What is Cryptocurrency Mining? What are Mining Rigs? How do miners make money and are they consistently profitable? What is Proof of Stake and how does it differ from Proof of Work? Read this transcript of an interview with a seasoned investment manager and Cryptocurrency proponent, Murray Stahl, and learn more about the economics of Cryptocurrency Mining.
Horizon Kinetics is one of the first conventional asset management companies to get actively involved in the field of Cryptocurrency and Cryptocurrency mining. Since 2015, the company lead by Murray Stahl and Steve Bregman has been publishing research on Cryptocurrency.
Last year, they started publishing podcasts in Q&A format, under the name Consensus Money. In the third episode of Consensus Money, published in April 2018, Murray Stahl answers question on the intricacies of Cryptocurrency Mining. The podcast is delightfully insightful and lays out the economics of cryptocurrency mining in simple and understandable terms.
What follows is a transcript of the podcast. We have changed the text slightly for a better reading experience. You can listen to the full Q&A session on the dedicated Consensus Money page on the Horizon Kinetics website.
- Horizon Kinetics has removed the Consensus Money Podcasts from their website.
What is Cryptocurrency Mining?
Hugh Ross: Welcome to another episode of consensus money, a podcast series with Horizon Kinetics about cryptocurrencies. I’m Hugh Ross, COO of Horizon Kinetics and today I’m joined by Murray Stahl, the chairman and chief investment officer of Horizon Kinetics. In today’s episode, we’re going to discuss the topic of Cryptocurrency Mining in general and also Horizon Kinetics’s approach to mining. So Murray, with that, why don’t we talk a little bit about mining? Can you tell us what Cryptocurrency Mining is?
Murray Stahl: Okay. Well, Cryptocurrency Mining refers to the process of validating transactions on the blockchain. So, essentially the so-called miners are really operating servers, which are endeavouring to validate transactions that are happening over, what’s called, a block interval. A block is basically the transactions that have happened over the course of 10 minutes and the validation process are things like making sure the public keys and private keys match and have the privilege of writing that to the blockchain.
You’re trying to solve an equation, a branch of mathematics known as elliptical functions. It is not really important what elliptical functions are, but more that people are competing to solve that problem. And whoever’s the first person to solve it within a 10-minute span, or the first entities, get the block reward. Which, at the moment, is 12.5 Bitcoin, in the case of Bitcoin. In the case of other coins, there are different block amounts of awards.
Different Types of Miners
Hugh Ross: Gotcha. So, I know there are different kinds of miners and mining rigs. Can you perhaps explain the difference between what an ASIC and what a GPU rig is and what does ASIC stand for? What does GPU stand for and why do you even need different kinds of rigs?
Murray Stahl: Okay, ASIC stands for application specific integrated circuit. GPU stands for graphics processing unit. Basically, the coins, the logic behind the Cryptocurrencies were designed to operate on different machines. And that’s why you need different kinds of machines for different kinds of currencies.
Now, with ASIC you’re using that for currencies like Bitcoin and Bitcoin Cash. GPU is used for currencies like Etherium, Etherium Classic, Zcash and oddly enough, Bitcoin Gold and there are some others as well. So you need different kinds of machines to essentially mine or validated transactions of different kinds of Cryptocurrencies, depending on how they were designed.
Owning Rigs vs Cloud Mining
Hugh Ross: Right. Okay, perfect. What is the difference between – you hear the term cloud mining a lot and then when you own the rigs, you’re mining with your own rig. So what’s the difference between mining with your own rigs and cloud mining?
Murray Stahl: Well, essentially what you’re doing in cloud mining, is that you’re leasing capacity. The same way as you want to store music or pictures or data on the cloud. Basically, you’re leasing someone else’s equipment and you’re not responsible for the operation of the equipment whatsoever.
You pay a certain fee and you get a certain revenue. And presumably, the revenue you get is greater than the fees you pay. The idea is similar, in the case of the mining that you do yourself, except you are responsible for the operations of the machines. There is no one to turn to other than your own faculties
Proof of Work vs Proof of Stake
Hugh Ross: Right. Perfect. Okay. Now, a couple of terms that are out there that I think it is important that people understand. So, there is what’s called Proof of Work and there’s Proof of Stake. Perhaps you could explain to everybody what is the difference between Proof of Work and Proof of Stake?
