There is a hope that cryptocurrencies have massive growth ahead, at least in usage if not in price. There’s hope that they’ll revolutionise banking and liberate man from the tyranny of fiat currency.
However I can’t leave this crypto energy handbrake thing alone. It seems such an obvious restriction that I cannot see how cryptocurrencies can become really mainstream in their current form.
Refresh: how does bitcoin work?
Every time a financial transaction happens, it is written down in a ledger. Some money gets added to one account and subtracted from another. In the normal world, a centralised institution like a bank writes the ledger and makes sure the money doesn’t end up in two places at once. In the crypto world, Satoshi Nakamoto devised a system to safely log a transaction on a distributed ledger shared across a network with no single centralised institution in control. To do this:
- A transaction happens and a competition is announced.
- Nodes on the network do a Piece of Work (completing a mathematical task).
- The node that gets the answer first wins the right to write the transaction in the ledger (add the block to the chain).
- As a reward the winner receives some cryptocurrency, which is created from nothing specifically for the purpose (mining).
- The difficulty of the Piece of Work is altered automatically to control how much new cryptocurrency is created.
What is the handbrake?
The Piece of Work acts as a handbrake on cryptocurrencies.
- Whoever wins the competition has to spend A LOT of energy on processing power to win the competition.
- Whoever doesn’t win the competition HAS ALSO spent a lot of energy on processing power.
- The more valuable the reward of currency is (if for example bitcoin has gone up and up), the more people will compete to win it, and so the more the winner will have to spend on energy to win, and, of course, the more energy will have been wasted by those who don’t win.
- The more transactions there are, the more such competitions will occur.
So in sum, the more valuable the currency becomes and the more it is used, the more energetically expensive it becomes.
(If you’re so inclined there’s more detail from Amy Castor on how bitcoin works, and Noah Smith on how energy use changes with price).
The cryptobull’s response
If you present this problem to a cryptobull they respond with two points:
- “Yeah it takes a lot of energy, but the global banking system uses even more, you’ve got to compare it to what it’s replacing”.
- “Actually it’s a pretty good way of finding the most efficient energy. You know they build farms next to hydroelectric dams to use energy that would otherwise be wasted”.
I thought maybe they were right, perhaps I was missing something and the handbrake wasn’t a real problem. So I looked into the arguments a bit more. I’m afraid neither stands up to scrutiny.
1. They say bitcoin’s energy usage doesn’t look bad compared to banking
I found a few articles discussing this point, all from a cryptobull perspective.
- Ark Invest say bitcoin consumes less than 10% of traditional banking.
- Most other discussions point to this 2018 article by an ‘energy efficiency startup’ called Zodhya, which seems suspiciously similar to an earlier article in Hackernoon. It sums branch, server and ATM energy estimates to get a figure of 140 TWh per year for banks. They estimate Bitcoin consumes 32 TWh per year, so less than 25% of banks.
- Then this Galaxydigitalmining article from few days ago says 240 or 263 TWh per year for banks (including card network data centres), 240 TWh for the gold industry, and 114 TWh per year for Bitcoin, so approaching 50% of an expanded definition for banking.
Are the comparisons reasonable?
Let’s start with the Zodhya numbers. Bitcoin.com shows there were around 15k bitcoin transactions per day in 2018 (it’s about 200k per now). In contrast, Statista shows the top five brands of payment cards process 1.2bn transactions per day in 2019 (which probably wasn’t that different from 2018).
In other words, very roughly, there were 80,000 times more payment card transactions than bitcoin transactions every day (1.2bn / 15k). And this is just consumer payments with cards. This is before we add in the huge numbers of bank transfers, B2B transfers, international trade remittances etc.
So whatever the real number is, we can be confident that there are MANY TENS of THOUSANDS more transactions each day for traditional banking than for bitcoin. For the sake of argument lets use a conservative estimate of 100,000.
So if bitcoin were to successfully replace the banking system as the cryptobulls hope, and energy usage increase was linear, our new system would take 32 TWh (Zodhya’s number) * 100,000 = 3.2m TWh per year.
This is a huge and totally unrealistic unfeasible number. To put it in context, it’s nearly 1000 times greater than the total energy consumption of the USA.
This is to replace something currently estimated to use 140 TWh, i.e., it’s over 20,000 times more energetically expensive than the current system. Further that 140 TWh is not a fair comparison because it includes things that bitcoin can’t do, like issuing physical cash in machines, and running the systems to approve loans, and all the stuff banks do that isn’t just transactions. In contrast, as Amy Castor reminds us, “nobody actually uses bitcoin to buy anything, except for the occasional large oil company that needs to pay off extortionists taking over their computer systems”.
If we use the more recent Galaxydigitalmining numbers it looks even worse. A 5:1 ratio for the combined energy of all banking, cards and gold production compared to that of bitcoin, when the ratio of what the former actually does compared to the latter is still going to be tens if not hundreds of thousands to one.
Also bear in mind, these numbers are just for bitcoin. Add in ethereum and the hundreds of other currencies and the picture gets even more ridiculous.
Or, to put it another way, in 2018 the cryptobulls estimated bitcoin used under 25% of the energy of banking. In 2021 they estimated is used under 50% of an expanded definition of banking. They estimated this while arguing energy usage isn’t something to worry about. But bitcoin is only used for a miniscule fraction of the things banking is used for, and if it becomes more widespread the energy usage will absoutely dwarf that of old-school banking in weeks.
To leave nobody in any doubt about the conclusion: the energy usage of cryptocurrencies is INSANE compared to that of the banking system.
2. They say bitcoin uses energy efficiently
They say bitcoin can absorb structural surpluses of ‘stranded’, ‘nonrival’, or ‘curtailed’ energy. This is energy that is wasted at production, because it can’t be stored, or during transmission.
I’m not entirely certain why bitcoin mines can succeed in harnessing this when others haven’t, but if they can, fantastic.
Even so, the amount of stranded energy isn’t growing very fast. If crypto is going to become used in any way comparably to fiat currency, energy usage would quickly exceed the amount of ‘stranded’ energy available, rendering this defence invalid.
The handbrake cannot be removed
I’ve discussed this questions with a few people now. Nobody’s yet been able to explain it away for me. The attempts I’ve found online look like attempts to bamboozle the reader with excess information. They do not address the key point that:
Bitcoin and other currencies are massively energy expensive compared to their usefulness, and this will only get worse if they’re used more. They have a handbrake that cannot be removed.
Now obviously Musk is just trying to move the market, but even he’s saying Tesla won’t accept bitcoin payment any more because of the energy toll, and that he’s working with Dogecoin to improve transaction efficiency.
And this is the second key point:
The proof of work that is integral to maintaining the integrity of the distributed ledgers underpinning cryptocurrencies is the source of the handbrake.
Decentralised digital currencies have incredible potential, but this potential cannot be released until some Nakamato 2.0 invents another protocol that keeps them legit but doesn’t cost the earth.
While I’ve been banging on about this for a while, I’m glad to see some mainstream articles saying similar things now. Most closely this piece on Bloomberg.
I don’t think this is going away.
Today, bitcoin is a third off its peak price.
I’ll leave you with Visual Capitalist’s illustration of bitcoin’s power consumption.
