Leo Zhang - Bitcoin Mining Institutionalization

JohnPaul: [00:00:00] Hey everyone, welcome to the podcast. I’m your host JohnPaul and this is Digital Gold. Known to many as the Bitcoin Kid, I started my own cryptocurrency out of my parents’ basement back in 2013. The goal of this show is to simplify the crypto world and explore how it changes the way the world thinks about money through conversations with thought leaders [00:00:16] in the space.

[00:00:17] JohnPaul is the founder and CEO of Orm Capital Ventures. All opinions expressed by JP and podcast guests are solely their own and do not reflect the opinions of Orm Capital Ventures. This podcast is intended for informational purposes only and should not be relied upon foreign investment decisions.

JohnPaul: [00:00:35] Today I’m joined by Leo Jung, founder of Anitsha Research, which is dedicated to studying the emerging phenomenon of hash power as a new asset class. Leo has been involved in the creation of numerous papers on Bitcoin mining and has a very in-depth understanding of hash power. [00:00:54] Leo received his bachelor’s degree in mathematics from Stanford University and has worked for various capital groups including Inter-of-Capital, Global Atlantic Financial Group and Morgan Stanley Global Capital Markets. Leo, welcome to the show.

Leo: [00:01:05] Leo Jung, thanks for having me. Leo Jung, I want to start with the age old question of why running a mining operation or why run one when you can purchase coins in the open market? Yeah, this is actually the opening question of my latest paper, The Intelligent Bitcoin [00:01:17] Miner. I think there’s definitely a historical reasons that people want to do this as your audience probably very familiar with some of the intricacy of mining. But generally, mining is a possible way for investors or mining operators to accumulate coins at a cost that’s potentially much cheaper than purchasing via open market. Obviously, this

Leo: [00:01:38] is a gross generalization. It depends on the miner’s physical condition as well as the expenses that he or she pays. Or in comparison to the open market strategy, this can be very different. But overall, I think this is more of a method but not guaranteed path to accumulating Bitcoin at a much cheaper cost. [00:01:59] People asking that question, as you mentioned, there’s not one mine fits all category. There’s the electricity cost, the age of the machine, and there’s efficiency. You mentioned that those all play roles in the profitability of a miner. But do you expect that the miners that have been in the industry for four or five years, the players that are really starting

Leo: [00:02:17] to sell operations, they must be generating coins at a discount if they are gaining traction. Do you think those miners are going to be able to say that they’re continuing to mine coins for less? Or do you think that it really we’re going to see the industry change a lot over the next couple of years? And that might not be the case. [00:02:33] I think it depends on how active the miner is managing their business. And we’ve seen some of the older generation miners who got very lucky when they started, started early when the first generations of ASICS came out and they were able to secure access to a cheaper electricity. But as we all know, that’s just part of the one side of the equation. The

Leo: [00:02:54] method that the miner managed the coins is also very complex. You can’t just sit on the coins and just waiting for everything to pump, especially those who have lived through 2018, 2019. I’m well aware that some kind of active management is critical for cash flow. And overall, there’s some mitigation of downside risk. So I think as mining becomes more competitive, [00:03:16] it definitely requires a closer paying closer attention in the past. Some mining operator, they simply just purchase the machines. And at the price that I think is reasonable and plug them in at a facility, they hired some technicians to maintain them. And that’s a very passive way of managing. I think unless the person had two cents electricity for all year,

Leo: [00:03:38] I think it’s very difficult to sustain this way. I think the newer players are much more intelligent, much more aggressive and much more methodical. Because when you’re mining a large mining operation, you’re not just running a physical data center, you’re not just doing the day to day maintenance of machines. Obviously, that’s a very important part. But at the same time, [00:03:58] you’re also managing a liquid portfolio. Every day, this coins come in, you have to figure out what you want to do with them, what kind of risks you want to take. And if you have to liquidate coins to pay for electricity, what’s the exact way for you to do that? And if you want to have even more aggressive trading strategy, you can buy low open market and you can sell high, you can

Leo: [00:04:20] use a combination of all these tools, you can borrow fiat with your Bitcoin as collateral, when the market is down, you don’t want to sell coins. I’d say overall, the kind of tools that available for miners also growing very fast, the kind of mindset of miners also changing as the competition becomes more and more fierce. And you mentioned that liquid portfolio, not only the coins, but [00:04:44] really the miners also are a liquid portfolio that can be bought and sold and are bought and sold across the industry. Can you touch a little bit more, maybe on the value of A6 rising over the past

JohnPaul: [00:04:54] two months? And how do you see the financing of miners and steady purchases from large corporate

Leo: [00:05:00] players improving the stability and distribution of A6 for the next year? Yeah. So when I said liquid portfolio, definitely meant coins specifically. Obviously, machines can be unplugged and transported and sold to someone else across the globe. But I definitely think that’s a relatively illiquid market, just from our experience. I’m sure you have the similar [00:05:23] kind of impression as well. Of course, there are many distributors out there that are connecting buyers and sellers, but overall, it’s still the kind of transparency and the efficiency compared to some more liquid assets such as coins, which has a higher level of standardization. These markets tend to be a little slower, tend to be a little bit harder to just engage with. So

Leo: [00:05:45] regarding the question of the meteor rise of the value of A6 in the past two months, I think, obviously, that’s a direct consequence of coin price rising and demand for hash power skyrockets. I’d seen on social media that people do back and napkin calculations of the static ROI days is two days, if you buy machines today. Of course, the static ROI, static break, even these metrics are [00:06:11] highly misleading, but at the same time, there are popular heuristics for people to think through, okay, where the market is right now, at least today. So I think a lot of people see that, a lot of people realize, okay, especially for people who are already in the mining space, and they realize, okay, this is a great time to deploy machines, especially hash rate has not really fully caught

