Proof of stake is a consensus mechanism that picks block producers based on how much of the native token they have locked up as collateral, rather than based on who burns the most electricity solving hash puzzles. A validator in a proof-of-stake system deposits a bond (32 ETH for Ethereum, 1 SOL for Solana though actual participation requires much more in practice, various amounts on other chains), runs software that proposes and attests to blocks, and earns rewards for doing so honestly. If they try to cheat — proposing conflicting blocks, voting for obviously invalid history, or going offline for extended periods — the protocol takes some or all of their stake through a process called slashing. The economic incentive is to stay honest because the cost of being caught cheating is losing real money.
The theoretical advantage over proof of work is that the security budget — the cost of attacking the network — is decoupled from physical energy consumption. A PoS chain can, in principle, have the same security as a PoW chain at a tiny fraction of the electricity use, because the “cost” of participating honestly is the opportunity cost of having capital locked up rather than the marginal cost of electricity. Ethereum’s transition to PoS in September 2022 reduced its energy consumption by roughly 99.95 percent, and Ethereum’s security — measured by the dollar value of stake required to attack the network — has risen rather than fallen since the switch.
How Ethereum’s Version Works
Ethereum’s PoS system — Gasper, a combination of Casper FFG and LMD-GHOST — works roughly like this: validators are randomly selected every 12 seconds to propose a block, and committees of other validators attest to whether the proposed block is valid. Attestations accumulate, and after two epochs (about 13 minutes), blocks reach finality — meaning reverting them would require a substantial fraction of stake to be slashed simultaneously. This finality property is stronger than anything Bitcoin offers; once an Ethereum block is finalised, you can be essentially certain it will not be reverted, whereas Bitcoin blocks are probabilistically secure but never technically final.
The minimum stake to run a validator is 32 ETH (worth roughly $100,000 at recent prices), which is deliberately set high enough to discourage spam but low enough to be within reach of dedicated individuals. Users with less than 32 ETH can participate via staking pools like Lido, Rocket Pool, or Coinbase’s staking service, which collect smaller deposits and run validators on behalf of depositors. These pools have become massively dominant — Lido alone accounts for a significant fraction of total staked ETH — which has raised centralisation concerns that Ethereum researchers are actively working to address.
The Slashing Mechanism
Slashing is the key innovation that makes PoS economically secure. When a validator signs two conflicting messages (proposes two blocks at the same height, or votes for incompatible checkpoints), their behaviour is provably dishonest, and any honest party can submit the proof to the network. The protocol automatically deducts a portion of the validator’s stake — typically around 1 ETH for minor offences, but scaling up sharply if many validators are slashed in a short window as part of a coordinated attack. The stake is burned, not paid out to anyone, and the validator is forcibly exited from the active set.
This creates a direct economic penalty for attacking the chain. On Bitcoin, a 51 percent attack is limited by the cost of acquiring mining hardware and electricity — expensive, but not directly punished; an attacker can mine honestly afterward and recoup costs. On Ethereum, a 51 percent attack requires putting up enormous stake and then getting that stake slashed when the attack is detected, which means the cost of the attack is also a permanent loss of capital. This “skin in the game” property is what makes PoS theoretically secure even without physical resource consumption.
The Criticisms
Proof of stake has critics — especially among Bitcoiners — who argue it is not as trust-minimised as PoW. The main points: The rich get richer. Validators earn staking rewards in proportion to their stake, which means initial distribution inequality tends to persist or worsen. Bitcoin’s PoW, by contrast, channels rewards to whoever runs the most efficient miners, which at least requires ongoing operational effort. The starting state matters more. If PoS chains can be attacked by acquiring majority stake, and majority stake is determined by who held the chain at genesis, the chain’s security is partly a function of its initial distribution — a political fact, not a cryptographic one. Weak subjectivity. A PoS chain requires nodes to agree on a relatively recent “checkpoint” as a starting reference, because long-range attacks (where an old private key is used to forge an alternate history) become possible in theory if you have the old keys. In practice, this means new nodes have to trust an out-of-band source to tell them which history is legitimate, which is a meaningful trust assumption PoW does not require.
These are real concerns, and the jury is still out on whether they are serious enough to matter in practice. So far, Ethereum’s PoS has been stable, has not produced any successful attacks, and has performed well through multiple market cycles. But “three years of stability” is not a permanent verdict, and the argument between PoW and PoS camps will likely continue as long as both chains exist.