Murray Stahl: Okay. Well, it can be a little convoluted. So, to begin with, Bitcoin was the first cryptocurrency. They wanted to have a certain issuance rate from its inception – which was roughly in 2009 – until the last Bitcoin was going to be mined, in the year 2140.
The idea was to make the issuance rate a little bit analogous to gold. So, why was gold, for the course of 5,000 years, frequently used as a currency? Because you could always get some gold. But it became progressively harder and harder to find gold. So, the amount of gold that entered the monetary system in any given year was, generally speaking, a small amount.
The idea of Proof of Work that came up, it was to make you do some work to validate the transactions. Basically, you have to get a rig, you have to consume copious amount of electric power and operate equipment and so on and so forth. And basically, you have an operating cost. The idea was to put an impediment upon issuance, (a) so it would be a controlled amount, and (b) if you set up a program like Bitcoin did, there would be a certain difficulty rate.
Difficulty refers to the probability of solving these equations or elliptical functions. And that can be actually fairly predictable. You can have a predictable issuance rate to the year 2140. Which means you will know what the inflation rate is going to be up to the year 2140. Beyond 2140 there is no inflation whatsoever because there is no net currency issuance whatsoever. And that would be discounted in the price. That’s the idea of Proof of Work.
Proof of Stake came up or was invented because a lot of people thought the Proof of Work concept is wasteful of things like electric power. It uses copious amounts of electric power. Some people feel it’s bad for the environment. So they understand that you want to put an impediment upon instruments.
So, instead of using electrical power and calculations, the idea was to actually put up Cryptocurrency and Cryptocurrency, presumably, is valuable. You put up a Cryptocurrency and the block awards are allocated randomly, pro rata, based on how much cryptocurrency you put up.
The only Proof of Stake system which might be issued, of which I am aware, is Etherium, which is currently Beta testing their Proof of Stake system . We’ll see if it actually operates or not.
Hugh Ross: Do you have a view of whether one is better than the other?
Murray Stahl: Well, no one has actully operated Proof of Stake. So, it has a good side and a bad side. Theoretically, or practically speaking, we can’t make an assertion because we don’t know Proof of Work. We’re all familiar with or we know how it works.
Here is the good side of Proof of stake and I’ll give you the bad side. So, we mine Etherium with GPU servers. Let’s assume that Proof of Stake is a valid concept. From our point of view, we’ve held a certain amount of Etherium that we´ve mined. We continue to mine every day.
In theory, we could put up that asset and then liberate the servers that are currently mining Etherium to mine something else, like Zcash or Etherium Classic, for example. Actually turning our Etherium into a doubly productive asset. It appreciates or has appreciated thus far. That’s one vector of productivity. And then we could actually earn more Etherium.
So theoretically, if Etherium has, let’s say, somewhere between a 14 to 15% inflation rate and 10% of all the people who own Etherium then put up Etherium for the Proof of Stake. That 10% would own the entirety of the inflation rate. So, the inflation rate spread among the currency is 15% but spread among the 10% who prove a stake, 150%.
So, theoretically, if 10% of people chose to participate in the Proof of Stake and we were one of them, we would earn a 150% rate of return, or nearly so, on our Etherium. Which would be pretty good. That’s the good side.
The bad side, in theory, is that if everyone thought that was so alluring, and you can clearly see – and this is why I said the good side first – if everyone saw that there are certain people that are making 150% on their Etherium cryptocurrency by just leaving it in a wallet – or in some cold storage or on the server – then a lot of people might want to do it.
So, you can have a situation where if enough of people who own Etherium posted for Proof of Stake, you can come very close to demonetizing all the system. Meaning Etherium wouldn’t circulate. Everybody would just, it would be one of these Gresham’s law situations. People would rather have the return than actually transact in the currency. So it can be very reflexive in the sense that it created a bad situation. No one knows what’s going to happen because no one has tried it.
Hash Power Supply and Demand
Hugh Ross: Could there be a massive amount of Hash Power that shifts from Etherium to, for example, Zcash, in the event that we go to Proof of Stake? Because then what do they mine? They’re not be mining any Ethirioum on it anymore.