Leo: [00:06:33] up with the growth rate of price yet. I think ultimately, the A6 market is still subject to a very strict supply and demand, especially in the past year due to COVID, many of the manufacturers have a hard time catching up, have a hard time managing your supply chain. It wasn’t really until mid to late last year, the manufacturing in China started to recover, and plus the little drama in [00:06:57] Bitmain seems like it’s just finally settled and it’s going to take some time for me, Cree, who’s going to take over the manufacturing side of business to reinstall himself into the what’s left of the empire. The Bitmain empire, I love it. On to my next question, which is how is the fair market value of a unit calculated? I know in your research report,

Leo: [00:07:19] you mentioned that when volatility is suppressed, the theoretical value of the machine drastically decreases, but when volatility is high, the theoretical value rapidly increases. And I would say we’ve seen that over the past three months. So I know you’re trying to touch on that in the paper, keep on that a little bit more on how that fair market value really comes in the back of the map [00:07:38] can math works and why it necessarily isn’t the best way to figure out the value of one of these machines, and maybe how much is going to earn? Yeah, definitely. But before that, I want to emphasize that the fair value of a machine is going to be different for every mining operator. And this is something that’s fundamentally different for asset like this compared to,

Leo: [00:07:59] say, a call option for equity, right? Because the underlying is fungible, it’s the same for everybody. But the same machine produced by minor A with force and electricity versus minor B with the force and electricity, but much better cleaner condition, these are different. So not all computes are created equally, not all hash power is being produced at the same efficiency. [00:08:22] So I’d say the fair value, and I definitely regret using that word because it’s just so subjective. But I think it’s more of a way to for someone who is aware of the parameters that is using the electricity, all that stuff, to calculate what this value is for him. The method that we introduced in this paper is has very deep roots in options valuation. One thing that’s

Leo: [00:08:48] very clear is that the much of the return of hash power by hash power really means machines comes from a future production. And this is something not something that you can really price right away. And for something like that, there’s definitely this kind of cash flow, there’s option valuation framework. A lot of the volatility rate comes from the uncertainty of coin price when you receive [00:09:10] them in the future, as well as the number of coins that you’re going to receive on daily basis. So these uncertainties make hash power a far more complex option to value compared to equity or just commodity options. And so we have to really leverage some of the more powerful tools to really understand what’s really going on. So the method that we introduced in the talented Bitcoin miner

Leo: [00:09:34] is through a numerical simulation. We used a Monte Carlo simulation basically to assume the path of future Bitcoin, at least in 18 months or 24 months, or something like that follows a jump diffusion model. And this is actually something that quite a few academic paper that when it comes to modeling the random process for Bitcoin adopts. So basically we create 10,000 possible [00:09:58] path for Bitcoin’s price trajectory following this distribution in the next 24 months. And each path leads to a hash rate growth path. Obviously here, we assume some kind of visible connection between hash rate and price over at least a horizon of two years. So we all know that intuitively, there are some positive correlation between hash rate and price, even though they

Leo: [00:10:21] don’t necessarily happen in the same timeframe, the demand for hash rate always follows price. So we assume some kind of function and that function is definitely nowhere near perfect. And in each scenario, we build up a similar path for hash rate as well. And then we calculate how much the machine is going to generate following each path and we take an average of the 10,000

JohnPaul: [00:10:42] path. Thanks for explaining that, Sis. And as you mentioned, it’s on that blog post,

Leo: [00:10:46] people who will be linked in the show notes for anyone who wants to see some of the charts. I read through the whole thing. It was very informative and learned a lot there. Leo, I want to touch on

JohnPaul: [00:10:55] one thing you mentioned, which is not all hash powers created equally. How do you see this

Leo: [00:10:59] affecting a synthetic marketplace or tokenized marketplace coming about for trading hash power? We’ve seen a couple of different groups try to do that. And then thoughts on Poland’s recent token launch with their their wrapped hash rate. Yeah, I have a lot of thoughts on that. Thanks for for bringing that up. So first of all, you’re absolutely right that standardizing all these [00:11:19] parameters is the key of solving this, at least on the south side. There’s a whole another side of the problem, which is how to solve for the buyer side, which we’ll touch later. But it’s complex enough for creating a synthetic hash power for even on the seller side. So the easiest way to do this is obviously just aggregate all these machines and have them managed by a centralized

Leo: [00:11:40] host, which is exactly the approach that pooling and this new coin BTC ST, which the launch down finances are these guys are doing. So the benefit of that is it can reduce the physical variance, at least to some extent control the physical variance that issuer understands where the physical variance come from. And at the same time, they can at least guarantee the hash rate [00:12:02] output. So if on some days their facility is down, they can compensate the buyer with appropriate amount of coins. Obviously, these are the risk that the issuer have to take. And so this model is definitely not new. The first time this something like this happened or something similar to this happened was I remember if it was 2014 or 2015, the first person who popularized

Leo: [00:12:23] ASICs, this guy called a Fredcat, he’s a yellow computer science PhD, took leave of absence to go back to China to help manufacture the first generation of ASICs. He wasn’t the first one to come up with ASICs, but he was definitely at one point owned like over 60% of the market share of the ASICs. I actually remember him posting on the Bitcoin Talk form as much back in the day, selling the [00:12:45] wafers when I was waiting for my butterfly arms order, hitting my head against the wall, saying, oh no, what is going on? But not to digress. Those are the real early times of ASICs. So at one point, he was the zihan of mining. And there’s like very substantial concern that he alone controls the majority of the hash rate, as well as the market share. And to fund his next generation of machines,

Leo: [00:13:09] he pioneered the cloud mining model. Basically, he has his own hosting farm, he has his ASICs there, and he sells the claim to the stream of BTC produced by the farm in the form of a IPO. But it’s the exact name of that is not really clear what that really counts. But basically, it gives people dividends in the form of Bitcoin, people pay Bitcoin to purchase shares for that. Of course, [00:13:35] there’s no token mechanism involved because at that time tokens are just issuing tokens not nowhere near as easy as on ERC 20. But the operation went wrong for many other reasons, and the Frick had subsequently disappeared from the face of the earth. And even today, he’s not found. But after that, there’s just so many cloud mining business sprung up. Some of them

