Saturday, August 15, 2020

Old School Analysis—Cryptocurrencies: Commodity Dynamics and Cartelization

From American Affairs Journal, May 2018:
Bitcoin and the other altcoins now have more “experts” than perhaps any other market. I am no such “expert”: I am neither a cryptographer nor a computer programmer. I am a currency and commodity trader of thirty-plus years, and I approach the cryptocurrency market from that perspective.
The cryptocurrency market cannot easily be dismissed, despite the fact that several high-profile people and institutions have called it a fraud, a Ponzi scheme, and a scam. The altcoin market surpassed $700 billion in market capitalization in late 2017, even though it is about half that now. This extraordinary volatility is caused by the immaturity of the market along with other factors that will be analyzed below.

Most discussions of cryptocurrencies revolve around two core questions: whether the trading values of Bitcoin and other cryptocurrencies are justified, and whether cryptocurrencies’ underlying blockchain technology will live up to its disruptive billing.
Such discussions, however, too often end up as abstract, irresolvable debates between blockchain optimists and pessimists. In contrast to these typical approaches, I will focus mainly on the quantifiable elements of the Bitcoin mining process, not only to suggest a reasonable range of values, but also to explore the larger economic and political issues surrounding cryptocurrencies and blockchain technology. While most commentators tend to emphasize the differences between virtual currencies and physical commodities, I will focus on the similarities. Indeed, it is the similarities to conventional mining that are likely to determine cryptocurrencies’ future trajectory more than their virtual aspects. As is the case with commodity mining (over the long term), Bitcoin mining will continue to increase in capital intensity; for Bitcoin, such a trend is effectively built into the system.
As a result, the shift toward corporate centralization, which has already begun, will likely accelerate. Although cryptocurrencies began under the guise of a libertarian ethos, market and mining conditions will steadily lead them back toward corporate cartelization and state involvement.

Bitcoin and the Blockchain
First, a bit of background on the origins of Bitcoin as well as the distinguishing features of blockchains. Bitcoin (BTC) was created by the person or persons using the pseudonym of Satoshi Nakamoto, who released a white paper on the concept right after the collapse of Lehman Brothers in 2008. The key component of Bitcoin, like those of the other altcoins, is the blockchain. The blockchain is a decentralized ledger. In contrast to many other tech applications—social networks, mobile phone ecosystems, and the like—the Bitcoin network is not administered by or dependent upon a corporate (quasi-)monopoly. In theory, it is the antithesis of one. No central authority is in charge, and its protocols are subject to change with the user base consensus. This is the main promise offered by cryptocurrencies, and, if nothing else, BTC so far has offered solid evidence that a peer-to-peer system can create a secure database: there have been no hacks of the system to date (although participants have been hacked). This fact in itself demonstrates that the blockchain has some value.
In a blockchain ledger, each transaction is recorded by a population of “nodes.” These nodes are the miners, which are required to solve computational puzzles before anyone else in order to receive compensation for their efforts. The miners are the backbone of any crypto system. It is the miners that have to validate every transaction on the blockchain. To do so, they need to build and store all the blocks on the blockchain and then reach a consensus on which blocks make it into the blockchain.
This network of distributed nodes replaces the trust and centralization needed in a hierarchical system. In fact, the Bitcoin network is the most powerful computer network in the world, and this is the true source of its value. The price of Bitcoin is directly correlated with the revenue opportunity for securing the Bitcoin network.

The first thing to ask about any market is whether it can become sufficiently liquid. BTC has achieved this, and the binary risk of its going to zero is for now alleviated: BTC is mainstream. To understand the underlying value of Bitcoin and its blockchain more precisely, however, it is necessary to understand the incentive structure that supports the blockchain. Without appropriate incentives, a decentralized system cannot sustain itself. Although the “decentralized” character of the blockchain has attracted the most attention, the incentive structure for mining Bitcoins is gradually adding crucial “centralized” elements to the cryptocurrency market’s overall character.

Unlike fiat currency systems, under which a central bank or other central authority determines the supply of money, in BTC the total supply is determined by supporting the blockchain, which occurs through the activity known as mining. Mining is governed by an algorithm: everyone knows when each supply of BTC is produced as well as the maximum total quantity, which is 21 million Bitcoins. For BTC, the mining algorithm initially offered a reward of fifty Bitcoins per block until the first 210,000 blocks were mined, and then the reward per block halved. The current reward is 12.5 Bitcoins.
Reaching 210,000 blocks requires about four years of mining, given that each block has been structured to be mined in about ten minutes. The ten-minute block average is maintained by a difficulty adjuster. This difficulty adjuster is changed every 2,016 blocks, or about every two weeks, based on how fast miners were able to mine the previous 2,016 blocks. This block reward system should end around 2140. There are also rewards through transaction fees, which have become a major component of compensation for the miners (though not without creating other controversies).1
Since Bitcoins are produced only by mining, understanding the costs and difficulty of mining is key to understanding the supply side of the equation. A major cost of mining is the computer hardware. Leading-edge chips are needed to compete effectively for blocks. The many similarities to the physical mining business make it quite appropriate that nodes are called miners. At block 512,500 in the blockchain, for example, mining computations become quite difficult and require a significant investment in hardware before one can even become a player in the field. Much like in physical exploration, there are no guarantees that the miner will get a return on this up-front investment. The most important pieces of hardware for BTC miners are ASIC (application-specific integrated circuit) chips, and only the latest, most expensive chips are sufficient for effective competition. Moreover, the lead times on the production of these chips are long and costs are prepaid to the vendors.
Finding the right location for mining operations is also crucially important. A mining location needs three major attributes—a cold climate, cheap and widely available electricity, and good internet speeds. Given the magnitude of computing power involved, mining uses an enormous amount of electricity and produces a lot of heat. To be competitive in the mining sector, costs of electricity typically cannot exceed five cents per kilowatt hour (kWh), and a miner will need a reliable electrical network capable of delivering megawatts of electricity per hour.