Murray Stahl: Yes. That could well happen. And even more the case, there might be GPU equipment right now that’s still being used for Etherium that is nearly obsolete. And if it moves to a currency like Zcash, Zcash has a much lower difficulty rating. So, what might be is that that equipment might be much more productive in a Zcash mode. Or maybe, there are some other forks of Zcash that have even lower difficulty ratings, where it might be doubly productive.
A lot of interesting things can happen. So from our point of view, we might be able to take our Etherium-based servers and move them to something like Zcash and increase the productivity of our mining activities. Might be very interesting.
Hugh Ross: Gotcha. As I know that you know, there was recently an ASIC miner released by Bitmain for Etherium and I wanted to just touch on how you think having an ASIC miner – which in theory is kind of a next-gen miner – will impact the mining of Etherium and Etherium Classic? How do you think this whole ASIC versus GPU hardware will play out and especially in the context of Etherium?
Murray Stahl: Well, the GPU miners clearly wouldn’t like that. And why wouldn’t they like it? Because if there is ASIC equipment that can mine Etherium as easy using that currency, it’s going to get a certain market share and it reduces productivity. It will make the GPU equipment worth less. Not worthless in the sense of zero but worth less money than it otherwise would be.
So, the natural impulse of the GPU miners is to fork the currency in such manner as it comes ASIC resistant. You can’t use it for that. Of course, it’s not necessarily a robust defence. Because just like you can fork the currency to be ASIC resistant, you can fork Etherium to be ASIC tolerant or ASIC sympathetic. Because all the code is open source. So you might end up having an Etherium fork which is ASIC compatible and you might have an Ethirium fork, rigid, ASIC resistant as possible.
Hugh Ross: Which is crazy considering Etherium has already forked once.
Murray Stahl: That’s right. So, it could theoretically happen. We just don’t know.
Cryptocurrencies without Mining
Hugh Ross: Right. Okay. So just for a kind of a compare and contrast: There are Cryptocurrencies that are not mineable. I thought we should touch on one or two examples of that, like Ripple. So if they are not minable and you, in theory, to quote-unquote, operate the system, in the world of Bitcoin, for example, there is an incentive to operate the system, right? Because you’re getting mining rewards. So what’s the incentive to operate the system in a Cryptocurrency where there is no mining?
Murray Stahl: Okay. So in order to discuss a Cryptocurrency with no mining, we have to introduce a little bit of new terminology. It’s not hard to use. It’s not hard to understand. There’s the dichotomy between a permission system and a non-permission system.
A non-permission system is like Bitcoin. If you have the money and go out and buy a server, you to can mine Bitcoin. So a non-permission system means; anybody who feels like it can take part in mining.
A permission system is; the group of people who are currently operating it have to collectively give their consent to admit you to join the network. If they don’t give you consent, you can’t operate the system. Now if they give their consent, you have an operating system, you have a cost of operating the system, but there’s no block reward to give out. So then the question becomes, to get back to the original question, why would you have the operating expenses and not have any reward whatsoever?
So, Ripple otherwise known as XRP was introduced to be a substitute to what is called the SWIFT network. Basically, it’s a global bank money transfer system. It’s actually not even a money transfer system. What it really is, is a notification that money is going to be transferred. So it’s cumbersome, it’s slow, it’s difficult to use. And then it also has the faculty that it operates in a variety of currencies.
So the idea behind XRP or Ripple as it’s sometimes called is that there will be this one currency. And you could pay any bank you wanted to in that currency on behalf of any correspondent, or any client. So it eliminates a lot of currency transactions and also eliminates a lot of accounting expenses. Because you’re only accounting in one currency as opposed to 180 currencies that are in the world. And there’s also, you don’t need trading in 180 squared pairs caused by any of the 180 currencies can be traded versus another currency.
So, if someone in Argentina wanted to pay someone in Sweden, you have to make a market. Argentine Pesos for Swedish Kroner. And that in itself is a great deal of work. So what you would be doing is – if you’re a bank or a global brokerage system and you are participating in this – you’re offering the service to your clients.