Leo: [00:13:58] have made it to a very large size, Genesis mining, and BitDear, you name it. And some of them they very quickly have gone out of business and disappeared from people’s memory. There’s still cloud mining games pop up every day. But in concept, I think there’s nothing theoretically wrong with cloud mining, right? It’s just people who don’t want the kind of hassle of operating machines, [00:14:24] wants the financial exposure, and don’t really care about which chain the hash power is building on. They should be able to purchase the financial return of hash power to someone at a premium to the electricity cost. But Leo, as that purchaser, they’re not getting any claims to the machine. Do you think that is an important aspect of a Bitcoin mining investment in owning the physical

Leo: [00:14:46] hardware? Or do you look at it as just a net zero because the end of the lifespan of these jeans are worth usually nothing? I think it depends on the preference of the investor. I think for some people who do care about that should definitely operate the machines directly or just do that through a special purpose vehicle that makes sure that he or she is the absolute owner of that [00:15:08] entity. And when something goes wrong, he at least have claimed to the assets that’s sitting in these entities, which is the machines. What I describe, the cloud mining platforms is a pure financial play. And this is suitable only to people who are strictly looking for a financial return. So I think this is a responsibility for the entity or platform that’s issuing this service.

Leo: [00:15:32] They have to be very clear what are the things that being offered. And it’s definitely the investor’s duty to understand what exactly it’s paying for. No, that makes perfect sense there on the different type of investor looking for different returns. I think we’ll see over the next couple of years facilities providing one off machines. Two clients, which we already are seeing [00:15:51] some companies approaching that model where you can still get ownership and still be a traditional hosting facility or run a traditional machine. But as you mentioned earlier, that does tie you to the lifespan of that individual hardware where you would have the risk spread out over multiple machines if you were in a larger hashrate token, like pool and token or something similar.

Leo: [00:16:12] Yeah, definitely. For instance, if you’re just someone from Wall Street and really interested in the financial return and you can invest in something like this, and when the whole thing expires, you end up getting a bunch of machines. What are you going to do with them? You don’t really know how to operate them. You don’t really know where to sell them. So that just creates [00:16:31] additional headache. So in that regard, just buying a pure financial swap is much easier, much cleaner. However, the problem with cloud mining platforms, at least with the format today is it’s just a very inefficient business model. These cloud mining platforms, they have to go out there, acquire machines, they have to put a huge risk on their balance sheet. And at the

Leo: [00:16:54] same time, they’re offering these contracts to the market at a premium to their electricity. They have to shoulder all the risk, all the problems that a typical large miner have to deal with. But at the same time, they’re not getting the kind of return that a single standalone mining operation is getting. The big problem is in a bull market, no miner wants to sell their machines. [00:17:16] And in a bear market, nobody’s going to want to purchase these cloud mining contracts, right? You can’t sell them. You have to sell them at discounts. So it’s very reactive. It’s a very inefficient play. And at the same time, these contracts, they do lock people into a long period time. Obviously, we’re seeing shorter periods now as more as the cloud mining space gets more competitive.

Leo: [00:17:37] But it is a problem. You pay something and you get stuck there for six months. And there’s no price discovery mechanism. There’s no adjustments. And the pricing is really just determined by the cloud mining platforms. And I doubt anyone there really understands what’s the rigorous way of pricing these things. So it’s not a good business model for the platform, not good business model [00:17:59] for miners who rent their hash rate to these platforms, not good for their buyer as well. But of course, this is a very blanket statement. There are definitely occasions where the buyers can make money from these purchases. But I think those are rare. So you’re saying that it’s not a good business model. Do you think pooling business model is better? Do you think it’s another business

JohnPaul: [00:18:19] model that would make more sense to provide hash rate exposure? What are your thoughts, I guess,

Leo: [00:18:24] on that? And do you think pools will start offering more of these contracts and purchasing from

JohnPaul: [00:18:29] facilities? Where do you see that industry going? Yeah, first of all, I think we should separate

Leo: [00:18:34] this pooling’s coin offering from pooling the pool because these are an entirely different thing. The reason is because the pooling’s coin is not offering with pooling’s customers. So the miners on pooling’s pool are not issuing their hash power through this vehicle. Those machines are controlled by pooling. It could be some other miners who are willing to sell their hash rate [00:18:57] to this process, but they still need to host your machines at a data center that pooling controls. So effectively, there is no selling mechanism. There’s no miners selling to them. There might be, but it’s not really a market mechanism. So I do think in that regard, from the seller’s perspective, it’s very similar to claw mining. Really, there’s no difference from claw mining.

Leo: [00:19:18] The machines are hosted in centralized spot and these contracts being generated from the centralized database. The only difference, obviously, it’s a one step further than claw mining because these contracts in the form of tokens are tradable. So people who have purchased previously earlier have mentioned that a big problem for buyers, that they get locked into [00:19:41] this contract for six months. And if price changes, they don’t want it, they want out. There’s no good way for them to do that. But now it’s possible for you to transfer these tokens. So in that regard, it’s definitely an improvement. However, the problem still is that these contracts are incredibly difficult to price. And this is definitely not the first time people have

Leo: [00:20:03] thought about this. In the past, we have Hashnast that issue these tokenized machines. There are various experiments that try to do something like this or tokenized claw mining. I think they all failed for various reasons, most of them for operational reasons. But I think if pulling electricity is low enough and if they manage their capital well enough, [00:20:26] they can definitely keep doing this. And they’re definitely taking advantage of the fact that the DeFi users are getting more and more used to complex instruments and instruments that have esoteric characteristics. I think that regard their playing something much smarter than traditional claw mining issuers. However, I do think that it sets a funny precedent. If this becomes something

Leo: [00:20:50] that’s successful, I do think every mining pool is going to or every mining correlation is going to try to replicate this because they all lack imagination. The problem for mining pools who look at this business model and think, oh, maybe I should try something similar is that they immediately subject themselves to the same type of operational risk that all large miners have to deal with. [00:21:15] And of course, they can try to convince the miners on their pool to issue this. But if the token is being created by the pool, then the pool is definitely a security broker from SCC’s perspective. So in your eyes, you would say that pool, even that token could be SCC regulated because it is fighting future profits on a machine in an investment-like contract.

