Leo Zhang, Principal at Iterative Capital Management discusses the unseen impact Bitcoin miners have on the overall market



Episode 32 of The Scoop was recorded with Frank Chaparro, Matteo Leibowitz, and Leo Zhang, Principal at Iterative Capital. Listen below, and subscribe to The Scoop on Apple, Spotify, Google Play, Stitcher, or wherever you listen to podcasts. Email feedback and revision requests to [email protected].
Leo Zhang is Principal at Iterative Capital Management, an investment manager, cryptocurrency miner, and wholesale dealer. In this episode Leo joins Frank Chaparro and Matteo Leibowitz to discuss:
- The unseen impact Bitcoin miners have on the overall market
- The initial public offering of mining firm Canaan and new entrants to the market, like Layer1
- The significance of rainy season in China and it's importance to mining ecosystem
- Leo's take on upcoming halvening and it's impact on the entire Bitcoin ecosystem
- Why Bitcoin on / off ramps are a main focus for Iterative Capital's investment strategy
The following transcript is taken from episode thirty one of The Scoop, The Block’s flagship podcast. Listen below, and subscribe to The Scoop on Apple, Spotify, Google Play, Stitcher, or wherever you listen to podcasts. Email feedback and revision requests to [email protected].
The transcript is provided for your convenience, please excuse any errors or typos resulting from the transcription process:
Frank Chaparro Ladies and gentlemen, thank you so much for joining me today on what is a very special episode of The Scoop. I am joined for the first time in what feels like a century with Teo Leibowitz. He's co-hosting this show as we interview our very special guest, Leo Zhang. He's the research lab at Iterative Capital. Iterative touches many different corners of the cryptocurrency market. Mining, OTC, they're an investor, but they kind of fly a little bit under the radar. Leo, a prolific researcher, writer, is here to explore the interesting, often overlooked mining landscape. There are tons of developments going on in the market as it pertains to mining, as Leo will explore with us in a moment. The halvening is on the horizon. You have new firms coming to market, new firms coming to the U.S. market specifically. Leo, are you enjoying your cup of coffee? We got him whiskey, but he's a little jet-lagged, so he opted for the coffee. I'm enjoying the whiskey. Leo, thanks for joining us.
Leo Zhang Yeah, it's my pleasure. It's actually a pretty interesting day to talk about mining since the first-ever mining company is now--
Frank Chaparro Canaan has IPO'd. We could start there, or we could start a little more high-level about your background and what you're doing at Iterative and what the firm's all about.
Leo Zhang Yeah. So do you want to hear about Canaan first or Iterative? Because one huge, one is very small. Iterative is huge. Okay, okay, so the small one. So Canaan IPO definitely has been in the works for many years, as many of you know. And the size of the IPO has obviously changed a lot in the past I think year, a little more than a year since they first started playing with this concept. So obviously, I think this is a pretty big deal in terms of bringing this to--in front of traditional investors, to the general public--because not many people understand mining, and I still don't understand mining. So these guys have been preparing for this move since 2015 and they've been consciously shaping their company, or the way they do business, for this specific goal. Just for some reason, IPO is incredibly important for the founder, Angie Jong. So, I mean, Canaan is is perhaps the most O.G. mining company in the space. And so they definitely deserve to be the first company to get listed.
Frank Chaparro Well, let's talk a little bit about Iterative. When did the company come online for folks who may not be familiar with what you guys do, obviously touching OTC mining, investing? What's the sort of underpinning story of the firm?
Leo Zhang So it Iterative started in 2016. So the first fund was a venture fund. And so the first fund generated about 13x return from 2016 to 2017. And, obviously, there was ICOs wave. And so in mid-2017 we quickly realized that there's something wrong with this much money going on in this space. And so we made the conclusion at that time that this token mania is likely not going to continue. So we closed the fund, returned the capital to investors, everyone's happy, and we started a second fund that's dedicated to mining, which is the fund we're operating now. So Iterative Capital, the investment center--the reason people, you know, our peers, other investors in the cryptocurrency space, don't hear much from us or about us is because we only focus on mining. We do our own operations. You know, we go flip the switched ourselves, myself. And that's it. We don't do venture investing anymore. We don't really invest in tokens. We definitely don't participate in ICOs or IEOs anymore.
Matteo Leibowitz Do you think there's any appetite to return to venture down the line? And I say that because within even just the last six months or three months, we've seen a lot of venture funds raise large amounts of capital and looking to deploy that in kind of equity bets.
