Design Staking Platform

Problem Statement: Liquid Staking at Protocol Scale

Liquid staking solves a fundamental tension in proof-of-stake: to earn staking yield on Ethereum you must lock 32 ETH into a validator and that ETH becomes illiquid, but DeFi wants the capital working. A liquid staking protocol pools deposits, runs validators through professional node operators, and hands the depositor a liquid receipt token (stETH, rETH, cbETH) that keeps earning yield while remaining tradeable and composable across DeFi. Lido (stETH/wstETH) dominates; Rocket Pool (rETH), Coinbase (cbETH), and Frax (sfrxETH) are the named alternatives, differing on operator permissioning and the receipt-token model.

The core mechanics. Users deposit ETH → the pool mints a receipt token → the protocol batches deposits into 32-ETH validators run by node operators → beacon-chain rewards accrue and are skimmed for a protocol fee before the rest flows to holders. The receipt token tracks accrued value one of two ways: rebasing (stETH — your balance grows daily) or share-based / value-accruing (wstETH, rETH — your balance is fixed but each token redeems for more ETH over time).

The two defining risks. Slashing — a misbehaving validator is penalized on the beacon chain, and that loss must be socialized fairly across all holders (correlated slashing across many validators is the tail risk that drains the insurance fund). Peg/liquidity — the receipt token can trade below ETH on secondary markets when redemptions aren't instant; the canonical event is the stETH/ETH depeg of June 2022, when the 3AC/Celsius unwind forced stETH sellers into thin liquidity (pre-withdrawals) and it traded to ~0.94 ETH.

Scale to anchor on: ~$25B TVL, ~4,000 validators, ~12k deposits/hour at peak, with target peg deviation < 0.5% and exit-ETA error < 6h p95. The design's whole job is to keep the receipt token fully backed and liquid while yield accrues — composability is the product, but backing and peg are the survival constraints.