Leo: [00:21:41] So I’m definitely not an expert on that. But I do think that it makes a difference that they are using their own machines. But I do think mining pools that use customers’ machines to issue a coin from the mining pool is more likely to become a security broker. For hash rate, that’s been viewed as security for selling it as a contract. What about selling the physical machines? Is that

JohnPaul: [00:22:05] in your eyes potentially, could you use it as security by the SCC? How do you see the

Leo: [00:22:09] difference between hash rate and the physical sale of a ASIC miner? Yeah, I think it’s the distinction actually from a financial perspective, the distinction is very small. If you’re someone who buys a hash-power contract versus someone who buys physical machines, of course, operationally, that makes tons of difference. But at the same time, financially, it’s the same. You’re swapping [00:22:29] USD for future Bitcoin. So I really don’t know how that should get categorized. I think it’s really unfortunate if for some reason, SCC deemed physical machines as security because that’s just funny. Who is going to be the security broker in this case? Are the manufacturers going to be deemed as security brokers now? It would be the first time in my eyes that electronic became a security

Leo: [00:22:54] where now, like you mentioned, Dell is now selling computers, which are securities because they could mine Bitcoin. That’s a good point. What is the biggest difference between or major differences that you experience mining in the US versus mining in China under the assumption that you’ve had facilities or you’ve run machines in both areas? Yeah, speaking of my own personal experience, [00:23:12] I think obviously the Chinese mining community has been growing for much longer. And also, I think that space has just been professionalized much faster. And just simply just because the proximity to manufacturers, proximity to mining pools, distributors, service providers, and did a whole ecosystem really built on proofwork. So that market is definitely getting more saturated. It’s

Leo: [00:23:36] definitely getting more and more competitive, although there are still facilities that are in Sichuan Yunnan that pop up every year that look for miners. But I do think that the time has given the Chinese mining community a tremendous amount of advantage in thinking through risk management and what’s the best way to service the machines. I think a huge disadvantage of US miners, of course, [00:23:57] the percentage of hash rate is definitely growing and would definitely see more interest demands and more serious players entering the space, which I think is a very good thing. But big disadvantages, I think, is really close to machines, as well as the people that you require to maintain these machines. You’re saying that experience in China was more professional, fully serviced solution

Leo: [00:24:18] versus in the US, where you might have had to have more hands-on on your machines. Is that kind of where you’re referring to? First of all, it’s just when you want to repair some parts, right? In China, it takes two days to ship to, you probably have less than that. It depends on where it are to ship to the manufacturer and get new components or get them checked up. But whereas here in the US, [00:24:37] and I’m definitely seeing more maintenance service centers and definitely seeing more technicians start getting trained. But the overall amount of time and also resources to require to maintain machines is definitely longer. I did see Core Scientific announce something, I think, late last year. I don’t remember time exactly that they’re going to become an official service

Leo: [00:24:57] center for Inkminer. I think that’s definitely an improvement. I think we need to see more service

[00:25:01] provider like that. Orm provides a bridge to the digital currency mining world for individual

Leo: [00:25:07] investors, financial institutions, and energy companies. By combining over 70 years of mining experience, 24-7 management and directly aligned incentives, Orm’s managed mining program is the simplest way to enter the digital currency mining market. To learn more, please visit forumcapitaltenters.com.

JohnPaul: [00:25:26] When and how did you first get into mining Bitcoin? Was it in the US? Was it in China? Can you talk

Leo: [00:25:30] a little bit farther on that experience? And if you still even have those machines? The first time I really got hands-on experience with mining is after I joined at this fund based in New York. That was 2017. So I started with some GPUs and moved it to A6. So that was a very hands-on experience for me and definitely learned a lot about the little annoying things that miners [00:25:52] experience every day. So before that, I definitely came from a much more finance background before joining crypto full-time. I was, as you introduced at the beginning of the podcast, I did a variety of Wall Street jobs. So my interests definitely do not start with mining necessarily, but start with just Bitcoin as a platform. And it just happened. I had a fund that I joined as it was my

Leo: [00:26:13] first full-time job in crypto. They made a killer return with ICO in 2017. I very quickly realized, yeah, that’s not sustainable. So started looking and mining started playing with educating capital to mining more seriously. And that’s how my journey started. And so that journey, that’s how you got access to the space. You do this fund. So you get started [00:26:36] in the facility. Now you’re still actively working with them to manage and run the facilities or be taking a more of a research role in your day-to-day. I left the company early last year. So I spent a lot of time researching the industry, not just the operation, but also the industry itself. I spent a lot of time talking to people in China. The huge advantage that I have actually is that I was born

Leo: [00:26:59] and raised in Sichuan. For people who are familiar with mining, they know that over 60% of, according to some rough estimate, a 60% of hash rate is situated in that province, because it has a very abundant hydropower in summer, summer, springtime. Obviously, I didn’t know that when I first interacted with the crypto community. But as I started to into mining, [00:27:20] I realized, oh, a lot of people actually speak the same dialect or from the same neighborhood. So it definitely helped me connect with some of the older players, some of the more experienced mining industry veterans. So that definitely gave me a lot of insights that otherwise not talked about or written about. So I started thinking about mining as an industry more broadly,

Leo: [00:27:41] and what are the problems that it currently experiences, which direction is likely going to evolve in the future. So if people who’ve read my blog post would probably realize that I actually talk about liquid hash power, liquidity of hash power, very extensively every now and then I will make a reference to it. It’s definitely very deliberate. So it’s something I realized and [00:28:02] something I start thinking very seriously in 2019. And I definitely think that the natural next step for this industry to evolve is to develop its own version of cap and market, similar to how oil or precious metal or other traditional commodity producers have built around those commodities. I think it’s not quite the same, but there’s definitely similarity for these kind of digital

