Bitcoin’s steady price gives clues to the secret of its valuation
Months ago, certain similar models predicted both flat prices and a $6,000 range for bitcoin.
The week ending 26 October was bitcoin's most stable since the start of the year, writes Kevin Davitt, senior instructor for the Options Institute at CBOE Global Markets in CBOE's latest weekly bitcoin futures commentary.
"If we look at the weekly range over the course of October, it's a mere 6.6 percent, which is far and away the lowest monthly average," he writes.
A swift lack of movement
This trend is quite unmistakable if you look at bitcoin prices in the year to date.
It's not entirely unprecedented though. The better part of a year ago, two separate studies that used similar valuations models made the following predictions:
- Bitcoin would be settling in for months of sideways price movements, sloping generally downwards to a more fair price
- A fair price for bitcoin was about US$6,000
Prescient, no? Both used a Metcalfe's Law-based valuation model as their foundation. This theory gained traction as a form of valuing bitcoin in 2017, when put forward by Fundstrat analyst and bitcoin bull Tom Lee.
This theory, proposed in the 80s, is a way of valuing networks based on the number of participants. Naturally, it's entirely imperfect because you're basically just swapping out the variables until you find something that works.
For example, are bitcoin participants measured by the number of miners, wallet addresses or transactions? How do you account for users with multiple addresses? Are bitcoin prices driving participant numbers rather than vice versa? And so on. The theory was proposed in the 80s, long before bitcoin was even a twinkle in Satoshi's eye, so it needs tweaking until it fits previous observations.
Here's a brief rundown of the two separate models that made these retrospectively remarkable predictions.
Metcalfe's Law + LPPLS model
This is the one used by those chaps from the University of Geneva, who in March predicted that bitcoin would be going sideways for many months, trending generally downwards. The full study is here (PDF).
This version essentially mashed together their version of Metcalfe's Law for bitcoin and a separate theory called the Log-Periodic Power Law Singularity (LPPLS) model.
For their take on Metcalfe's law, they used bitcoin wallet addresses as an imperfect proxy for network participants, while acknowledging that people's tendency to hold multiple wallet addresses, bitcoin's wealth inequality and other factors were all making it quite unscientific. The reason they just used it anyway, they said, was because it worked really, really well even though scientific gut feelings said it shouldn't.
They then combined it with the LPPLS model.
The LPPLS model is a theory for predicting financial bubbles and busts. There are also a lot of different ways to imperfectly apply it. It works by acknowledging that there are two types of traders: the rational ones who use charts, algorithms and big words like "fundamentals" for their trades, and the so-called "noise traders" who tend to form herds and drive much of the market action. The idea is that real-world elements, such as irrational exuberance, mean bubbles are a natural part of any market.
The bigger the bubbles are, the theory goes, the more fragile it is, and the more likely you are to get a "straw that broke the camel's back" effect.
In the case of the big bitcoin bubble of late 2017, the University of Geneva researchers' suggest that straw was simply the fact that bitcoin busted the psychological $20,000 barrier. Despite ostensibly being good news, the bubble was fragile enough that this made it all crash back.
So, by applying these two models (Metcalfe's Law and LPPLS) in tandem, they suggested the following back in March:
- Bitcoin was still overvalued, based mostly on the number of bitcoin wallets and network participants
- The bubble of over-valuation was quite firm though, and wasn't in any foreseeable danger of an abrupt crash or boom one way or another
Put those two factors together, and you can predict months of sideways prices trending loosely downwards. That's exactly what happened.
If you want to roll with that, the most important question then might be how one goes about finding a fair price for bitcoin.
The $6,000 range
In January this year, Credit Suisse analyst Damien Boey reckoned $6,000 was a fair price for bitcoin. His model also used Metcalfe's Law, but aimed to compensate for its shortcomings by correlating it with other real-world factors.
Metcalfe's Law + Credit spreads
In this case, Boey used a version of Metcalfe's Law that incorporated average bitcoin transaction value and number of wallets, then aimed to account for its imperfections by searching for factors that improved its accuracy and mashing them together.
The idea is that Metcalfe's Law strongly correlated with bitcoin prices by itself, so other sensible factors that increased its accuracy theoretically should be playing a role in bitcoin's price movements as well. In this case, Boey found that adding yield spreads of BBB-rated bonds to the mix made the model more accurate than just using Metcalfe's Law alone. Ergo, it looked like bitcoin prices correlate partly with yield spreads of BBB-rated bonds.
