Saturday, December 30, 2017

Pricing Cryptocurrencies Is Basically Impossible

This is NOT investment advice. This is a web developer, with no background in economics or finance, thinking out loud.

With the price of bitcoin skyrocketing, and other cryptocurrencies following it up, there's a pretty obvious question: what are cryptocurrencies actually worth?

Back in the day, Rick Falkvinge argued that bitcoin's killer app was international money transfers, and, therefore, that the total value of all bitcoin in circulation should be equal to the total value of the international money transfer market, which he estimated at 1% of total money in circulation. I liked this argument and found it compelling enough to buy a few bitcoins at $14 (which I then sold at around $60, because at the time I didn't take it seriously). But there are multiple problems with this argument.

The first is that there are many other cryptocurrencies in circulation, and bitcoin's value would have to be determined not only relative to its use case, but also relative to competing cryptocurrencies which were similarly well-suited for that use case. The second problem is that Falkvinge compared money currently in circulation with the total number of bitcoin expected to ever exist. The third problem is that this is basically a streetlight effect. There's no compelling argument that I'm aware of which makes the case that international money transfers are the only use case for Bitcoin specifically or cryptocurrencies in general. Logically, even if Falkvinge's argument is correct — which I don't know for sure — since there are probably other use cases, the total value of all cryptocurrency in circulation should be greater than the total value of the international money transfer market.

And again, that's the value of cryptocurrency in general, not bitcoin in particular.

Another, equally inconclusive way to estimate bitcoin's value is to look at how the market behaves when there are no compelling reasons to suspect that influential pump-and-dump scams are underway. In a YouTube video, economist Neil Gandal charts the impact of two such pricing scams. This was in a presentation given at an academic conference on blockchains at Berkeley in September; bitcoin's subsequent surge in price also arouses serious pump-and-dump questions. If you look at the time period free of pump-and-dump scams, however, you see a range from around $300 to $500 per bitcoin. A person could argue that this is the "true" price of bitcoin, but removing fradulent influences from an analysis does not magically grant omniscience to the remaining market participants, especially not in the context of an emerging technology with untested potential, so this argument is still very flawed.

A further compounding factor in any attempt to come up with a plausible price for cryptocurrencies is that the prices of cryptocurrencies in general go up whenever the price of bitcoin goes up. You might believe, for example, that blockchain applications hold sufficient potential to justify at least some optimism re: cryptocurrencies, and that Ethereum is a much more sophisticated platform for building these applications than Bitcoin. But you don't see the price of Ether go up when the price of bitcoin goes down. The market reflects surges in optimism and skepticism re: the general topic far more than relative evaluations of the different instances of the concept. And it's an apples to oranges comparison anyway; not only do the tokens exist in different quantities, but also, bitcoin has an absolute limit on the number of coins produced (unless the network changes its mind about that), while Ethereum plans to always release new Ether.

With all this, it's essentially impossible to invest responsibly in cryptocurrency, since the uncertainty is so high that there's no way to determine any plausible "correct" price for the stuff. Falkvinge's argument for a million-dollar bitcoin value was basically this: take 1% to 10% of the (loosely estimated) total money currently in supply, worldwide — which he says is $60 trillion — and match it against the finite limit of about 6 million bitcoins that will eventually ever be in circulation. But this is apples to oranges again: a projected number in the future (bitcoin in circulation) vs. a number (very loosely) based on today's economy. And even if this math were flawless, estimates that the current dollar value of cryptocurrency in circulation is $568,661,188,172.

So, using a modified version of Falkvinge's argument, if the total value of cryptocurrency should ultimately be 1% of $60 trillion, well, hooray, the future is already here. One percent of $60 trillion is $600,000,000,000, which is essentially how much money is in cryptocurrency right now — although that's including bitcoin prices which arguably appear to be artificially inflated via a scam involving Tether. Falkvinge's argument included an enormous fudge factor: basically, if bitcoin is ever used for anything above and beyond international money transfer, then its total share of the money supply could go up from 1% to as high as 10%. Although Falkvinge was by far the most disciplined thinker in bitcoin at the time, this is still not a super rigorous approach. He basically pulled that order of magnitude fudge factor straight out of his ass.

So it could be that we've reached the useful limit of cryptocurrency in circulation, and it could be that we're only one-tenth of the way there. It could be that we have the "correct" amount of money in the sector, but it's not allocated to the "correct" coins. Plenty of other interpretations exist, and might be true instead. This is just the degree of certainty we get from the only analysis I've seen that ever even tried to do any math; more mainstream commentators seem to go with "we don't know yet" and call it a day, and that does seem to me to be a more realistic approach. If the only math you can do involves made-up numbers, it's better to admit you don't have enough information to make a precise estimate.

The best lens I've seen for guessing at the values of cryptocurrencies comes from Twitter snark. A popular trend in snarking about bitcoin in particular is to say that it's like a 1990s dot-com stock, except without a company attached. This is supposed to be a dis but it's easily the most flattering thing you can say about cryptocurrencies. If I could have bought Rails stock back in 2005, I would have. To me, the uncertainty in Internet stocks has always been the companies attached; the underlying open source technologies are much easier to evaluate. From that point of view, bitcoin looks like Java; it sucks, but people are going to build a lot of stuff with it. And Ethereum reminds me of SmallTalk; it's trying to build too much of a world around itself to ever really become what it wants to be, but it's still an impressive leap forward.

This is all very much not investment advice. But I am totally qualified to give you advice on how to talk about tech without sounding like a jackass, even if I haven't always followed that advice myself. So here's some advice in that vein: if you're a dev and you're curious about cryptocurrencies, read the whitepapers and look at the code. Although the dot-com crash was overall a very overdue market correction, which cleared out a bunch of trash companies and let the real leaders emerge, at the same time, plenty of complete idiots made a mint during the dot-com boom, and plenty of great technologies tanked in the crash. The only certainty in any of this is that randomness will play a part. But if you know which cryptocurrencies are better designed than others, it'll at least be a useful piece of the picture. It might make all the difference in the world.