Leo: [00:28:28] commodities. And once I started thinking about that, I could not stop thinking. And in beginning of last year, I left the old firm to start miniature research. On one hand, I want to push people’s understanding of mining to free them from the everyday dirty operation of connecting the wires because that does take a lot of time. And just try to start conversations, how to push the industry [00:28:54] towards that direction. And so when you’re talking about pushing the industry in that direction, are there any opportunities or areas that mining operations can focus on today in order to gain financing quickly and easily, or even improve the overall value of hash rate or the economic value of hash rate? Yeah, so I think both financing as well as the ability to increase the yield on a

Leo: [00:29:16] hash rate are been actively explored, especially in the recent one to two years. Because before that, these people don’t really talk about these, right? Or at least not discussed it very seriously. So on the financing side, I think we’re seeing more and more active players focusing on miners. There’s this, I want to say, fund. I actually don’t know how they categorize themselves, [00:29:35] called Arc does capital run by Trevor Smith. What they do is machine back the lending to help people purchase machines by lending them fiat. And that debt is collateralized by the machines that purchase. So these kind of activities are starting to become more popular because when miners purchase machines, they are taking a lot of risk. Your timing can be wrong. That’s just

Leo: [00:29:56] so many things can go wrong. So it would be really nice to have an insurance mechanism to transfer at least part of that risk to people who are willing to share that risk, of course, with appropriate compensation. I think some of the bigger desk, I don’t know if Galaxy also offers that. DCG Foundry is one of the newer players that’s coming to do this. I think the overall trend [00:30:16] is that the financing options are coming more open. Some of the traditional lending desks, such as BlockFi, Celsius, they’re also offering these kind of service to miners. In China, this is already a very established business like Babel Finance, a bunch of others. They’re all lending desks that have a very strong presence among the mining community. I definitely think that

Leo: [00:30:37] machine-backed debt is the way to grow financing options for miners. So that’s actively improving. As we speak, I remember in 2018, when we tried to raise funds for mining operation, it was difficult. It was just it takes a lot of effort first to educate people what the hell the mining is and what kind of risk that’s associated with it. I think as people’s understanding, [00:30:57] become more clear, people’s awareness of what kind of risk they’re facing, become more understood. They’re definitely willing to support a miner more and more. As for enhancing the yield on hash power, I think there are several directions that this is going. One, obviously, for non-BTC stuff, there are these profit switchers, nice hash style or Luxor also does this. I have

Leo: [00:31:20] to pull. You don’t really know what you’re mining actually. The pull decides at that specific time what is the best thing to mine. Of course, assume this is a GPU. At the end of the day, it’s just deliver the equivalent amount of ETH or BTC. So it’s possible that through services like this, the miner can get higher return compared to figuring out what to mine themselves, especially [00:31:42] for GPU miners. And of course, there’s limitation because if let’s say some chipcoin that’s going to pump and it’s very difficult for the service provider to realize that and gamble on that and make that decision. However, I think as a GPU mining also becomes more industrialized, migrating from the hobbyists seeing, I think this kind of service will become more and more popular. And on the other

Leo: [00:32:06] hand, you have open source firmware like brains is also improving the efficiency for a miner. Everyone’s still eagerly waiting for them to roll out what’s miner support. It’s according to themselves, it’s a soon TM. And so besides the things that to do with the hash rate itself, there’s definitely other things that can help mitigate risk such as more financial play, [00:32:28] collateralized lending is definitely one way to do it for miners to not oversell their coins to pay for electricity, especially during a down market. There are also miners who are a little bit more well versed in trading terminology that can hash their price risk with futures, perps, whatever. And obviously for Ethereum miners, the universe is much, much bigger, the kind of stuff that they

Leo: [00:32:49] can do with it. Ease is also just things growing. So I think there is definitely a lot of exciting things that’s built on what comes after you receive your reward in your wallets and let it sit there. I think that was one of the things that I was thinking about a couple of days ago, which was this idea that capital management in the mining space is one of the most important things you [00:33:11] can do as a miner versus when we’re talking to energy companies in this space, they’re really looking to liquidate that Bitcoin or Ethereum. Immediately for cash, they view it as a way to go from one megawatt hour price to another megawatt hour price. Whereas a miner who’s more crypto-focused is viewing this as a way to collect more coins. And then as you mentioned, utilizing those coins

Leo: [00:33:31] to provide additional yield or to increase risk and increase upside through these different mechanisms. Do you see that, as you mentioned, the managed funds being a very core position?

JohnPaul: [00:33:42] How do you see energy companies coming into that space? Do you think that they will immediately

Leo: [00:33:47] start to start using these vehicles? Or do you think this will be mainly used for the foreseeable future by miners who have a long Bitcoin outlook or long Ethereum outlook? So I think all miners fall on somewhere around the spectrum. On one side of the spectrum, you have these typical energy providers who are really just looking for a way to get more juice. They would sell immediately and [00:34:07] they don’t want any of this. Actually, traditional oil producers, traditional depression, metal producers, they also just immediately sell very rarely the whole inventory and take that risk. Just definitely, when we do it, they definitely seem very smart. But in market where coin price continues to increase, their upside is very limited. So this is also

Leo: [00:34:26] some scenarios I ran in the last blog post as well. And on the other side of it, the spectrum, you have miners who are really here just to get as many coins as possible. They would not sell. Under any circumstances and just wait for coins to pump. I think older generation of miners, people who started in 2013 who have seen the crazy price that broke their brains [00:34:49] are definitely leaning towards the latter. I think most miners today, they fall somewhere on this spectrum. I think it really depends on where you’re coming from, what’s your perception of this market and your risk tolerance. I think people who come from energy space, just dipping their toe in Bitcoin, I think it’s definitely easier for them to

Leo: [00:35:10] lean towards the more conservative side of the spectrum as a start. I see no problem with that. So Leo, talking about those different industry players and where they stand on that spectrum,

