Cryptocurrency networks employ distributed consensus algorithms to ensure that pseudonymous participants reach agreement over the state of the public ledger – i.e. who owns what. In proof-of-work, electricity serves as the proving resource that imposes an irrevocable financial cost on miners. In recent years alternative algorithms have been proposed that promise similar security features and network properties at substantially less energy overhead. Most famously, proof-of-stake is a now-widely deployed alternative that reduces network power demand by more than 99%. In proof-of-stake, electricity is replaced as the proving resource with staked capital in the form of locked-up cryptocurrency. Validators – the pendant to miners – can vote proportionally to the amount of collateral they have provided.

However, it is unclear to date whether alternative consensus algorithms like proof-of-stakes can replicate the same security assurances without engaging in substantial trade-offs. For instance, proof-of-stake cryptocurrencies require social coordination outside the system when the network is partitioned (‘forked’): unlike proof-of-work, there is no objective yardstick (‘longest’ chain in terms of computational power) that enables pseudonymous actors to naturally converge towards a single chain. The introduction of subjectivity in this supposedly trustless process creates centralisation risk. It also raises ethical issues as wealth directly determines voting power. Users who can only stake smaller amounts of native tokens are unlikely to be chosen as validator nodes, leaving them with no other option than delegating their tokens to a pool. In contrast to proof-of-work mining pools, however, users cannot redirect their stake to a different provider without first being 'allowed' to withdraw, introducing the risk of vendor lock-in. Furthermore, networks become more vulnerable to potential attacks involving stolen or ‘borrowed’ private keys that control a substantial stake.

While wealth disparities and the formation of oligopolies are certainly a concern in proof-of-work as well, particularly with only a few hardware manufacturers, technological innovation can always rebalance the prevailing power dynamics. Proof-of-work may appear enormously ‘wasteful’ at first glance, but it provides a uniquely trustless mechanism for arriving at decentralised consensus in an open, pseudonymous network. Other consensus mechanisms may appear less ‘wasteful’ at first sight but have their own downsides that may not always be immediately apparent. Ultimately, a comparison between different consensus mechanisms is difficult to draw as it requires an assessment of network properties in the context of the intended utility, defined by each network's respective community and users.

It is not unheard for networks to move to a different consensus algorithm after launch. The technical and operational complexities associated with such a transition make this a challenging undertaking, however. How can a decentralised network coordinate a system-critical update in the absence of a central authority, potentially putting billions of dollars of user funds at risk? How would this transition impact network incentives, and consequently affect network security and properties? These issues are certainly among the reasons why Ethereum, the second-largest cryptocurrency and public blockchain network, has been delaying its planned transition to proof-of-stake on a regular basis – for years in a row.

Bitcoin network power demand

Theoretical lower bound

3.31

Estimated help

9.94

Theoretical upper bound

36.40

Estimated