Leo Zhang Yeah.
Frank Chaparro So some of which were mining bets, which would give you an edge.
Leo Zhang Yeah. I definitely see more and more mining operations or so-called mining-related companies popping up, especially here in the U.S.. But I think, you know, here we're relatively late compared to, you know, many things that people in China have done in the past couple of years or experimented and tried and failed. And regarding your question on venture, I don't think that's something--it may or may not be the thing for us. But right now, I think this is this game is still very much a music chair. And for us, we want to focus on our product, Essure, which I think is something that we will work on for many, many years to come, if not, it's the product for us to focus on for the rest of our career.
Frank Chaparro How do you see the investing landscape changing over the next twelve months? And could there be a catalyst that maybe puts you into venture again?
Leo Zhang So I actually haven't really thought about that. So I myself am very much focused on mining, and there's already a lot of things going on in the mining space on a day to day basis. So I don't really, I'm not really fully subscribed to the DeFi and all that stuff. So I'm certainly not very psyched about what's going on in some of the altcoins. So, you know, just given my general preference and my own appetite, I don't think there's much that really excites me that fits into the venture-style investing.
Frank Chaparro Sure. Well, let's focus then on mining for a minute, and the halvening that is six months away at this point. You've written about the relationship between the rainy season in China and mining operations. Walk us through a little bit about what you're talking about.
Leo Zhang Yeah. So this is actually a very peculiar phenomenon. So, as you know, mining is very much--it depends on where the cheap electricity is. And miners in China have started this operation, and specifically, miners in Szechuan, have started playing with large-scale mining since 2013. I think the first-ever full-scale mining facility happened in Szechuan using hydropower. So there's a long history of large-scale miners situating there for a variety of reasons. One, the electricity is insanely cheap, and secondly, it's relatively remote. And people, you know, get away with a lot of things. And, you know, the government is generally pretty friendly to--the local governments, generally, are pretty friendly to stuff that as long as you claim it's some kind of innovation and the politicians, they don't, definitely don't understand, right? So, especially now--
Frank Chaparro And if you bring jobs, probably.
Leo Zhang Oh, yeah, absolutely. You pay tax. And there's tons of extra electricity generated, especially between April to October, which is the rainy season in Szechuan. So a lot of, you know, additional electricity just could go to waste. So, you know, using that capacity is definitely a beneficial thing for the local power plants, as well as for miners who are trying to make a profit. So that's why so much hash rate is concentrated in Szechuan. I think a conservative estimate done by a variety of researchers shows that it's probably around 50 to 60 percent of hash rate are in China. And the majority of that 50 to 60 percent are in Szechuan, and another province that also hosts a lot of hash rate is Yunan which is to the south of Szechuan.
Frank Chaparro Interesting. Well, from my perspective, I find it so interesting and I wish we covered mining more from a business perspective because there are so many different things that can either impede as a headwind or act as a tailwind. Weather. Geography. Government relations. All these things can impact the business in a way that's unique compared to other firms that operate in this market.
Leo Zhang Yeah, for sure. And also--
Frank Chaparro Walk us through some of those as a mining operator. How do you navigate some of these things that could serve as an impediment?
Leo Zhang Yeah, and just to finish on that thought, I think miners actually have a much bigger impact on what's going on in the market than people realize. And just because they're the only natural sellers and they need to sell, right? There's some miners who even sell on a daily basis just to capture that USD, sorry, fiat spread. And there are people who would only sell enough to cover their electricity and just hoard the rest. So miners definitely have a lot of coins. And once you have a lot of coins in a commodity game, you can play a lot of games. And, you know, it just naturally becomes more and more sophisticated when the coins disperse to other arenas. So just going back a little bit, back to why the mining industry is the way that it looks like today, at least in terms of geography. So I think, you know, obviously, there's this history there and also mining manufacturers are largely based in China as well. So shipping these machines to these locations is much, much easier compared to shipping these machines when you have to pay insurance, and potentially tariffs, and maintenance becomes harder, you don't have the right people who are trained on working with these machines once these machines leave China. So most of the large-scale miners who are still operating today, they only host the new latest gen machines in China because they're easy to fix and they ship some of their older S9s or just half-broken machines to places like the Middle East or Russia, where they, you know, they don't really care if the machines break as long as they get the cheapest electricity. So because so much hash power is concentrated in these regions, mining cycles behave--are influenced by the climates of these regions as well. So April to October, as I mentioned, is the rainy season and from November to March, pretty much, is the dry season. So this difference actually has a significant impact on the cost basis of and the general capacity of miners. And so when miners actually plan around these times in the year or to think about, you know, deployment schedule and when they buy new machines or sell some of their machines and facility owners, they plan around these schedules as to when do they do marketing conferences to attract all the suckers, yeah. And even manufacturers, right? They plan around these times before, you know, right before rainy season to sell their latest gen machines. So there's always big conferences in Chengdu, which is the, you know, the capital of Szechuan get more Shenzhen, which is the capital of hardware, around March, April. It's always very, very festive.