JohnPaul: [00:35:19] I want to ask you, we’ll just jump into it. Are governments mining Bitcoin? And when does the

Leo: [00:35:23] first publicly announced that they have started mining Bitcoin? And does that start a cascade from these governments across the world to start using energy resources to acquire Bitcoin? Yeah. So the honest answer is I don’t really know. I’ve seen use such as the Pakistani government recently just announced that they’re going to do a government mining operation of Venezuela, [00:35:43] Iran. I think we’re all seeing news announcements like this. But it’s very difficult to know how substantial they are. I think on grand scheme of things, it doesn’t really matter because I think it just really depends on the purpose of these guys, right? Are they trying to hold as many bitcoins possible as reserve? Because someone high up that sees the potential of Bitcoin

Leo: [00:36:05] and wants to protect their treasuries and something like that. Or they’re just simply a way to get rid of the waste of energy. I don’t think it’s as big a deal as some people are concerned, because ultimately Bitcoin is a Bitcoin. I guess on that note, have you seen the industry consolidating in 2021? Is it at the mining facility level or are you predicting that

JohnPaul: [00:36:25] mining companies will begin to merge? How do you see this landscape playing out? And then on another

Leo: [00:36:30] note, the value of the mining and mining facility, do you see that increased a year from now? Or

JohnPaul: [00:36:35] where do you see the overall market cap of Bitcoin mining and Bitcoin miners and daily revenue going?

Leo: [00:36:42] Yeah. So I think this comes in faces or somewhat cyclic. So during an explosion when coins just goes crazy, right? Like we saw just now, I think it definitely opens up a lot of opportunities for new players to come into the space, whether it’s facility owners or energy providers or just manufacturers. Everyone’s a little bit more capitalized than before, and everyone’s a little bit more [00:37:04] incentivized to break the current paradigm of control or market share. So I think every time there’s an explosion of coin price upward, I definitely think we would see more new players entering the field. I think for facilities, this is a longer process because it involves the conversion of a lot of the stranded energy asset holders to turn into Bitcoin mining facility providers.

Leo: [00:37:26] And I’m sure if you talk to tons of them as well, just in my experience in the past few years, there are a lot of these stranded energy asset owners that are looking to get Bitcoin miners into their space. But the problem is they always take a lot of time or a lot of upfront capital in order to get the warehousing in the proper shape. I think the increase in demand for mining in [00:37:47] North America definitely facilitates that. But the question remains, what kind of player has the most advantage in games like this? So that’s on the facility side. And as for the places that we see the fastest expansion will be financial service providers to miners. Imagine if you’re a public company like Marathon, and you’re placing order for a lot of S19, and they are delivered seven

Leo: [00:38:11] months from now, there’s definitely something you have to do between this period of time. You definitely need to try to find some way to hedge or try to, I wouldn’t be surprised that they raised debt to, I’m not Tom American specifically, but players like that would raise debt to fund that or convertible notes to with some kind of conditions. If the whole deal falls apart, [00:38:30] something gets returned to do that kind of funding. Because when I saw the news and I was thinking, if I were a very conservative traditional player who has a public company and just getting monitored all the time, I wouldn’t want to take that kind of risk. So how I would do that, and if it’s the only option that’s available, and I have to do it, the way I would do it is by selling someone some

Leo: [00:38:50] kind of convertible note that if by the time we receive machines, something goes wrong or the market goes sour in certain direction, there definitely pre-built conditions, I’ll have to do something about that. In your eyes, how many machines a month are the major manufacturers producing either in aggregate or individually? That’s a tough question. I don’t really have the [00:39:12] number on top of my head, and I think it’s really difficult to estimate. The way that from my experience of interacting with the manufacturers, it’s very demand-based. It’s because a lot of these guys are deeply scarred by 2018. The context is that in the 2017, everyone was a little shocked by the Buran, and everyone’s really misjudged how high the market was going to go in the first half of 2018.

Leo: [00:39:36] So a lot of the manufacturers bit me. What’s the matter? They all ordered tons of components, and throughout 2018, it was just basically a process of painful liquidation, painful process, clearing, and inventory at a discount. And the same thing happened with NVIDIA as well. We all know that 2018 was a time of inventory flush for a media, and you absolutely hated crypto after that. [00:39:58] Do you believe that large tech companies, such as NVIDIA, AMD, Google, Samsung will start building ASICs? Now that we’re past that maybe 2017, oh, Bitcoin’s a bubble. It’s not going to go anywhere after this. Now to 2021, oh, Bitcoin is still here. This industry is still massive, and it’s growing even at a faster pace. First of all, I don’t think NVIDIA is probably going to make ASICs, but

Leo: [00:40:19] for Intel, Samsung would definitely hurt rumors like this, and that definitely hurt rumors that sound a little more real than just average rumors. I think the question comes down to how sustainable the demand is. So in order for these older guys to really play with this market, they’ll want to make sure that there is consistent demand, that not like one-off thing, [00:40:39] means that I think that’s something that really scares them if there’s huge demand one year, and there’s no demand the next year after. I think that’s just like very devastating for someone who to have to be really careful in managing supply chain. I do think that there’s always, especially with given the recent price rally, it definitely incentivized new players to come

Leo: [00:40:59] to play. Are they more startups, IC designers, or big players who already have a lot of expertise in IC designers? I’d say it wouldn’t be surprised players from both backgrounds are looking at this very closely. And I would agree with you on that, that players from a wide range of backgrounds are looking at it closely. But the question is, are they going to pull the trigger in these [00:41:21] manufacturers? They have a pretty big IP moat. Of course, the companies we’re talking about do have enough engineers and very smart people to build the ASICS. It will take a while. And on that note, I’ve heard that fab vocation of chips is really the biggest delay on getting those wafers. It takes about four months to get a wafer order into a machine. Can you touch a little bit more on

