Written in October 2022
- PoW: Proof-of-Work, Bitcoin consensus algorithm
- Miners: participants spending electricity for the right to write blockchain history and earn newly minted BTC
- PoS: Proof-of-Stake (validators), Ethereum consensus algorithm
- Validators: participants publishing the blockchain history in PoS. They do this by staking (locking) ETH, and earn newly issued ETH + transaction tips.
- Consensus algorithm: in a network, the mechanism allowing all participants to agree on the state of the system
Open blockchains require a robust consensus algorithm to be decentralized. It is what allows all participants to agree on the state of the network (what happened, and who owns what) without relying on a centralized company, bank, or government.
The strength of its consensus algorithm is what makes a blockchain resistant to attacks. Ethereum recently changed its consensus mechanism, making it very different from that of Bitcoin. Let’s see what it’s all about!
The need for a “cost”
In open blockchains, specific types participants called miners or validators, write the history of the networks’ activities so that everyone can coordinate themselves. All the other participants need to ensure that these publishers remain honest and don’t cheat the network.
To do this, publishing history requires a cost that is “wasted” if the miner or validator try to cheat. In the case of Bitcoin’s PoW, this cost takes the shape of energy. In Ethereum’s PoS, this is staked ETH.
If these actors behave honestly, they receive incentives in the form of newly issued BTC/ETH and transaction fees paid by users. If they cheat, they miss out on their rewards and waste the resource they dedicated for their participation.
To make sure a decentralized network stays secure, this resource needs to be 1) costly, and 2) hard to monopolize. If it isn’t, it becomes very easy to attack the network and stop it from working. This is why energy appears like an unmatched choice to secure a global and valuable network of this type.
PoS removes this energy requirement and introduces new safeguard mechanisms to disincentivize attacks on the network. In the following part, I look at Ethereum’s Proof-of-Stake mechanism. Then I compare it with Bitcoin’s Proof-of-Work under some aspects.
Ethereum moved from Proof-of-Work to Proof-of-Stake on September 16th 2022. Instead of consuming energy, Ethereum 2.0 participants stake 32 ETH in order to become a validator and have the right to randomly publish a block at the tip of the blockchain.
Each validator need to attest the validity of every block, and 2/3 of validators attestations are required to consider a block valid. If a validator doesn’t attest a block, they have a penalty the size of the reward they would have earned for this attestation.
Validators are also regularly chosen to propose a block. When this is the case, they get additional staking rewards (newly minted ETH) plus the tips of the block’s transactions. If they try to cheat and propose an invalid block, they get “slashed” and lose some of their stake instantly and a little bit every day. They are also banned from publishing blocks. After 18 days, a “correlation penalty” takes place. The more validators have been punished recently, the more of their staked ETH each of them will lose (up to their total balance). This mechanism was implemented to discourage organized attacks.
To summarize, in Proof-of-Stake validators:
- stake ETH — their cost, which has to be put up once
- earn ETH from staking rewards (inflation) + transactions tips — their reward, currently at 5–7% return per year
- and lose their staked ETH if they try to cheat — their punishment
An aspect to keep in mind is that any “staking” entity can run multiple validator nodes. This leads to some stakers controlling a very large share of all validators: they publish more blocks than the others. A malicious actor gaining a very large share of validators could damage the network. The mechanisms detailed above are the network’s protection against this.
Contrary to Bitcoin miners, Ethereum validators don’t have any recurring cost. As long as they remain honest, they just need to stake ETH once to validate blocks. Miners spend energy continuously.
As a consequence, Proof-of-Stake (PoS) validators need much less reward to compensate for their cost. This allowed Ethereum to drastically reduce its ETH inflation when moving from PoW to PoS.
Changes in ETH issuance
Under PoW, ~13,000 ETH were being minted to miners every day as block rewards to pay for their electricity cost. They were also earning 100% of the transaction fees paid by users.
Since the transition to PoS, there is no recurring electricity cost to be paid anymore. As staking is much less costly, validators need much less revenue to make this profitable. This allowed Ethereum to decrease block rewards down to ~1,600 ETH per day, or a ~90% decrease!
Additionally, validators don’t receive all of users transaction fees anymore*: they only receive a “tip” specified by users and the rest is burned, effectively reducing the total ETH supply.
💡 * this was not implemented during the merge, but in an earlier improvement called EIP-1559.
This is a major change for Ethereum, as it decreases the yearly inflation rate from ~4.62% to ~0.49% at most, and potentially even less depending on network activity.
*these rows are subjective, and only based on my (limited) knowledge of how PoW and PoS work.
Bitcoin’s Proof-of-Work and Ethereum’s Proof-of-Stake are two consensus algorithm based on very different mechanisms. Despite these differences, they allow both networks to actually work in a coordinated and decentralised way.
There are a lot of debates going around to know whether each of them will have what it takes to resist to attacks through time. For now, there doesn’t seem to be a major threat on any of the two that can’t be handled.