Ending the Bitcoin energy drain
Bitcoin uses energy. Lots of energy. It does this because as an incentive to ‘run the network,’ the Bitcoin protocol sets up a proof of work mechanism whereby to certify a block on the blockchain you need to win a computational contest. The great news is that this is a contest that everyone can have a chance of winning. Back in the olden days (circa 2009), that reward was 50 bitcoins and that added up to about 15 cents. These days the reward is 12.5 bitcoins or about $250,000. Suffice it to say, while being a miner in the olden days was all about the revolution, today it is all about the money. It just isn’t the same.
Now that it is a commercial enterprise, the amount of electricity used to run the system is given by:
Prob{Win per Unit}*Reward = Average Resource Cost per Unit
Because someone is going to win, what this means is that:
Reward = Total Resource Cost of the System
As a block is mined every 10 minutes, that means that over the course of a day, today, the Bitcoin network will cost $36 million. Over the course of a year, that comes to around $13 billion. By contrast, in 2010, the same calculation would have led to an annual cost of around $8,000. So in the course of the lifetime of Bitcoin, the costs of running the network have gone from basically free to something akin to the GDP of all about two thirds or so of the world’s national economies.
This is a little different than the ‘bottom up’ calculation by Digiconomist. The logic of my calculation is based on the simple idea of ‘rent dissipation.’ That is, people will spend enough in a contest to dissipate all of the rents. So even as more entry causes the computational difficulty of the Bitcoin system to rise (it is designed to adjust so that a block is mined every 10 minutes), rent dissipation will still apply.
Not all of this is energy consumption. There is also the cost of the chips and machines designed to enter the Bitcoin contest. In the olden days, anyone with a computer could mine for Bitcoin. Then it evolved to GPUs and now to specialised ASICs which I understand have no alternative use case other than mining. Having an edge in technology will allow you to earn inframarginal rents.
The other way to earn rents is to arbitrage on energy usage. Bitcoin is global. You don’t have to search for a gold mine. What you search for is a place to mine where electricity is cheapest. Then you face no transportation costs. It is the purest form of energy arbitrage we as a species have invented to date. If it continues, it has the capability to equalise the price of electricity world-wide.
The thing about all of this is that there is no mechanism to stop it. Bitcoin was designed to be decentralised and robust to attack. So long as the Bitcoin price (its exchange rate with USD) remains high, the resources will be spent. If the Bitcoin price rises, the resource expenditure will rise accordingly. So if we reach Bitcoin $100,000, the resource cost will rise to $65 billion (putting it equivalent to the 80th ranked country in terms of GDP). This is terrible news because we are getting nothing for this casino.
One piece of good news is that the party will end by about 2032. At that date, the cap of 21 million Bitcoins will be minted and so the reward to the running the system will rely on transaction fees. Suffice it to say, one would not expect that to be much more costly than running the Visa network — which is a fraction of these costs.
Can we stop all this much sooner? The quickest way to do that would be to undermine the Bitcoin system so that people lose confidence and the price falls back to below $250. At that price, the resource cost would be about $165 million per annum. Absent a concerted international government attack, it is unclear how to do that.
Bitcoin enthusiasts offer three other ways out of this jam. The first is to change the Bitcoin protocol to move to a non-resource intensive mechanism. But who would change that? In effect, the miners would have to support a change (or a fork) that produced this. But the miners are invested in the current casino. That is not a recipe for change.
The second suggestion is better computational technology so that mining can be done at much lower energy cost. The problem with that is that it would not change the fact that the resource cost will equal the reward. If you have more energy efficient technology, miners will deploy it to dissipate any rents. In other words, the main incentive to produce such technology is solely ‘business stealing’ so as to grab more rewards from other miners. Importantly, this means that investments in such technologies are purely socially wasteful.
The final way is through competition. Someone will invent a better cryptocurrency that is not as resource intensive. People will realise that this is a better way to go than Bitcoin and will switch. This is another way of saying that Bitcoin’s price will crash through competition and that is how it will be replaced. This idea, not surprisingly, is not pushed hard by either miners or those who are banking on a return to their Bitcoin holdings.
It is also not that plausible. If Bitcoin was all about the currency and the transactions, then this force of competition would make sense. But it isn’t. Bitcoin is currently competing as an asset class. It’s value is driven by people thinking it will have high value. How can someone just compete with that? It isn’t clear.
In summary, we are going to look at the Bitcoin era as an important innovation running off the back of a big design mistake. It all could have been avoided had Bitcoin minted its 21 million coins right from the start or with a much shorter expansion. We may have seen a speculative run as we do now but it would have occurred without the additional resource cost that is just completely unnecessary. Yes, transaction costs might have posed an issue but they would be bounded by an economic force at least. Mining just isn’t.