Leo: [00:41:43] maybe the overall wafer capacity of these facilities and anything you’ve heard on that end? Yeah, definitely. So I think the game is still relatively early on the grand scheme of semiconductor space. So even though Bitmain seems like very well established, very powerful, it’s like I have a lot of market share, but we’ve seen how market share in the space come and go. And especially [00:42:06] not Bitmain is experiencing such a damaging, such a painful internal struggle. They’re definitely very delayed. I think that opens up space for new players. And what’s minor has really is what people consider the top contender, but they haven’t really fully get their supply chain management up to speed. And the reason this is so hard for these smaller manufacturers,

Leo: [00:42:29] by small compared to Apple, Qualcomm, Kuplace, massive order at the Foundry, compared to these guys, they don’t have that much negotiation ability compared to these bigger guys, especially now that the back end process is events to a seven nanometer, eight nanometer. That means they have to compete with much more powerful players. Unlike back in the days when the mining was to rely on [00:42:51] 16 nanometer, there’s much easier to get wafers from when apples of the world are not looking for those, or you’re not competing with those. So I definitely think as mining becomes more competitive in the Foundry process, I think the for-afer management for manufacturers also becoming harder and harder. But I do think that a lot of these manufacturers, in order to keep

Leo: [00:43:12] momentum, they’re going to shift their focus from simply racing to build a role models on the more and more advanced back end, keep pursuing seven five nanometer instead of that, they’re going to shift their focus on heat dissipation, some more interesting ways to structure and stack these machines. Several years ago, I’ve seen these server racks model. I think they’re going to come back [00:43:34] again, hopefully with better heat dissipation model and immersion cooling. So basically, the equivalent of Moore’s law for hash power production is also slowing down. The focus should more improve the efficiency from an architecture level or just the environment of the facility. Do you see a moment in the future, Leo, where you’re buying almost like shipping containers of

Leo: [00:43:57] miners going back to early days of computers where they took up buildings, thousands of ASIC boards controlled by one or two controllers and really instead of buying one machine, for $10,000, you’re buying one machine for half a million dollars or a million dollars. Do you see an industry moving that direction in any sense or do you see it more of just improvements on the current [00:44:16] box style design we have today? I think that’s the direction, but I don’t think that’s going to happen immediately. I think that’s going to happen with one or two players who have the luxury of spending that much CapEx. So the architect who designed Google’s data center has this famous quote that says, the data center is the computer. And the idea is that the way that the data center

Leo: [00:44:38] is structured, the entire facility is tailored to maximize the function that the machine’s in the data center is going to provide. So I think a similar mentor also applies here that in the future that the facility is the minor. We’re going to see more efforts on tailoring the data center itself, whether it’s how to deal with temperature, how to deal with dust, and or just [00:45:01] how to arrange these machines. I think people are going to be more thoughtful around these things. Obviously, there’s no one way to fit all right, because every data center is different. But I do think that facility owners who are a little bit more intelligent on that regard will become popular very soon. I definitely agree. There’s going to be a lot of expansion space as we look for more

Leo: [00:45:21] term key options. They may try to do that with their main inbox as like a first level, but it’s not the level that I expect to see 10 years from now. What does a mining environment look like in your eyes supported only by transaction fees? And do you think we’ll still see the level of machines and

JohnPaul: [00:45:35] growth? I know that is pretty far out. But what are your thoughts there? I want to say that the

Leo: [00:45:40] growth of ashtray, at least the rate of growth of hashtray is slowing down. But we know that kind of if you’re a stick, break the moment price 10x, and that’s definitely going to happen again and again. I’m definitely hesitant to say that the production of the machine is slowing down. But I do think that the ability for each single manufacturer to respond to price change is slowing down. That [00:46:03] is because the kind of machines that are manufactured S19 is objectively harder to manufacture than S9. It takes much more efforts to get wafer, to get everything in place in ship one of those. But obviously this observation is based on the assumption that everything else states constant, but in your environment where the price constantly 10x, I think it definitely introduces new

Leo: [00:46:27] employers that definitely introduce more options for buyers. So it’s hard to tell. But I do think having a schedule in the Bitcoin space is sufficient and has shown that it does balance out the system and bring that equilibrium. What are some of the things you’re researching the most right now that you’re willing to share with the audience? I spend a lot of time thinking about liquid [00:46:46] hashrate. I discussed earlier, I think they’re still quite a bit at work. It involves a lot of details and details really matter in designing something like this. It’s definitely worth waiting to see what is the right way to do this. And I think it’s likely going to take multiple iterations for someone to get it right. Besides that, I’m actually very, I don’t know if your audience

Leo: [00:47:05] are necessarily maxis, but I’m very interested in Ethereum mining recently. There are rumors about Ethereum A60, despite that the uncertainties around moving to proof of stake. I think Ethereum mining is more interesting to me because one that at this point in time, the fees as percentage of revenue is much higher. So I think it definitely introduces a new type of thinking on revenue model [00:47:31] and definitely introduce a lot of uncertainty to the predictability of mining return. And there all these new games that’s being introduced to Ethereum that can potentially leverage for miners makes it more colorful. I was thinking Ethereum was going to be stop mining with this recent upgrade. How long until those block awards stop and will miners still be able to get

Leo: [00:47:51] transactions fees or how is that going to work? Yeah, so the consensus on ETH 2.0, first of all, it’s going to take a long time. And secondly, the consensus is that the proof of work chain is probably going to happen in parallel with the proof of stake chain. So I think most likely, but of course, this is personal opinion that we’re going to see two ecosystem existing parallel. [00:48:11] There’s going to be some kind of bridge that allow you to go between one and ETH 2. I think from ETH 2.0’s perspective, I think it effectively is really a hybrid proof of work proof of stake. The proof of work really provides the consensus and the proof of stake is the process moving to one point out to ETH 2.0 is equivalent to Decred’s purchasing ticket.

