A moderate take on cryptocurrency

With Bitcoin prices back up high, people are again variously hyping or disparaging cryptocurrency. I don’t often see moderate takes on the technology, probably because it’s extremism that gets attention, but I thought I’d share my take.

Should Bitcoin Be Valuable?

I agree with the many skeptics that Bitcoin cannot be a currency. The value has always fluctuated greatly and had long periods both inflation and deflation. Denominating prices or contracts in Bitcoin injects a lot of risk into a transaction due to this volatity. For now, this is a non-starter.

But, Bitcoin can replace gold as an asset. Yes, gold has industrial uses that Bitcoin mostly does not (setting aside timestamping value of the blockchain), but Bitcoin, assuming the protocol and code is secure, offers an asset that is definitely finite: there are at most 21 million Bitcoins. This provable scarcity alone has some value. Since Bitcoin is more convenient to hold and transfer than gold, it could take over much of the role gold and other metals play in investment. At the end of 2019 there was about 198k tonnes of gold mined, about 77k tonnes in investments and holdings, which could be valued at about 5 trillion USD. If Bitcoin gets 10% of this market, you would expect a Bitcoin to be worth about 24k USD (close to current values).

Next, it seems like Bitcoin could actually provide value in the space of settlement of large transactions and cross border transactions. Banks or corporations looking to make large moves, especially across borders, could buy in, make the trade, and then get out. So, the amount of float of active transactions will put additional demands on Bitcoin which will drive the price above the price set by the e-gold use case.

Between the two use cases of hedge asset (e-gold) and conduit for very large transactions, Bitcoin should have some value significantly greater than zero. In addition to these use cases, there is at least the promise that cryptocurrencies could be useful in machine-to-machine protocols that are impractical today (see Filecoin as a way to run a distributed p2p Dropbox alternative).

Could Cryptocurrency Have Bigger Impacts?

The chief objection to cryptocurrency is the cost of operating proof-of-work. The whole point of the protocol is to be expensive to run. So, fundamentally, transactions have to be expensive, or you can have double-spending. This concern is serious, however, it seems the transaction costs may be low enough still to enable the e-gold and large transaction market to operate. To make bigger impacts, cryptocurrency has to address at least two more problems: lowering transaction costs, and reducing volatility in value. There are several approaches that could solve both of these problems.

How to Reduce Costs?

Two promising approaches to reducing transaction costs are: 1. moving transactions off-chain, 2. moving away from proof-of-work. The lightning network is an approach to use the Bitcoin blockchain as a settlement layer for a network of debt transactions. Here is a sketch of the idea: to send money from A to B, there would be paths found through the lightning network that would update their debt logs. If someone makes bad on the debts, the commited data on the blockchain allows one party to reclaim funds. Since a transaction with the blockchain is only needed for a small subset of operations, the cost is greatly reduced and privacy is increased.

In addition to moving off chain, there are several proposals to move cryptocurrencies away from proof-of-work. One interesting approach is proof-of-stake (PoS). In PoS, nodes that hold more assets have more say in setting the next item in the log. Tezos was an early adopter of this approach and Ethereum 2.0 will use a version of PoS. There are many other proposals as well (e.g. Stellar’s consensus protocol). Additionally, lightning network and similar proposals are generally orthogonal to non-PoW approaches so both can be combined.

It seems quite likely that efficient cryptocurrency can dramatically reduce costs and be competitive with current centralized banking systems.

How to Deal with Volatility?

The second major problem with cryptocurrencies is volatility. We don’t want to set prices, agree on rents, borrow money, etc, in an asset that fluctuates wildly. Again, there are at least two solutions to this problem. The most obvious solution is to piggy-back on existing stable assets: coins that are backed by one stable currency or a basket of them. In this picture, trustworthy actors are audited to prove that they have a 1:1 backing of some coin to some currency asset. Tether is an example of this, however it is not at all clear that they are sufficiently audited and can be trusted. Facebook’s Diem (previously called Libra) is based on this idea: coins are backed by currencies, baskets of currencies (e.g. half a dollar and half a euro), or short term government securities.

A potentially more interesting approach is Dai. Dai uses smart-contracts on Ethereum. Smart contracts allow the assets on the blockchain to become programmable. Dai uses smart contracts to make a market that tends to push price of Dai towards 1:1 exchange with USD. The protocol is a bit complex, and has had a couple of hiccups, but has generally performed well (extreme events around 5% deviation).

So, like the problem of costs, it seems that fluctuation in value is also a solvable problem.

Should We All Be Crypto Bulls?

Why do I say this is a moderate take on crypto if I have outlined so many positives? Briefly, I love technology, and I am generally a techno-optimist. Technology is exciting and it is breath-taking how much it has advanced even in my own lifetime. That said, much of cryptocurrency work smacks of solutions looking for problems and outright fraud.

For instance, American consumers don’t seem to have any problems that crypto can solve, except maybe moving money a bit faster. But the slow pace of money movement in the US can, and likely will, be solved by upgrades to the technology that the banking industry uses. Next, crypto fans often say how great it is that crypto transactions can’t be reversed. But, reversible transactions are a feature, not a bug. American consumers demand this from their credit-card providers, and without it they would be much more reluctant to use credit cards (trying selling a credit card that does not permit reversing transactions). These features will have to be added on top of cryptocurrencies by providers, or by smart-contracts to compete, in my opinion. Aside from the e-gold use case, most of the wins of crypto today probably are limited to those living with poor national currencies, or those often doing expensive cross-border transactions.

Secondly, a technology can be valuable, have a future, and still be packed with scammers and speculators. The Internet is an important and impactful technology, but that doesn’t mean that many or even most of the companies in the late 90s were not scams trying to exploit hype and many investors were simply gambling on finding a greater fool. It could be that literally all of the current cryptocurrencies will fail, yet the idea still makes an impact over the next decade. I think technologists should be interested, but I am very concerned that retail investors are simply gambling and that many of the companies are knowingly fraudulent.

Finally, I’m excited about the prospect of cryptocurrencies and distributed finance. I think they will make an impact over the next ten years, and yet, I would discourage any non-expert from investing any funds in anything cryptocurrency related.

Machine learning, Programming, Hadoop, Scala.