Although results can be highly variable, miners can use a Poisson distribution to predict the expected blocks to be mined. The Poisson distribution is used for modeling the number of times an event occurs within an interval of time. A mining operation can calculate the probability of finding blocks over the life of its chips in order to decide whether the operation is likely to be of value. The block probability is directly affected by the global hash rate (the speed at which an operation in the Bitcoin code is completed) versus the amount of hash power the miner has. Usually miners have a set capacity of hash power until additional investments are made in new chips. Miners can thus expect their probability of successfully mining blocks to decay over time.

To reduce block volatility, many miners have joined mining pools. Herein lies a major risk, at least in theory. The top twelve mining pools currently control 13,225 petahashes per second (at the time of writing), while the total network speed is about 20,000 petahashes per second. The top three mining pools control almost 40 percent of the power of the whole network, and the top six control a majority of network power. That is a lot of power in the control of a few top miners. Remember that cryptocurrencies are supposed to be decentralized systems. Even though each pool represents many miners, there is a real risk that decentralization could be undermined, should the top six mining pools initiate a 51 percent attack on the system. The main point here, however, is simply that the power of the system is controlled by a few, given the complexity and cost of mining. Another potential challenge to maintaining BTC’s decentralization is that only two companies dominate the market for the hardware underpinning cryptocurrencies: Bitmain Technologies Ltd., a closely held company in China, and Bitfury, a Georgian company. Bitmain, moreover, is believed to be much larger than Bitfury.

Such growing concentration and cartelization is relevant because everything in the Bitcoin world is done by consensus. When consensus breaks down and some miners want to follow a different set of rules, a fork is created in the blockchain. How much support (via mining power) each fork gets will determine the success of that fork.

Although the miners are the key players in the Bitcoin ecosystem, there are other important members, including the core developers and the buyers of BTC itself. Other components of the Bitcoin ecosystem include those who use Bitcoin for transactions, such as merchants and their respective customers, along with any intermediaries and payment services.

Bitcoin Valuation:
The Limits of Demand-Based Calculations
Since the buyers of or investors in BTC represent the demand side of the equation, we need to understand how they value BTC. As we shall see, these methods cast some light on BTC’s possible valuation, but their predictive value is limited. Some investors in BTC and other altcoins use exponential models to project prices for cryptocurrencies. These models use variants of Metcalfe’s law and Zipf’s law. Metcalfe’s law looks at the change in nodes and squares them (n2). Zipf’s law takes a less aggressive approach, using logarithms and multiplying the nodes by the log of the nodes or (n log (n)). For example, let us assume a network of 100,000 members that generates $1 million. In this example, if the network doubles its membership to 200,000, Metcalfe’s Law says its value grows by (200,000/100,000)² times, quadrupling to $4 million, whereas Zipf’s law says its value grows by 200,000 multiplied by log(200,000) divided by 100,000 multiplied by log(100,000), or to only $2.12 million. Applying the two models to the node growth from last year and looking at the price of BTC/USD, we see the two models imply the following price: One year ago there were 5,625 nodes and today there are 11,714. So using Metcalfe’s Law, the BTC price should have increased by (11,714/5,625)² times or 4.34 times. Last year, BTC was $920. So using Metcalfe’s Law, BTC should be $4,000. Under Zipf’s law, BTC should be (11,714 × log(11,714))/(5,625 × log(5,625)) or 2.26 times $920—about $2,100. Needless to say, these methodologies have not proven especially useful in predicting prices and constitute little more than applying basic formulas to the major unknown variable—network size.

Clearly, expanding the network is extremely important to the success of a peer-to-peer system. Attempting to quantify this, other investors seek to value BTC based on relative values. One such approach to BTC valuation includes comparing the market capitalization of BTC to the narrow money supplies of the major economies of the world.
When BTC hit $20,000, the total BTC market capitalization was about $340 billion. If one BTC equals $10,000, the market capitalization for BTC is about $170 billion. For perspective, the M1 “narrow” money supply of the United States is $3.6 trillion. This puts BTC at 5 percent of the narrow supply of money of the United States.2
The comparison is probably a poor one, however, given that the cryptocurrencies are not really transactional currencies, even though the amount of entities accepting BTC is increasing daily. In addition, the multiplier effect for cryptos is probably close to one as there is no real lending market to date, unlike for fiat currencies....