On Bitcoin’s Security Budget

The recent past has shown a renewed euphoria about alternative Layer-1 blockchains such as Avalanche, Solana or Elrond which receive a fair portion of the crypto community’s attention by advertising scalability capabilities and -as always- low costs per transaction.

Apart from the fact that all of the above mentioned blockchains still have to prove themselves with regard to decentralization efforts, it’s always wise to challenge their long-term security concepts, not just technically, but especially economically.

As the crypto sphere gains broader attention we observe more and more nation states tightening the noose around blockchains through regulatory measures. That is why in addition to decentralization and scalability maximising blockchain security turns out to be a particularly important task in order to prove resistance to harmful attacks by powerful players who are potentially capable of raising the necessary resources. The internet of value won’t be instantiated on long-term insecure Layer-1 blockchains. In light of new, alternative blockchains emerging, it seems reasonable to reassess the economic security model, i.e. the security budget, of the oldest, most valuable, and most exposed blockchain.

Why bitcoin has a security budget issue and why it’s not 100 years away

First of, the correct statistic for measuring the security budget of a blockchain — which could also be referred to as the disincentive to attack the network — is independent from its consensus algorithm (e.g. PoW or PoS). The security budget is measured in relative terms (i.e. miner revenue / market cap, whereby miner revenue = block reward + transaction fees). Market cap is ultimately the metric to be secured (as opposed to e.g. transactions), because the incentive to attack the network grows with the network itself. That is why nation states all have similar military/defense budgets in relation to their GDP.

(Note that hashrate can be left out of consideration as the cost per hash decrease over time.)

Graphic 1 shows that the security budget of the Bitcoin blockchain has decreased from around 10% in 2015 to around 2% in 2021 as a consequence of two halvenings.

Also, it does not matter whether the security budget is denominated in BTC or in USD terms as long as it is done consistently. This means that it is irrelevant for the security of the Bitcoin network if BTC price appreciates (or depreciates) in USD terms, because its growth in market cap is inevitably secured by a relatively decreasing miner revenue making the costs of an attack cheaper and cheaper.

Furthermore, in the case of BTC transaction fees as part of miner revenue are fairly negligible, because they make up only ~.x % of miner revenue (see graphic 2). Plus, as the higher the ratio: fees / miner revenue gets, the less predictable the miner revenue turns out to be and thus the less predictable the mining rig payback time will be. Overwhelmingly relying on fees as component of miner revenue also incentivises chain re-organisations to capture and steel fees from previous blocks — harming chain stability.

Graphic 2 shows the ratio of BTC annualized fees and its market cap in %

To summarize briefly: with every halvening (every ~4 years) BTCs security budget shrinks.

This is why it is not unlikely that a few more halvenings down the line more and more miners are likely to retreat since the budget paid by the network insufficiently incentives them to protect it. The time between blocks will likely increase correspondingly (protocol rules demand it to be ~10 minutes). The less reliable the incentive to mine, the more variable and subsequently longer block times will be. Remember also: difficulty adjusts only every 2016 blocks (~2 weeks).

Once a predetermined number of coins have entered circulation, the incentive can transition entirely to transaction fees and be completely inflation free. — Satoshi Nakamoto, 2008

Ethereum Classic serves as an example for a blockchain which is too cheap to attack and is consequently suffering from frequent chain re-organisations harming its usability. In practice, this kind of chain instability pressures exchanges to enforce two weeks of waiting time for final confirmation of transactions to further process coins — imagine these consequences for the usability of the Bitcoin blockchain…

It is worth keeping in mind that when people praise Bitcoins immutable economic policy to be its biggest strength, in fact -if no changes applied- from a security perspective it is going to be its greatest weakness.

thought construct inspired by publications from @hasufl and @drakefjustin