Article
NYDIG
Exploring Bitcoin Valuation Methodologies
For traditional asset classes, there are widely accepted ways for estimating future values. To value a stock, open up a handy discounted cash flow model1 and plunk in your assumptions. For bonds, do a little math2 and a bit of credit analysis. Want to know what a house is worth? Look up the comps.
Of course, nothing is inevitable, and each technique can have personal touches that result in widely different results. Indeed, valuation is more art than science. But at least there are guideposts and general rules of thumb.
When it comes to bitcoin, there's no consensus on how it should be valued, but that doesn't mean people aren't trying.
So far, attempts to predict where bitcoin is going have pulled on techniques used in various asset classes. That isn't surprising. People who have come to bitcoin from the world of finance see bits and pieces of their former lives in the new asset class. Former stock analysts see the network effects so prominent in the biggest tech companies. Some bond traders price it as if it were insurance against sovereign currency defaults. Investors in precious metals think stock-to-flow models are the way to go.
In this article, we take a look at some of the ways for estimating bitcoin's future price. Ultimately, only time will tell which approach will prove best. With hindsight, perhaps none will provide an accurate forecast. But by considering the different methods, we might establish a framework for mapping what is undoubtedly a new frontier.
Network Effects
With no cash flows to analyze, one way analysts value bitcoin is by measuring network growth. Here's the idea in a nutshell. First, take bitcoin's price and see how it has historically tracked user growth. Then, assuming user growth is a leading indicator, you can forecast how the price will react as the network adds more participants.
Many analysts rely on an insight from the early internet boom to make these predictions. Metcalfe's Law posits that the value of a network grows exponentially rather than linearly with each additional user. Intuitively that's a tricky concept to grasp, as anyone who has read Malcolm Gladwell's The Tipping Point can attest to, but here's a simple example that brings it to life.
Imagine the beginning of the telephone. With just one phone, the invention is worthless. With two phones, there's only one possible connection. But as you get above three phones, each new line creates more than one new link. So a four phone network has six possible connections, and a five phone network has twelve.
In a 2018 paper,3 Timothy Peterson showed that the correlation between bitcoin’s price and network growth was more than 80% over medium to long-term time horizons. Explaining his methodology, Peterson wrote:
“Traditional currency models fail with bitcoin, but various mathematical laws which explain network connectivity offer compelling explanation of its value. Our purpose in conducting this research is to examine bitcoin’s price as a function of the network effect.”
Similarly, NYDIG Global Head of Research Greg Cipolaro found that “the number of Bitcoin addresses squared explains 93.8% of the variation in the level of Bitcoin's market cap.”4 Cipolaro explained:
“Given our view that, as an emergent successful money, bitcoin’s fundamental value derives from its network effects, bitcoin’s value should roughly adhere to Metcalfe’s Law. Building on and simplifying prior research in the area, our analysis shows that bitcoin’s valuation is well described by the most fundamental factor intrinsic to its network: the number of addresses that hold bitcoin. This may be an important insight for investment professionals who, understandably, require anchoring around a fundamental valuation framework as a necessary component of their allocation diligence and analysis.”
Building a model based on user growth isn't as straightforward as it may seem. One complication is determining how many individuals use bitcoin. For example, an address could be for an exchange, in which case it represents countless users. On the flip side, a single person can have multiple addresses, which would lead to overcounting.
Stock-to-Flow
Whereas network effect-based models draw a comparison between bitcoin and the internet, other models highlight bitcoin's resemblance to precious metals like gold. For example, gold has been revered by humans since the dawn of civilization for its physical properties: it is lustrous in appearance, scarce in supply, malleable into items like coins and jewelry, and ductile for industrial use cases. But like other commodities, the supply of gold can be modulated by its price. Even so, the supply of gold has grown 1% – 2% per year, that is to say exponentially, over the past 100 years. While bitcoin does not share all of the physical properties of gold, it does share scarcity, but takes it a step further. Bitcoin's predetermined supply schedule is insensitive to changes in demand.
With this knowledge in hand, some analysts use stock-to-flow models for forecasting bitcoin's price. Stock-to-flow is simply a measure of the rate of production versus existing supplies. A higher stock-to-flow ratio typically corresponds to a higher value, advocates contend. Currently, bitcoin's stock-to-flow multiple sits around 60, just shy of gold's 62. But because bitcoin production is cut in half every four years, its multiple will soon surpass that of gold.