Matteo Leibowitz So another cycle that we've seen in the mining space is that these issuance cycles and namely the hardening of block rewards on a 4-year basis. Iterative put out some very interesting research projecting costs of production under various different assumptions. I'm just gonna run through some of the numbers. Under the assumption that you're using the watts miner M20S at three cents per kilowatt-hour, you'll see cost of production somewhere around seven thousand dollars. I think that's under the assumption that difficulty doubles after the halvening, and then, perhaps, under a slightly more conservative seven cents per kilowatt-hour, you're looking at cost production closer to seventeen thousand. So Bitcoin is currently trading just below eight thousand. Considering these projections, is now an especially good time to get into the mining industry, and how are you guys preparing for the next six months?
Leo Zhang Yeah, so I'd say those calculations are definitely done as a snapshot. And obviously, the mining industry is heavily influenced by the price, despite, you know--a lot of people don't realize that because it seems like hash rate is moving independently from price. But there's definitely a connection and hash rate definitely follows price, but there's a huge delay just because hardware, they take time to manufacture, and they take time to sell, and they definitely take time to get up and running. So I think there are a lot of uncertainty around halvening, even for miners. And so people who deployed early this year, they're definitely--especially people who have bought a lot of cheap S9s for April--the price jumped. Those people have earned back their principal many times over already. So I think there are a good amount of people who are expecting to retire some of the S9s or just, you know, older generation machines around halvening because your cost production literally doubles after crossing that threshold. So I think the general narrative is that people who buy these, you know, 7 nanonmeter to 10 nanometer machines, these are machines that prepare you through the next cycle, which is next April to October. And people who are operating S9s or some of the less power-efficient machines, they will likely retire or transfer some of the machines to people who have more resources. So it's hard to say whether halvening is going to have a very noticeable, drastic effect on hash rates. It really depends on where the price is.
Frank Chaparro So as someone operating in the mining industry, that you can't really say whether or not this will be bearish or bullish for your business or other miners in the space, it all depends on where the price goes?
Leo Zhang Yeah. So mining is actually a very--it depends a lot on timing, when you deploy. The best time to deploy is actually after right after a market crash, a big market crash, actually. Because one thing people who are outside the mining business don't realize is that mining machines are incredibly--the price of money machines are incredibly volatile. And this is because mining machines are inherently a bivariate option. It's an option whose variables are price and hash rate. So the pricing of these options is extremely sensitive and almost impossible to price. So most of the market right now uses some very simplistic metrics, such as static days to ROI, just number of days that you earn back your principal, which is clearly a very flawed, a very naive way of looking at the investment. But overall, money machines, their price fluctuates a lot. Especially after a big market crash, the expectation is that, you know, it takes much longer for you to earn back your principal and therefore a lot of machines, their price would drop as well. So those are the best times to approach mining manufacturers or vendors of secondary machines to acquire the machines that people don't want.
Matteo Leibowitz Sure. Yeah, so it seems it seems like a lot of these answers essentially come down to "It depends," in a lot of ways, but the trend itself is fairly clear, which is that, you know, assuming things stay relatively constant, miner revenue is going to fall over time--perhaps not over this next four year cycle, but, you know, give it perhaps eight years or twelve years or sixteen years. The solution there, ostensibly, is the emergence of a healthy transaction fee market. Is this something that you're thinking about? And I guess a follow to that is if we don't see the emergence of this transaction fee market, what are your preferred solutions as far as maintaining some kind of constant revenue stream for miners? Is it, "Lets transition to proof-of-stake?" Is it reintroduce steady inflation, re-appropriate some unmoved coins? How are you guys approaching this?