Presumably, it’s going to be faster and it’s going to cut your expenses and cut your errors. So, on the one hand, your expenses go up because you are operating the system. On the other hand, your expenses are going down for a number of reasons.
You’re not operating the SWIFT system anymore. It’s displacing it. You’re not having all of the accounting problems and tracking problems working in 180 currencies. And you also don’t have to maintain what’s called in banking Nostro and Vostro accounts. So, Nostro and Vostro accounts refer to kinds of accounts that balances other currencies you have to maintain in different countries.
So, when you’re paying someone in another country, on behalf of one of your clients, what’s really happening is that they are taking the money out of the account you maintain with a foreign bank and then eventually you reconcile with your client and get paid. There is enormous accounting expense and you’re also leading balances around the world in different currencies. And some of them may shift in value and you have significant hedging expenses.
Hugh Ross: So, so using Ripple as an example, it sounds like there are financial or economic incentives to use that currency, that don’t revolve around money. You don’t necessarily need money, but there are just, basically, another set of incentives.
Murray Stahl: There are other incentives. You can create other incentives to do it. And it is possible. There’s one other incentive. It’s minor, but it may not be a minor one day. The minor meaning m-i-n-o-r as opposed to a miner. Minor meaning of little significance right now.
There’s a tiny insignificant transaction fee for transacting in Ripple or XRP. That transaction fee doesn’t go to a banker or a brokerage firm. It just currency, it gets burned. So, Ripple started out with 100 billion units. Now it has 99.9 billion units. It does no transactional activity. The supply of the currency may decline a lot. So the higher the velocity of the currency goes, the faster the supply of money is going to shrink. And therefore, all things being equal, the greater the value of the currency itself.
Hugh Ross: So, why don’t we turn to Horizon Kinetics’ approach to mining. Why don’t we start off just by asking you when and why did Horizon Kinetics first get involved in mining?
Murray Stahl: When we got involved was late April, early May of 2017. Why we got involved was because it seemed like a very robust return was available. But it wasn’t only that. This plays a role in asset allocation.
So, mining in asset allocation should be known as Seniorage. Seniorage is basically the profit you make from printing money. So, historically when the United States Bureau of Printing and Engraving creates a $1 bill, it costs them 5.4 cents to print up the $1 bill. So they get a dollar out of it. The profits of mining aren’t quite as good, but they’re very robust. So it’s a way of earning a profit.
But it also plays a role in asset allocation because when you mine, the value of your equipment is being depreciated over the course of three years, in accordance with generally accepted accounting principles. So assuming you hold back the cash for depreciation, you’re unit value, month by month should be basically stable. What’s going to change is your profit. Or your dividend, if you decided to pay it out. Because the Cryptocurrency fluctuates.
- Kunav Crypto Research: Is Crypto Mining a Future Asset Class?
So it plays a major role in asset allocation because of the problem right now with bonds, which are much more stable than equities. The lower the interest rates go, the greater the convexity you have in the bond portfolios. So if rates were to go up, you might take mark-to-market losses in bonds. And you don’t get a high yield anyway.
You get a much higher yield by mining cryptocurrencies and you also have a stable unit value. And that is real diversification versus publicly traded securities, that don’t have stable valuations. So I think it’s very important to asset allocation. I eventually think it’s going to be a major-sized asset class.
Choosing What to Mine
Hugh Ross: That’s great. So obviously I know we mine a bunch of different Cryptocurrencies such as Bitcoin, Bitcoin Cash, Litecoin, Etherium, Etherium Classic, Zcash and so on. What is your approach to or how do you think about optimizing the mix of Cryptocurrencies that Horizon Kinetics is mining, in terms of profitability?
Murray Stahl: Well, the returns fluctuate dramatically. Week by week, month by month. So, basically, we’ve weighted them more or less in accordance with the inflation rates. So Bitcoin and Bitcoin Cash have roughly 3.9%. I think it’s actually 3.94% inflation rates and that’s much lower than Litecoin right now. Litecoin right now I believe, right now, is over 10 per cent.
