By Stan Higgins
The University College London’s blockchain research center has found that bitcoin’s current proof-of-work (PoW) mining costs are necessary to maintain the network’s trustless and decentralized nature.
The findings are notable given the longstanding debate over bitcoin’s use of mining for transaction verification, one that has often found institutions and academics deeming the process either excessively wasteful or an unnecessary component of the system’s architecture.
Still, in the note, Aste calls the amount of money spent per block justified when observed through the lens of maintaining a distributed transaction network.
He wrote:
“I conclude that the current cost, although large, is of a justified order of magnitude for an anonymous systems[sic] operating between untrustful parties.”
At current bitcoin prices, a block reward of 25 BTC nets miners or the distributed members of mining pools roughly $15,600. An upcoming block reward halving, scheduled for this weekend, will see the subsidy fall to 12.5 BTC.
Assessing cost
Bitcoin mining has been compared in the past to a form of energy arbitrage, by which mine operators profit by generating more money than they spend on electricity. The cheaper your power – and the bigger your mine – the more bitcoins you can attempt to generate.
Aste estimates that as much as $50,000 is expended on electricity per hour by the world’s bitcoin miners.
Using a standard a transaction block roughly every 10 minutes (which, in practice, can fluctuate significantly), Aste suggests that it costs about $8,333 in power per block. This amount will fall by half imminently, he writes, “leaving very small margins for profit accordingly with the above estimations”.
“The overall mining electricity bill for a year of bitcoin mining sum up to over $400m which is a large amount and, somehow, a big waste,” Aste writes, going on to say:
“On the other hand, proof of work is the mechanism that keeps the blockchain pure making an entire community competing to verify validity of transactions and making attacks costly. The question I address in this note is whether this cost is justified.”
Attack scenario
Aste arrives at the estimate for the fair cost of proof-of-work by envisioning a double-spend attack, using a transaction block with a value of about $1m as an example.
He notes that in order to ensure a successful double-spend, the attacker would need to somehow guarantee that their transactions are confirmed at least six times on the network.
Accounting for the fact that a double-spend would likely go unnoticed – “it is quite unrealistic to assume that nobody notices the propagating fork for such a long time,” he writes – Aste posits that the equilibrium fair cost of bitcoin’s proof of work is equal to the duplicated fraction of a block’s value divided by the number of blocks required to settle, in this case six.
“With the current values, and to make calculations clean, we can assume that the attacker duplicates 60% of the typical value of a block, double spending therefore $600,000,” Aste writes. “Requiring six blocks for settlement this yields to the following estimate for what should be the fair cost of the proof of work per block at equilibrium….$100,000.”
Aste goes on to write that the actual fair cost, accounting for the notion that a double-spend would be spotted early on and that most double-spends would likely be attempted on a smaller scale, is likely 10% of that rough estimate.
“This is indeed the order of magnitude of the present electricity cost for the proof of work in bitcoin,” Aste notes, going on to conclude:
“We can therefore conclude that the current cost for bitcoin proof of work is large, wasteful, but necessary.”