Leo Zhang Yeah. So I think as money rewards diminishes, there will be more and more consolidation happening in the industry. So I think it will definitely be, especially now that the iteration on the chips has become slower and slower--after 10 nanometer, it's gonna take long, long time before the latest tech known becomes available to--and longer period before the latest tech known, such as 5 nanometer, becomes available to integrated circuit designers. So I think mining as a business is going to become more and more expensive to run and requires more and more financialization. What I mean by that is people will have to manage their rewards in a much more careful way and manage their expenses in a much more careful way, and manage ways to hedge a much more thoughtful manner as opposed to, you know, between 2013 to 2016, people were just doing whatever because they were making so much money.
Frank Chaparro What are some new possible ways that a mining firm can hedge in that type of environment? I want to talk about hash rate derivatives.
Leo Zhang Yeah, anyway, we'll get to that. So there is no real efficient way to actually hedge. And the best you can do is keep your expenses as low as possible.
Frank Chaparro Best you can do is cross your fingers and hope for the best.
Leo Zhang Yeah. And for most people who have been in the business for a while, their cost basis is so much lower than where the market is, so they're fine, you know. So really, it comes down to your discipline around selling the rewards to make sure that you have enough electricity so that you don't scramble when the price is too low and you oversell your number of coins. So the important thing is make sure you purchase the hardware at as low as possible price and you're managing your electricity, your ongoing expenses, at the lowest possible costs. And when you rotate some of the coins into Fiat to pay for electricity, you're doing it in a relatively frictionless way. And obviously, you can hedge with Bitcoin futures, but the problem is that obviously volume is an issue. The volume has been growing a lot in the past year, but some of these exchanges that have the largest volume of these Bitcoin futures are not really trustworthy. So it's hard to scale. So for really large-scale miners, the most important thing still is managing your expenses.
Frank Chaparro Interesting. I feel like the trend, at least, you know, Layer 1 announced a massive, somewhat massive, I think, 50 million dollar fundraise at a 200 million dollar valuation with some pretty significant backers. They're making a bet on Texas. There was a merger announced yesterday between two firms that, frankly, I have never heard of, Whitestone US and Northern Bitcoin A.G.. They're merging to set up, in quotes, the largest Bitcoin mining facility worldwide. We'll see if that actually pans out to be true. If it's the largest, it's covering 100 acres of land, which is the equivalent to about fifty-seven soccer fields, and that'll also be based in Texas. There is this trend that, you know, we're moving to Texas now. This is the new--is this the new land of milk and honey for miners, is this the promised land, or are these firms misguided?
Leo Zhang Yes. Well, first of all, everyone's building the largest ever facility on the planet, everyone is doing that.
Frank Chaparro We're building the largest news and research site--here at The Block--in the world! So is there something to Texas is the question.
Leo Zhang Oh, definitely. There's a lot of cheap electricity in--the United States has a lot of power, right? And frankly, these people are just beginning to realize that, "Oh, OK, we can actually use this otherwise wasted electricity for Bitcoin mining." However, a lot of these facilities being advertised, they are not fully built. So we talked to a lot of these guys actually since early last year, people would tell you, "OK, we have a place, you know, in this very remote location in West Texas or upstate New York, and we have sub 3 saying electricity, 40 megawatts readily available, and we can scale to 700 megawatts in a year and a half," something like that. But you start talking to them, and it's like, "Oh, actually, we don't even have a roof. You know, you need to put in 10 million first so that we can build the roof." This happens so much, and this is really because mining for North America is still relatively new, especially mining at scale. Most people who are building these facilities, they are not miners themselves. So they, you know, as opposed to in China, where people who own facilities were miners first and then they realized, "OK, to fully capture these economics, I have to own my facility as well." Because once you own a facility, you can use that facility, or equity of that facility, as collateral to start borrowing Fiat to buy, you know, machines and just play that game again and again and again and scale from there. Whereas here, the facility owners, they just see, you know, a lot of Bitcoin miners as suckers who will pay them a margin for them to get rid of their electricity. So I think there's still a lot of work, a lot of infrastructure needs to build out before we see mining in the US, in North America, truly flourish. I think that we have seen some big facilities popping up, but I think these are rare, or relatively fewer, and they get to charge a much higher price because there are few. But we definitely see this trend. Although how sustainable is it? That's a completely different question, especially, you know, manufacturers are all based in China and getting the machines here, which is obviously a lot of hassle, maintaining them is a lot of hassle. So you have to make sure that your electricity is absolutely, absolutely dirt cheap.