So, we mine much more Bitcoin than Litecoin and you could make an argument of saying that we really should do the reverse. Because the increase in value in Litecoin when the inflation rate drops, as it will about one of the third years. From let’s say 10 to 4.9% or whatever the number happens to be. That might lead to a greater percentage increase in value than you might get in Bitcoin. It’s possible, but at the moment we decided to do it the way it is.
Also, it’s arguable that Bitcoin Cash has a potentially greater profit than Bitcoin because a certain amount of mined Bitcoin Cash is going to be burned, very much like the way it’s burned in Ripple. So the supply of Bitcoin Cash will ultimately then be lower than the supply of Bitcoin. And Bitcoin is going to have a better monetary policy.
So all these things need to be taken into account. And there are some other Cryptocurrencies where when they forked – for example, Ethireum Classic, Bitcoin Gold, Bitcoin Cash, there are some others – it’s not clear that very many people got their forks. And they may be out there in the ether and no one will ever have access to them. So the supply of money, it may be much less than we think it is.
So all those things have to be taken into account and we’re always reevaluating our mix. At the moment GPU miners are uniquely expensive. It might have something to do with the idea that Etherium is going to move to Proof of Stake, and therefore you might be able to get alternative uses of the equipment. That said, we just don’t know. So at the moment, we have had, at least for the last couple of weeks, we haven’t really bought any GPU miners.
Hugh Ross: So, I know a lot of miners out there definitely look at profitability. Let’s say, hour by hour, minute by minute and they’re probably shifting their resources around from one Cryptocurrency to another, pretty frequently. What’s your approach to shifting resources from one Crypto to another? Do you do it often? Do you do it daily? Do you do it weekly or what is your approach to that?
Murray Stahl: Well, we haven’t done it often yet. The reason is that we are still building up our supply of service. So, one way to do it is when we’re making the next order. That’s a way of shifting your mix. So, if I had a lot of ASIC equipment and I wanted some GPU because I had in mind to mine Zcash, let’s say. Well it’s very easy, I’ll just order some GPU servers and one order is ASICS servers and now my mix is going to change.
The prices as they have been – as anybody who’s looking at the currency exchanges would realize instantaneously – is that the prices fluctuate enormously. Therefore the profits fluctuate enormously. Sometimes in the space of hours, sometimes even the space of minutes. And I think one of the reasons they fluctuated is people are constantly making these shifts. They try to optimize.
And I think when they make these shifts, that’s part of the reason these things fluctuate as much. So I decided I didn’t want to add to the variability by trying to outguess everyone else. So I’ve left it alone more or less and we have been very happy with the result.
Investing in Mining Equipment
Hugh Ross: Gotcha. I’ve certainly experienced and I know you have, that at times equipment can be hard to come by. And I think there are people out there, that take the approach that when they raise capital to invest in mining equipment, that they’re basically just going to go all in when they raise that money. I know you have a different approach. Maybe you could explain why you think it’s maybe better to have more of a prolonged, shall we say, approach to acquiring mining equipment?
Murray Stahl: Okay. Well, the first point is just a minor point. When you have a go-all-in approach you buy a tremendous amount of equipment at one time. You have to believe that you are actually impacting the price of equipment. So you want to buy a lot. The supply is not changing based on what your motivations are at that minute. Because the mining equipment manufacturers have no idea what you’re going to do. And so from their point of view, these are totally random events. When you put in a big order, you might actually make the price higher, not lower.
More importantly, however, over time the chipsets are going to improve. The software is going to improve. Everything about it going to improve. And if you believe in Moore’s law, eventually the chipsets will generate less heat. It will use electricity more efficiently and the price is going to come down as the manufacturers learn how to make your chips get higher yields.
There is a certain rejection rate when you’re making these ships. So the lower the rejection rate, the more profitable a given fabrication batches and the better a deal they can get for you. So if you put in an order for everything at once, you have no ability to react to changes in technology and changes in pricing.
So now, sometimes, like what happened recently, ASIC miners dropped in price. In our case, we didn’t invest that much money up front. We were investing gradually. So, now we have cash flow as well as the initial capital to buy many more miners at lower prices than we had before. You just have to assume that you’re going to get a better deal in future than you have been getting in the past.
