Proof of stake
Replace the electricity bill with a deposit that can be confiscated. Cheaper, and it changes what the security actually rests on.
Proof of work makes block production expensive by burning electricity. Proof of stake makes it expensive by requiring a deposit that misbehaviour destroys. Both answer the same question — how do you make participation costly when identities are free — with different resources.
The mechanism
- A participant locks up a quantity of the network's own token as a bond, becoming a validator.
- The protocol selects validators to propose and attest to blocks, with selection weighted by stake.
- Honest work earns rewards, paid in the same token.
- Provable misbehaviour — signing two conflicting blocks, for instance — causes part of the deposit to be destroyed. This is called slashing.
The transition, known as the Merge, was executed without halting the network and without altering balances. It reduced Ethereum's energy consumption by an amount usually stated as more than 99% — the electricity-intensive mining step simply ceased to exist. It is the largest live change of consensus mechanism attempted on a major network.
What it improves
- Energy. No mining hardware, so consumption falls to roughly what running ordinary servers costs.
- Attack cost is recoverable to defenders. An attacker's stake can be destroyed by the protocol, whereas an attacker's mining hardware survives a failed attack intact.
- No hardware race. Participation does not depend on access to specialised chips or cheap power.
- Faster finality. Explicit finalisation is achievable in a way that probabilistic proof-of-work confirmation is not. See consensus and finality.
The criticisms worth taking seriously
- Wealth concentration. Rewards accrue in proportion to stake, so large holders' shares grow. Defenders note that mining rewards also accrue to whoever can afford the most hardware.
- Internal security. The bond is denominated in the token the system itself issues, which is circular in a way that burning external electricity is not. How much this matters is genuinely disputed.
- Barriers to solo participation. Ethereum requires 32 ETH to run a validator alone, which pushes smaller holders into pooled staking services and concentrates influence in a handful of providers.
- Complexity. The rules are substantially more intricate than proof of work's, and complexity is where bugs live.
Proof-of-work advocates argue that security anchored in an external, physical cost is more robust than security anchored in the system's own token. Proof-of-stake advocates argue that destroying an attacker's capital is a stronger deterrent than merely wasting their electricity, at a fraction of the environmental cost. Both arguments are coherent. Neither has been settled by evidence, because a serious attack on a major chain of either type has not occurred.
Staking as a product
Most holders stake through an exchange or a liquid staking provider rather than running a validator. This is convenient and reintroduces a counterparty: you are trusting the provider to operate correctly, to remain solvent, and to return your deposit. The yield advertised is a payment for that risk as well as for the protocol's reward. See exchanges.