the Uniswap pool and be the first to arb it, basically. Obviously, you can’t do a flash loan, because you can’t close the loop right? You can’t buy YFI/USDT on Binance using … and then use that YFI immediately to deliver to Uniswap.
H: Centralized exchanges are kind of the ugly part of flash loans.
S: Exactly, exactly, exactly. So you always need market makers to be borrowing lots of coins to be able to arbitrage centralized exchanges versus DEXes. That’s actually what fuels a lot of the borrowing and lending activity in the markets – it’s not people short selling, it’s not people doing crazy spoofs or whatever, but it’s really just market making.
But, I will say that even if they were short selling, that there’s absolutely nothing wrong with that, and like we saw with the LINK short sell attempt on AAVE, that those guys got liquidated – because the market was bullish, and they realized that they could push the price much higher than the liquidation price. So, there’s no risk free money in short selling, right? It actually takes quite a lot of conviction to short an asset, because a short … you can have an infinite loss on a short, you can lose your whole account on a short.
So, from that point of view, I think that the Defi space in particular risk making themselves look a little bit silly if they try to vilify short sellers. Because, that’s also the main point of Aave and Compound and CREAM, right? Which is that, you have two-sided liquidity – you have suppliers and borrowers. If you now are like, mad because you have borrowers selling your asset, well, you’re probably not cut out for Defi, right?
So I think that’s now become … I think in the last few hours on Twitter, I now see people saying okay, “yeah, he should be able to do whatever he wants.” But, yeah, you definitely have a few people who are sitting on a lot of losses now, and trying to find things to blame.
H: Yeah, and I mean also the borrowers pay interest to the suppliers, so (S: Yeah! Exactly.) the supplier gets, they get compensated for whatever the borrower does with the money. And you have some people saying, you can earn more interest by staking your YFI in the.. in a vault, governance vault for example. But if everyone did that, and no one supplied to CREAM and Aave, then of course also the … basically whatever revenue people in the governance vault get, would also be way lower. So basically, I think it all balances out in the end.
But, this story was interesting to me, not just because it shows the irrationality that you described, but also because it also kind of showcases a fundamental problem, something that really bothers me actually with Ethereum, and Defi actually. Which is that’s it’s getting easier and easier to track accounts with software like nansen, but I reckon that funds have even better proprietary data and tracking software. So in that sense, there’s especially no on-chain privacy, and it’s so much more deadly in Defi chains and basically any kind of stateful chains, than it would be in Bitcoin, where basically transactions can’t be frontrun or otherwise manipulated against you in any way, right?
So if, if I hold money in a Bitcoin account, or send it to an exchange, or send it to my… [trails off]. You know, whenever you just, [mumbling / garbled] just hold it, there’s nothing really malicious parties or like other market participants can do with your transaction. But when your transaction interacts with the global state – when they know what you are doing – then they can mess up your whole transaction, and future transaction as well.
So, how do you – as one of the largest firms in Defi, doing plenty of transactions – how does this lack of privacy affect you?
S: That’s a very tough question, because there’s no solution right now in Ethereum, right? So most of [the] big players, they don’t even bother being private, because it’s not worth the hassle of trying. Even if you own a large amount of supply, even if you try to deposit it to a centralized exchange and then withdraw it back, people will … can easily see it’s you, because who else would of done that, right? So, there’s absolutely no way to get privacy.
And I think that is a huge barrier to adoption as well, because I don’t think that companies want… Like, imagine if everyone’s bank account was just on Etherscan, like you could go on Etherscan and see how much money you have in the bank, you probably wouldn’t like that too much, right? So I kind of think that that, that is a very understated problem, and it mainly impacts whales, which is why its not as talked about. But I think that its a huge impediment for institutions to come in, [a] huge impediment for privacy-minded people to use Defi.
It’s not unsolvable, I think that people are working on it, but yeah – Its a huge issue, right? Because one is what you mentioned, which is the miner frontrunning, right? Where you can actually have censorship very easily through that. But then two is that you could have all your trades be seen, so you have to … basically you have to justify that you’re buying or you’re selling.
Whereas in practice, in the stock market, when Bill Ackman buys stocks or he sells stocks, he doesn’t have to tell the world why he’s doing it, he doesn’t have to tell people “yeah I’m bullish now” or “I’m bearish now”, he doesn’t have to give a speech about it, he can just do it.
But in Ethereum Defi, there’s this big meme around like, oh you never sell, oh you never do this or do that. So like, Blue Kirby had that issue, where like I’m selling from my ENS address, and I’m going to buy back somewhere else, because I don’t want everyone to follow my 30 YFI all day long.
So I think that, that kind of a community-feel, or that kind of a … mantra, it’s definitely not healthy. It’s definitely not the underpinnings of global finance, right? And it’s a very provincial kind of mentality. So I do hope that people build the needful tools there, and that it’s usable by real firms, by real people that just want to do finance – they don’t want to be in a community, they don’t wanna be … y’know, on Twitter talking about coins all day. They just want to use finance, because that’s real people, right? And I think that Ethereum does run the risk of being a circlejerk in that respect, where it’s mainly an elitist, or a kind of, a very niche subculture in the end.
H: Yeah, totally agree. And I think the YFI borrow story also shows us that the game theory just changes completely between a paradigm with hidden information, and one with open information. Its not necessarily unwinnable when basically all your information is [not] private and maybe all the smaller players is hidden, but it requires some changes in strategy, right? If the market is so fickle, and people are paying so much attention that borrowing YFI on Aave can make people dump, then basically firms should start doing that as a bluff, right?
S: Yeah, and eventually it won’t work, and it will all be equilibrium. But, the danger is that these kinds of things become very normalized – that people just track whales all day, and they enjoy the game. Because if you’re going to build a global financial system, you have to move beyond that – and if you’re not able to move beyond that, then the big money won’t come into play, they won’t want to join these games. So yeah, it has created some very interesting money games, and there are definitely ways to exploit those games, and exploit psychology, but it is a little bit sad for me to see too, because I think that if Ethereum had better on-chain privacy, or had some ways to get at that, then I could see it being much more usable.
I mean, everyone knows the top 20 addresses, pretty much who they are, in terms of farming, in terms of trading. I would say that people are generally pretty good about not doxxing, but at the end of the day, it’s a very small community, and I don’t see how newcomers will be comfortable with having all their financial transactions seen globally. That, to a newcomer, seems strictly worse than how finance already works. Like, way worse.
Because you’re telling the guy that he can both lose all his money from a hack – and by the way most custodians don’t support these smart contracts yet – but also everyone can see your transactions: so that they can see when you capitulate, they can see when you buy, they can see all these things. I just think that it’s not gonna fly.
H: I think it’s a bit of a sober way to end this episode, but thank you very much for the discussion – it was fascinating, but I think you leave some real alpha.
S: Alright, take care.
H: Take care.
