As Bitcoin’s issuance subsidy continues to fall, the network’s security will be increasingly reliant upon transaction fee revenue
Without sufficient security, Bitcoin will be susceptible to attacks from both profit-seeking and non-profit seeking actors
While there are several viable solutions to Bitcoin’s impending security problems, it remains unlikely that the traditionally conservative community will support any significant changes to the protocol
Bitcoin acolytes often cite the network’s predictable, finite, supply schedule as its most alluring feature. A marked departure from global central banks’ propensity to inflate money supply, the hard-coded 21 million cap ensures that savings held in Bitcoin will retain value over time. This property ostensibly confers ‘sound money’ characteristics, positively positioning Bitcoin as a viable contender for global reserve currency status.
And yet, despite this noble narrative, monetary theorists often overlook the equally significant technical component of Bitcoin — its blockchain. Without a secure underlying ledger, Bitcoin is susceptible to censorship and double-spend attacks. If the network itself is rendered unusable, the ‘sound property’ features ultimately become irrelevant. After all, who would possibly choose to store their wealth in an asset that has become impossible to predictably transfer?
Source of security:
Employing Proof of Work as its sybil resistance mechanism, Bitcoin’s blockchain currently derives its security properties from the vast expense required to gain control over the network, and thus, execute censorship and double-spend attacks. This expense takes the form of hardware, electricity, and operation costs and is provisioned by miners.
Miners are not altruists: they are economically rational actors, willing to devote resources in return for economic reward.
The security of a Proof of Work network is equal to the cost required to mount a 51% attack, and because miners are profit-seeking actors, over the medium-long term the cost will always be capped at the value they are receiving as payment. Taking inspiration from Jordan McKinney, we can call this value the Security Budget.
Over the medium-long term, security cannot exceed the Security Budget, as that would require miners to run loss-making operations. Likewise, under competitive market conditions, security will not underperform the Security Budget, as any margins between revenue and cost will eventually be consumed by profit-seeking operations.
Under these conditions, we can calculate a simplistic Attack Cost as equal to the Security Budget: if miner revenue is equal to $1m/day, an attacker would have to spend just over $1m/day in order to attain 51% of network hashpower.
This Attack Cost calculation assumes that all existing miners are honest and that sourcing the hardware and electricity necessary to attack Bitcoin is relatively straight forward. Likewise, this equation assumes that miners will not pause operations as more hash power comes onto the network, further driving down profitability as revenue is spread among more actors.
In reality, sourcing enough new hardware and electricity to mount an attack is logistically difficult as the majority of available hardware is currently being put to use by profit-seeking miners and cheap electricity sources are far and few between. Similarly, it is sensible to expect honest miners to stop operations as an attacker starts adding significant hash power to the network, thereby cutting into profit margins and rendering inefficient operations unprofitable. In fact, the two assumptions are somewhat related in that hardware will come onto the market as mining becomes increasingly unprofitable. As such, we can work under the assumption that the two factors negate one another.
The Security Budget is currently subsidized through a combination of new issuance and transaction fees, with the former source of income set to halve every four years. Since Bitcoin’s launch, the network has already undergone two ‘halvenings’, from a initial schedule of 50 Bitcoin per block. The next ‘halvening’ is scheduled to take place in 2020, bringing block rewards down from 12.5 BTC per block to 6.25 per block. By 2032, issuance will have fallen to 0.78125 BTC per block.
The question therefore arises as to how Bitcoin can continue to maintain a significant Security Budget under the impending condition of negligible issuance.
One variable that remains unknown is exactly what absolute value of Security Budget is necessary to ward off attackers. Is an Attack Cost of $1 million per day enough to ward off attackers? How about $10 million per day?
To answer this question, we must consider who these…