"In the extreme, this could imply that leverage has been used to fund Bitcoin investments through time. Alternatively, it could simply mean that bitcoin's valuation is highly sensitive to whatever has driven credit spreads over the past few years."
So, what are all the factors impacting those credit spreads at any given time? Good question. Please inform your local reserve bank if you know the answer.
Either way, adding this factor to the Metcalfe's Law model produced a formula that saw a fair bitcoin price, at the time, being somewhere around $6,000, which is close to where it's settled this month.
How does bitcoin move?
Cryptocurrencies and other digital assets are notable for their ability to remain largely uncorrelated with other asset classes. This is one of its key value propositions as an investment, and why there's a solid mathematical case for holding bitcoin and other cryptocurrencies for diversification, regardless of how one feels about blockchain, hype and dark-web currency.
That bitcoin prices are maintaining their unprecedented stability against relative turbulence in traditional markets might suggest that bitcoin is largely free from the influences of other markets, despite previous correlations.
It raises a few questions though. What happened to its previous correlations with other markets? And why is bitcoin settling in at this particular price range?
Bitcoin has found a few different correlations in traditional markets over time, beyond models like Metcalfe's Law plus credit spreads. One of the more compelling recent correlations is between bitcoin and emerging markets.
But Travis Kling points out that by hanging tight, bitcoin and the wider crypto markets have diverged from the emerging markets index that it had previously been following.
Over the last month, BTC and Crypto markets broadly have hung in INCREDIBLY well in the context of traditional asset class stress. Not making a call on whether or not this will continue, just pointing out the recent divergence. pic.twitter.com/gDhYjUORSe
— Travis Kling (@Travis_Kling) October 29, 2018
Elsewhere, analysts have found inverse correlations between the US dollar and bitcoin, posed theories about the impact of the futures market on bitcoin prices and wondered at the market consequences of whales and potential manipulation.
By all appearances, it looks like bitcoin itself is something of a morphing asset, changing over time and slotting into different niches over time. Or maybe all these correlations at different points in time are just meaningless statistical noise.
Opinion: What's a bitcoin worth?
A bitcoin is worth precisely US$6,463.17 at the time of writing (6 November 2018), although your mileage may vary depending on where you look, where you buy and where you sell. Plus, that's just how much people will generally buy or sell it for, rather than how much it's "worth."
The only fair answer to the question so far is a congenial shrug.
But here's another theory to try on for size, should you feel so inclined:
The solid line that ends up on top is bitcoin's hashrate, and the spotty line is bitcoin's price. Both are smoothed out with a 30-day simple moving average for that manicured look and because it makes the theory look more compelling.
Note how they both really flattened out at about the same time from September onwards.
This is in keeping with the theory of bitcoin as a commodity, where the floor price of an asset is the cost at which it can be obtained. In this case, it would suggest that bitcoin miners are roughly breaking even at current bitcoin prices, which is why neither hashrate nor prices are moving a lot.
The idea is that this has to be the floor price as long as there's some organic demand in the market because if production is not profitable, production slows and demand relative to supply increases which drives prices back up. Of course, bitcoin "manufacturing" will continue regardless of the hashrate, so that might not work quite so neatly and demand for the coin is driven largely by price and speculative future value. But it does suggest that there's a close tie between mining profitability and bitcoin prices, although it's perhaps just the hashrate responding directly to bitcoin prices rather than vice versa.
Still, it's perhaps as neat a model as any.
The only thing anyone can say with certainty might be that bitcoin's valuation is still a bit of a mystery, and there are a lot of potential pieces to the puzzle. Just like most other economic theories.
Disclosure: At the time of writing, the author holds ETH, IOTA, ICX, VET, XLM, BTC and ADA.
- 7 tips for Bitcoin and crypto beginners
- Tether to launch a stablecoin pegged to the British pound
- Today’s crypto movers: Bitcoin (↓4.70%), Ethereum (↓6.27%), ApeCoin (↓9.75%%), Waves (↑45.04%)
- Ethereum’s low price provides a discount for anyone long on the Merge
- Bitcoin price stabilises as tech stock correlation drops to new low