Scale versus Flexibility
Hugh Ross: Got it, that makes sense. So, in terms of physically housing the equipment or locating the equipment, do you think it’s better to, basically, put everything in one large facility because of Economies of Scale or do you think that it maybe makes sense to choose multiple locations to house all your equipment. And if so, why?
Murray Stahl: We’ve taken the approach of housing our equipment in a variety of locations and we are actually interested in increasing the number of locations. And the reason is you never know what’s going to happen. There can be electrical outages and given the area is totally beyond your control, there can be natural events that are totally out of your control. You can have one hosting facility that just decides it doesn’t want to operate the facility anymore. You have to move your equipment.
There are a variety of things are going to happen and there’s always the possibility of disruption. So if you have all your eggs in one basket and you watch that basket very carefully, no matter how carefully you watch it, something might happen to put all your servers out of operation simultaneously. You don’t want that because they are not productive until you (a) find another hosting facility, (b) negotiate proper terms, (c) move and install all your equipment, which may be cumbersome.
Especially if you have a lot of servers and so on so forth. So I don’t think it’s a good idea to put everything in one basket. Just as I don’t think it’s a good idea to order all your servers in one fell swoop, which apparently everybody wants to do.
The Cost of Mining
Hugh Ross: Gotcha. And then in terms of the of the income statement of a mining operation, what are the biggest components of expense when you’re running a mining business?
Murray Stahl: Well, there’s really two, and I’ll break them down. One is the hosting expense. The hosting expense is, basically, that you’re paying for the electric power marked up by some suitable quantity and you’re paying essentially rent to put your machine on a shelf and you get a little oversight. That someone at least is looking at it, making sure it’s operating properly. And there’s at least some security in it. That somebody is making sure the facility isn’t broken into. If somebody does something untoward, they will call the appropriate authorities and so on so forth. So that’s one expense.
The other expense is depreciation on equipment. So basically you have to assume that in 36 months, your equipment’s is going to be fully appreciated. Meaning that it may work, but there will be more productive models out there. So, you’re not getting a high return on it although they might have some salvage value. So your depreciation is really a function of how much you pay for the machine. So essentially you’re using machine ratably over its useful life.
And those are the two primary components and some other minor expenses of operating a company. And it applies to every company. There is a tax return and there are certain filings, administrative expenses that every business has and cryptocurrency mining operators are not immune for that, but they are relatively minor. Those are the two significant costs.
Hugh Ross: So let’s talk about the revenue and the net income side of the financial statement. The profit margin associated with crypto mining, is that something that’s fairly consistent or does it vary? What causes it to vary? Can there be times when mining actually becomes unprofitable and what happens then?
Because, if you think about it, using Bitcoin is as the example, the miners have to have an incentive, right? To operate the system. Their incentive is they’re making money from mining. So, if mining becomes unprofitable because of either the cost of machines going up or the price of bitcoin falls enough, what happens? If the miners are not incentivized to operate the system, what happens to the whole system?
Murray Stahl: Okay, so to begin with, profit margins are not consistent. They vary and they vary very greatly. And what makes it vary the most? Well, the most important thing is the price of Cryptocurrencies is constantly fluctuating. So, as a generalization, the higher the price of the Cryptocurrency, all things being equal, which they never are, the better it is for you. If your profit margins lower, it is the worse it is for you.
Of course, the other variables vary as well. The cost of the equipment can go up it or it can go down. The electricity prices don’t seem to vary all that much, but they can. At least in theory. And then you could also have some machines that don’t last the requisite 36 months. As with any piece of equipment, something can go wrong with it. Although, that doesn’t happen all that frequently. So, those are all ways of varying the profit margins.
Then you might say: Whoa, what causes all those things to vary? Look at it this way. If you want to have a given cryptocurrency, Bitcoin or Etherium, you basically have two choices. You can buy it or you can create it by mining it. Those are ways you can obtain a crypto.
So let’s look at the example of the end of 2017, beginning of 2018. There were, I think, well over a thousand initial coin offerings. A lot of people compare them with Initial Public Offerings. They have a salient difference from our perspective and the mining perspective, in as much as, if you want to buy one, they don’t want fiat currency. They don’t want Dollars or Pounds or Swiss francs, Euros. They want crypto. So you had a whole series of people that wanted to participate in the so-called ICOs, initial coin offerings, that went out and bought Etherium and bought Bitcoin.
