Killing Bitcoin for $8 Billion: A Shocking Thesis Questioning the Reliability of Cryptocurrencies

A 51% attack made profitable by derivative markets

Bitcoin relies on Proof of Work, a mechanism where miners mobilize computational power to secure the network. A 51% attack would theoretically involve controlling a sufficient share of the hash rate to reorganize recent blocks, censor certain transactions, or execute double spending.

Canadian economist Campbell Harvey, a guest on the podcast The Wolf of All Streets, reminded that this risk has been known for a long time, but such an attack was previously considered economically irrational.

According to his own calculations, purchasing a sufficient number of mining machines to attack Bitcoin would represent about 0.5% of the total BTC market capitalization. Based on an estimated capitalization of $1.6 trillion, the cost of the operation would therefore reach around $8 billion.

Harvey adds that this multi-billion expenditure could be turned into profits if the attacker simultaneously took significant leveraged short positions in the derivative markets:

This thesis may worry more than one person, as the security of Bitcoin precisely relies on economic incentives and the cost of an attack. However, the initial estimate of $8 billion seems to overlook several important constraints that challenge its thesis.

Machines would not remain at the same price

Buying $8 billion worth of mining machines is not the same as placing an order on a liquid and deep market. The ASIC market is industrial, with limited stocks, production lead times of several months, and centralized among a few dominant manufacturers.

If a state or a large organization began to massively purchase these machines, their price would likely not remain stable. Demand would drive prices up, lead times would extend, and existing miners would have an interest in defending their position by increasing their computational power as well.

Harvey's static calculation is actually misleading, as the supply chain adds constraints that exceed the attacker's wallet size. The more the attacker buys, the more they reveal their intention, the more the market adapts, the higher the real cost increases, along with the extraction of raw materials serving other industries.

Energy, logistics, and raw materials complicate the scenario

Even assuming that the machines already exist or that enough of them are seized by the attacker, they would need to be transported, installed, cooled, maintained, and powered.

On the scale of an attack against Bitcoin, this represents several tens of millions of machines and several tens of TWh, which, as these detractors like to remind, exceeds the electricity consumption of small countries like Belgium.

Thus, a state could have a political interest in attacking Bitcoin, especially if it perceives BTC as a monetary threat. But conversely, the attacker would have a greater interest in using and protecting it, because it transforms hard-to-value energy into profit.

Shorts on Bitcoin do not guarantee a profitable attack

Finally, Campbell Harvey asserts that only taking positions in derivative markets could make this attack less costly, or even profitable.

However, to gain $8 billion with a 50% drop in BTC, approximately $16 billion in short positions would be needed. Even in the extreme scenario of BTC falling to zero, it would still require that platforms and counterparties have enough liquidity to open these positions and actually pay the generated gains.

Such a significant position would move prices, funding rates, margin requirements, and the attention of exchanges. Even if the technical attack worked, it would still require that platforms could absorb and then pay these gains in the midst of a liquidity crisis.

Thus, Campbell Harvey's thesis merits raising good questions, but claiming that it only takes $8 billion to attack Bitcoin while being profitable through the derivative market is more of a science fiction scenario than a real risk.

Indeed, although such an attack is theoretically possible, it does not seem viable as the Bitcoin network was precisely built to be more profitable to protect than to attack.