While stock-to-flow models are among the most cited in the bitcoin community, they’re also widely criticized. The model, critics say, takes for granted that scarcity alone leads to higher valuations. Others point out that the approach has failed to work for other digital currencies when it should be just as applicable. Still, others object because models based on stock-to-flow yield unrealistic predictions. Taken to their logical conclusion, as the block subsidy drifts to zero, stock-to-flow models expect bitcoin’s price to converge on infinity.
On-Chain Analytics
Bitcoin is famously an open ledger. Anyone can inspect the economic activity that occurs on its blockchain. Because of that, there is a wealth of open-source data with which to make measurements and observations in a new and growing field called on-chain analytics. This field attempts to quantify the fundamentals of the Bitcoin economy, often creating unique time series analyses and ratios. This can all happen in real time, giving a glimpse into the network in a way that is unprecedented for other technologies or assets.
“In the same way that a government statistical agency publishes data about a country’s population and economy, or a public company publishes quarterly financial statements disclosing growth rates and earnings, Bitcoin provides a real-time, global ledger that publishes data about the network’s activity and inner economics,” Ark Invest research analysts wrote in a report.5 “In the absence of central control, Bitcoin’s blockchain provides open-source data, its integrity a function of the network’s transparency.”
However, there are limits to what you can learn from blockchain data. Identities on Bitcoin are pseudonymous, meaning we can’t precisely know specific actions by an individual or organization. The data only allows for educated guesses. For example, Bitcoin’s blockchain style requires the movement of all of one's coins, even if the intention is to just move a portion. The portion that the sender wishes to keep, known as the change, is a best guess by blockchain analytics companies.
Comparable Market Analysis
Since bitcoin's earliest days, people have tried to peg its value against the markets it could disrupt. A week after the first block, Hal Finney, one of the first bitcoin developers, talked about how to value the new currency:6
“One immediate problem with any new currency is how to value it. Even ignoring the practical problem that virtually no one will accept it at first, there is still a difficulty in coming up with a reasonable argument in favor of a particular non-zero value for the coins,” he wrote. “As an amusing thought experiment, imagine that Bitcoin is successful and becomes the dominant payment system in use throughout the world. Then the total value of the currency should be equal to the total value of all the wealth in the world. Current estimates of total worldwide household wealth that I have found range from $100 trillion to $300 trillion. With 20 million coins, that gives each coin a value of about $10 million.”
Of course, not everyone is using the world's total wealth to come up with a valuation. The most common addressable market targeted by analysts is gold. A recent estimate pegs the total value of all the gold in the world at about $12 trillion.7 Then there are others like negative-yielding bonds8 or pricing bitcoin as a credit default swap against sovereign debt.9 If bitcoin grew to just the size of the current global gold market, the price per coin would rise to more than $500,000.
A comparable market analysis helps layout possibilities, but it's only relevant over extended time frames. Thus, while few would criticize taking such an approach, the chosen market will always be subject to debate. What might seem like a reasonable comparison today could be considered absurd tomorrow.
Marginal Cost of Production
A common criticism is that bitcoin has no intrinsic value, but just because it doesn’t have cash flows to discount doesn’t mean it doesn’t have value. Bitcoin, unlike fiat currencies, can’t be produced with a simple keystroke. Instead, creating bitcoin requires computational work, and that involves using electricity. With this fact in mind, analysts have valued bitcoin in relation to its cost of production.
In a 2018 paper, Adam Hayes of the University of Wisconsin-Madison, examined the relationship between bitcoin’s price and its cost of production.10 Hayes found that bitcoin’s price tended to “fluctuate around the model price.”
“This finding is striking given the volume of recent media accounts and research projects that have supposed no fundamental value at all for bitcoin,” Hayes wrote. “Moreover, it suggests that attempts to find a causal link between the long-run price of bitcoin and various exogeneous factors may be misguided, as well as attempts to value bitcoin as if it were a traditional financial asset.”
A marginal cost of production model might best measure the floor rather than the fair value since it doesn’t consider market demand. For example, commodities don’t trade at their cost of production.
But unlike commodities, the bitcoin mining rate is completely insensitive to price. It is governed by a programmatic supply function that reflexively adjusts to the network processing power to continue to produce new bitcoins every 10 minutes. The implication of this is that the cost of production model might have the relationship backward. Bitcoin’s price arguably drives the cost of production, not the other way around. The idea being that higher prices attract more miners which leads to the difficulty adjustment being ratcheted up, making mining a more expensive endeavor.