Leo: [00:48:36] Okay, that makes more sense. So you’re saying that this transition, it’s not just like immediately, I thought it was like when the upgrade was done, mining awards are going to be done, but you’re saying that can continue. And then also, you do believe a chain split will occur like we have with the Ethereum Classic. And that you think Ethereum POW will still be a chain that [00:48:52] people utilize and participate on. So nobody knows what exactly is going to happen at this point, but that’s my personal guess. It’s kind of crazy. We don’t know what’s going to happen with this massive network. People are developing and people are asking to buy GPUs all the time, putting more capital into the space. Do you see any other coins really stepping in and taking

Leo: [00:49:10] Ethereum’s place as one of the most profitable GPU coins for GPU miners? If and when they move completely over to proof of stake? Yeah, that’s a hard question. Obviously, I don’t have anyone on top of my head and I don’t think anybody comes close. But if it were ever to come to the point that proof of a work chain just disappear from the face of the planet immediately, [00:49:29] I think that opens up the game for big people already spent so much money on the GPUs. There’s going to be a next wave of GPU launched coin like we saw in the beginning of 2019. Grinn was absolutely failure, but I think that just kind of trigger that wave again. Is it fair to say that miners will be making coins just to have them to be mine as we move

Leo: [00:49:52] in the future? Because I feel like GPU money has been on for so long, it’s been resilient. For the past eight years, nine years, if not right around that time, it’s almost solving a problem. We’re creating a problem and solving it with some of these new coins is what I’m hearing you say. Oh, absolutely. It’s just like you said, the economics at play and we’ve seen with crypto [00:50:09] since you’re able to just copy, paste and about building community, there’ll be the opportunity to paint that picture however you want. If it’s Mimblewimble, which was that other algorithm reports or if it’s Ethereum. One of the last questions I have for you, Lira is, where can our listeners connect with you online? What’s the best way to stay in touch?

Leo: [00:50:27] Yeah, so I’m on Twitter and I have a blog called AnishaResearch.tech. I don’t write very often, about once a month, but every time I publish it, I definitely make sure it’s spent a lot of time on it and also it’s polished over. Definitely check out the blog and then is there anything else, Lira, that you wanted to talk about? Yeah, so I think overall mining is at the crossroad. [00:50:48] I think it’s moving to a very interesting direction. Like you said earlier, the third having really changes things, whether consciously or subconsciously for people’s mental model for revenue. I think two major trends that I’m seeing. One is obviously further industrialization, which has been happening for quite some time. I’m specifically talking about ASICS, Bitcoin specifically, and two is

Leo: [00:51:09] financialization, whether it’s liquid hash rate or financial services built on servicing miners to help them mitigate risk. I think these are two biggest directions. The former definitely is a little bit harder to play because that requires a lot of CapEx and a lot of convincing people to show up with money and definitely requires smart people to really understand how to design these [00:51:29] data centers. Whereas the second trend, I think it’s going to have a lot of players, a lot more fragmentation and it’s going to take a long time for people to figure out what exactly is going on. I think that’s definitely the trend for ASICS mining, for GPU mining. I think it’s also at a very interesting place. Several months ago, most miners were hesitant to deploy, to purchase cards and

Leo: [00:51:51] because the East 2 is such a huge uncertainty and overall Ethereum community is not that friendly to miners. But just this DeFi that started this explosion of transaction fees and just the recent prosperity made people change their attitude and decide, okay, we’re going to do ASICS. So yeah, I think these are big observations at the moment. In changing the foundational [00:52:17] layers we have here today, anything out of left field regulation? So I think some of the conversation has started to become a little bit concerning and on the emphasis that I’m not a super ideological person. But I do think that regulation placed on the wrong level can be potentially damaging to the industry. There’s a lot of talk about mining pools needs to be KYC

Leo: [00:52:37] that I think that’s probably going to happen. The question is when, but I don’t know how they’re going to enforce that. There’s still a lot of open questions, but what they can possibly do is require all the large mining entities, marathons, and I wouldn’t be surprised if that already happened, require them to only go through KYC mining pool or they have to do all sorts of disclosure. [00:53:00] I think there’s definitely a lot of wiggle room for individual miners, smaller miners with just a couple of machines because they can’t really enforce you to put your hash rate on F2 pool. There’s no way to enforce that. But for large players, for large industrialized players, I think that process is definitely happening as we speak.

Leo: [00:53:20] And if that’s the case, do you see decentralized pools coming with stratum 2.0? Do you see they’re being an opportunity to build almost like a DeFi decentralized pool where there isn’t no information given or no really central authority? It just is like a protocol? Yeah. So first of all, even without all these things happening, I think stratum V2 is very [00:53:39] exciting. And I think that is absolutely the right direction for mining pool software to move to the stratum. Stratum V1 just have some funny problems, even without this minor delegation problem. So I don’t know if that necessarily further incentivized development in that because the regulators can simply just require the large mining players to force them to use centralized

Leo: [00:54:02] KYC pool. And they can’t say, oh, we’re just going to join a decentralized pool. And then they get rested in the middle of the night. But I do think that when stratum V2 becomes more mature, I think new generation of pools, although I don’t know why would anyone build a new pool at this point, new generation of pools will start to take advantage of that stratum version one that’s been here [00:54:24] since they got introduced and hasn’t changed. Now we’re trying to push for more privacy, more improvements in big proponent as well for stratum version. Thanks to the guys that slush

JohnPaul: [00:54:32] over there for pushing that and everyone else. Leo, thanks again for coming on the show. It was

Leo: [00:54:36] great talk. I really appreciate it talking about how not all hash powers created equally. These capital markets based on hash rate. And it was really interesting to hear you say that the facility is the mine. And because as you know, it’s not just the miners that going to make the hash

JohnPaul: [00:54:50] rate and keep the sustainability, but it all comes down to the overall environment. So thanks again

[00:54:54] for coming on. Remember to mine on. I hope you enjoyed today’s episode of digital gold. Be sure to subscribe so you’re notified when the new episode drops. Don’t forget to leave us a

Leo: [00:55:03] five star review to support our journey to become the number one crypto podcast. Thanks so much

[00:55:08] for listening. And until next time, mine on.