The economics of cryptocurrency
Governance
Cryptocurrencies do not have central banks to regulate the money supply or oversee financial institutions, but no one should neglect the importance of cryptocurrency governance institutions. We focus our discussion on two separate but interrelated ways that cryptocurrencies can be said to be governed.
Algorithmic governance
Rules for what are considered valid cryptocurrency transactions are embedded in the peer-to-peer software that cryptocurrency miners and users run. One valid kind of transaction is the creation of new coins out of thin air. Not everyone can execute this kind of transaction – miners compete for the right to execute one of these transactions per block (on Bitcoin, every ten minutes or so). When a miner discovers a valid hash for a block, they can claim the new coins.
A transaction in which a miner claims new coins, like any other transaction, has to conform to the expectations of the network. The network will reject a block that contains a transaction in which a miner awards themselves too many new coins. The growth of coins is limited by a pre-determined amount per block. On Bitcoin, the pre-determined amount is not scheduled to be constant over time, but rather is set to halve every 210,000 blocks, or about every four years, as described above. The total supply of bitcoins will asymptotically approach, but never exceed, 21 million. It will reach 20 million in 2025 and stop growing altogether in 2140.
Open source governance
The astute reader will note that the Bitcoin software that enforces particular rules about valid transactions and the rate of money creation does not appear out of thin air. Rather, the rules embedded in the software emerge from an interplay between leaders of the open source project that manages what is known as the ‘reference client’, other developers, miners, the user community and malicious actors. The dynamic between these players is as crucial to understanding Bitcoin as that of central banks, traditional monetary institutions and monetary politics is to understanding fiat currency. Bitcoin, like all other even moderately successful cryptocurrencies to date, is a non-proprietary open source project. Users tend to look with suspicion on cryptocurrency projects that are closed source, that feature significant pre-mining in order to reward insiders, or that have other proprietary features. Other expectations of the user community also impose a check on developers. For example, the hard cap of 21 million bitcoins, while in principle subject to change through a software update, appears to be non-negotiable for Bitcoin, although other cryptocurrencies have different money supply rules.
The division of Bitcoin software into a ‘reference client’ and so-called ‘alt-clients’ also has implications for Bitcoin’s evolution. The community looks to the Bitcoin Core team for leadership as to the direction of the network. An alternative approach would be for the community to agree on the specification for the network, and then let independent teams write clients that implement the specification. The fact that Bitcoin has such a dominant reference client means that evolution can occur more quickly, although it may also have hidden costs. For example, the community has to put a lot of trust in the Bitcoin Core developers not to make bad changes to the network. A less concentrated approach to cryptocurrency development would slow down development, which would prevent any changes to the network without full deliberation of the community. It’s possible that over time Bitcoin could move more to this model, but for now, the advantages of rapid evolution might outweigh the costs.
Miners also play an important role in governance. Because miners cryptographically guard against double spending, their consensus on what counts as a valid transaction is necessary for a cryptocurrency to function. A majority of miners must adopt any change to Bitcoin, and therefore the miners are able to impose a check on developers. Miners also exert influence through mining pools. Miners join pools in order to earn a more consistent payout. A single miner working alone might go for some time without discovering a block. But if miners pool their work and split their rewards, they can earn daily payouts.
Mining pools raise complications. For example, the biggest Bitcoin mining pool often has a third or more of the computing power of the Bitcoin network. If a pool ever obtained more than half of the network’s computing power, it could double-spend. Double spending would destroy confidence in the Bitcoin network and would likely cause the price of bitcoins to plummet. Consequently, we observe some self-regulation by the mining pools, which are heavily invested in the success of Bitcoin. Whenever the top pool starts to approach 40% or so of computing power of the network, some participants exit the pool and join another one. So far this norm has persisted, but many in the community are concerned about mining pool concentration. Recently, the GHash.IO mining pool briefly exceeded 50 percent of Bitcoin’s mining power. There is no evidence that the pool used its position to double spend, but many observers were alarmed that it was able to happen. Concentrated mining pools have benefits as well as risks. In a crisis, it is useful to be able to assemble the key players. Such a crisis occurred on the night of 11 March 2013, when it became clear that a change in version 0.8 of the reference client introduced an unintentional incompatibility with version 0.7. As a result of the incompatibility, the two implementations of Bitcoin rejected each other’s blocks, and the block chain ‘forked’ into two versions that did not agree on who owned which bitcoins. Within minutes of the realisation that there was a fork, the core developers gathered in a chat room and decided that the network should revert to the 0.7 rules.
Over the next few hours, they were able to confer with the major mining pool operators and persuade them to switch back to 0.7, sometimes at a non-trivial cost to the miners who had mined coins on the 0.8 chain. The fact that mining pools are relatively concentrated meant that it was relatively easy to coordinate in the crisis. Within about seven hours, the 0.7 chain pulled permanently ahead and the crisis was resolved.
Another problem occurred in February 2014 when Mt. Gox, the oldest and largest Bitcoin exchange, claimed that its bitcoin holdings had been depleted through ‘transaction malleability’ attacks. Although it remains unclear whether Mt. Gox losses were really due to attacks, it became clear over the next several days that misunderstandings about transaction malleability were creating vulnerabilities. Some Bitcoin sites temporarily suspended withdrawals while the issues were addressed by the core development team, which updated the Bitcoin software and helped educate the community about transaction malleability, which, when properly understood, is a feature of Bitcoin, not a bug.