They had no interest in owning Etherium. They had no interest in owning Bitcoin and therefore they were pricing different buyers. Bitcoin and Etherium were merely modalities by which they paid for initial coin offerings. The only way to participate was to buy one of those two currencies. The orchestrators of initial coin offerings, therefore, received either Bitcoin or Etherium. In some cases, both.
They have no interest in holding Bitcoin or Etherium. It’s merely to fund some project. In most cases to fund their cost of creating the next generation of Cryptocurrency itself and making themselves very wealthy. So, in almost all instances, they were still in Cryptocurrencies.
So you had a few months when there was enormous pressure to buy. Virtually everyone was buying and then there was an enormous pressure to sell because they want to liberate this capital and put it into their coffers. You might say there were some traders that would look at these trends and maybe added a little fuel to the fire. Or you could debate whether that happened.
So, there were times when cryptocurrency mining went to break even. It happens. But it’s not going to stay there very long because cryptocurrency mining is designed to equilibrate. So let’s just say that it was unprofitable for a number of weeks. Well, most of these companies were very poorly capitalized.
So they can’t operate without profitability. They didn’t have huge cash reserves to operate unprofitably. What they would do to save cash is that they would turn off their machines. Which you can do in 30 seconds. If they turn off their machines, what happens is that the difficulty rating? Which you will recall, is the probability of solving this equation. If they turn up their machines, the difficulty rating goes down because there are fewer machines trying to solve the problem.
So when difficulty rating goes down and you leave your machine on, your machine necessarily becomes more productive. The probability of earning a coin goes up. So ultimately the whole thing is designed to equilibrate.
So, I’ve never really seen it being, maybe I’ve seen it for a day or two, get modestly unprofitable. Of course, it really isn’t moderately unprofitable, even so. Because when people calculate the cost of mining a coin, remember they are adding in the hosting fee plus the depreciation. The depreciation is not a cash expense. So on a GAAP basis, you might not be profitable but on a cash basis, you’re very profitable.
So in theory, you don’t have to take your accumulated depreciation expense and buy another machine if you don’t want to. It’s entirely a discretionary item.
The Least Efficient Entity
In any event, the way cryptocurrency mining works, it is unlike normal business activity. Because remember, this is the most important point. This is a permissionless system. And what makes cryptocurrency valuable are two things: a network effect, which is all people using it and the robustness of the network, which is the difficulty of hacking it, which is a function of how many people are actually operating it. So the more people operating it, the more robust it happens to be.
Now all those people don’t have the same cost structure. In normal business, there’s going to be some company that happens to be the most efficient company. So in theory, whatever good they’re producing, whatever service they’re providing, in theory, they could lower their price to a level where that individual enterprise makes a very high return on investment capital, but nobody else can match it. So, they basically drive everybody out of business or nearly so and they become the dominant player.
And therefore in most businesses, a dominant player sets the ceiling on return on equity or return on capital. So that’s the highest and nobody can compete with that. And that’s why most businesses have a dominant company or sometimes two dominant companies.
In cryptocurrency, if that happened, no one would use the cryptocurrency because it is hackable. So what needs to happen is, to make it robust, a lot of entities or a lot of people if you prefer, have to be operating it. Meaning mining it.
So the return on invested capital or return equity has to be sufficiently alluring so that the least efficient entity, whichever entity that is, has to make enough of a return to be interested. Everybody else is, by definition, more efficient than the least efficient entity. So everybody else is making a higher return.
So unlike the situation in all business, where the most efficient company sets the ceiling, in the world of Cryptocurrency, the least efficient company or entity or person, sets the floor on return on invested capital. Everyone else is going to make a higher rate of return.
Hugh Ross: Okay. Well, um, we are, we’ve been talking for over 35 minutes so I think we should wrap it up now. So I want to thank Murray so much for having this discussion with us. And thank you to our listeners. If you have any questions about what you’ve heard today, please reach out to us on email@example.com and please stay tuned for future episodes of Consensus Money. Thanks, Murray. Thank you very much.
Murray Stahl: Thank you.