Frank Chaparro And not to interrupt, but I feel that's the trajectory we're going in; more-so about, "How can we bring the electricity costs as low as possible?" versus, "How can we get our machines to work as effectively and efficiently as possible?" But at the same time, I wonder how things like maybe the trade war with China might impact some of these businesses, right? You have, you know, Layer 1 that is working with a certain type of chips, you know, silicon, which could become a casualty of this Trump trade war. Is that something you think about?
Leo Zhang Yeah. I mean, I am a Chinese national who lives in the U.S.. My parents are very concerned about my safety.
Frank Chaparro Are they, really?
Leo Zhang Yeah, of course. Because they, like, people in the U.S. Who read about these draconian headlines about what's going on in China, my parents also read about these draconian stories about, you know, people's lives suffering U.S. and racial war and all that stuff. But anyway, so it's just what we have to deal with, or, you know, people like ex-pats have to deal with. But just quickly before taking this too far, I think the tariff on the semiconductor industry is definitely going to have an impact on importing miners. But mining manufacturers are have been working around this for quite some time. So, for instance, you know, silicon is building a assembly line in Canada, I believe. I think Bitmain definitely has the ambition with miners coming to U.S. as well. The thing is, it doesn't matter where the integrated circuit is being designed. I mean, sorry. Layer 1's ICs are--they use a Beijing-based design firm and frankly, they should because at this point starting a new design shop for shot to 50 60, it's just stupid unless you're Intel. And anyways, so there are certain advantages to where certain things get manufactured. And this is just, you know, basic econ. So the way to get around this tariff is just simply having the assembly built here and the design still happen in Beijing or Shenzhen because the actual taping of the chips they're, you know, they're done by TSMC or Samsung, and these are Taiwanese or Korean companies.
Leo Zhang Do you think some of Layer 1's claims around 2.3--I think they're quoting 2.3 cents per kilowatt-hour for their El Paso-based operations. Do you think those estimates are realistic?
Frank Chaparro We had one source who described it as an utterly ridiculous estimate and that their assumption relies on a price floor for unused renewable energy, equating that to a pipe dream. Where might you stand?
Leo Zhang Yeah, so Alex is actually a good friend of mine, so I'm gonna be careful.
Frank Chaparro Yeah, and we're nice to Alex, too, we wrote a great story about Layer 1!
Leo Zhang Yeah, so I think that price, that raw price, is definitely available. The question is how much hidden cost, or cost that needs to add up to building out the infrastructure or labor? These things are expensive, man.
Frank Chaparro But will that price always be available, at all times? And what happens if the people--I think that one's based in El Paso, yes--are unwilling to bear the costs that mining operation might place on the power grid.
Leo Zhang Yeah. So that's definitely something that we saw in New York State last year. You know, the locals, they pushed back on these monstrous operations. So I think there is a lot of negotiation that needs to be done on the local level, which is another thing that is much easier to do in China than here, because you don't have to deal with other people's opinions. But anyways, I think I think 2.3 cents is not the most ridiculous thing I've heard. And, just to your point, the sustainability of that, and also the hidden costs associated with that, is going to have a very big impact on his business.
Matteo Leibowitz Let's switch over to Canaan, which we did address very briefly at the beginning of the podcast today. So they were listed in the U.S. today on the Nasdaq. I think it is currently trading slightly below its listing price of nine dollars. So Canaan is one of these O.G. basic manufacturers. In their IPO filing the firm said it controlled 23.3 percent of the global Bitcoin mining machine market share in Half1 2019. So they raised 90 million, but they were initially looking to raise 400 million. In Q2 of 2019, they actually lost 37 million dollars, and I think they have roughly 39 million dollars in cash left in the bank right now. Some estimates suggest that they actually need to raise on the order of two- to four-hundred million to actually maintain their operations. Do you think this latest raise was borne out of desperation or was it more of an opportunistic listing?
Leo Zhang I think it's both. So, like I said earlier, this company has been preparing for IPO since 2015, and they deliberately structure their business in a way to make sure that they're clean for IPO, whereas, you know, Bitmain gets all these troubles because they, you know--some messy stuff there, right? So even in 2015, Angie Jung, the founder of the company, deliberately chose to only accept Fiat as payments, whereas Bitmain accepts Bitcoin, Bitcoin cash, Litecoin, everything. So this is to make sure that their revenue is easy to account and also much cleaner when the regulation becomes tighter. So I'd say IPO is a big thing for them and a big milestone for the industry overall, but it's more symbolic than there is anything substantial there. I really don't think so.
Frank Chaparro Why Is it symbolic?
Leo Zhang Yeah.