Lido's 32 ETH validator bond matches default-route fee income only at about 747 ETH of stake

On its first day, a 0x02 key is fully collateralised by its own operator. EIP-7251 lets such a validator activate at 32 ETH and compound to 2,048 ETH ^[7]. The draft bond for a first key is also 32 ETH ^[5], so collateral starts at 100% of the stake being run ^[2]. On CryptoSlate's own formula, that key earns 0.02 of the period's gross yield per ETH bonded. A first default key on a 2.4 ETH bond earns about 0.47, roughly 23 times as much ^[11][3]. The ratio improves only as the protocol adds stake. At the 2,048 ETH ceiling, the same bond is 1.5625% of delegated stake ^[9].

CryptoSlate's 747 ETH parity point ^[11] comes from two separate gaps. On collateral alone, a default first key bonds 7.5% of its 32 ETH, and the new key reaches that ratio at about 427 ETH of stake ^[4]. The lower fee share accounts for the rest. Operators on the new route would take 2% of staking rewards ^[10] against the 3.5% in the default-route formula ^[11], and 427 multiplied by 1.75 gives 747 ^[4]. The fairer comparison is an operator who spreads the same 32 ETH over 23 default keys, bonding 31 ETH against 736 ETH of stake ^[12]. Against that portfolio, parity moves to about 1,330 ETH, roughly 65% of the module's maximum balance ^[12][5]. Every one of these figures is before infrastructure costs, gas, penalties and funding delays ^[13].

Posting the bond does not buy an allocation. The bond is a stETH security deposit, and the protocol supplies the validator's stake separately, with no guaranteed amount ^[6]. CryptoSlate argues that first-in, first-out top-ups, the proposed cap on the module and penalties that scale with balance leave net returns depending above all on how much stake a key is given and how long it runs ^[14]. The module's Staking Router parameters go to a later vote, with mainnet expected in Q4 2026 ^[4]. The source does not give a cap size or a queue depth, so how long a new key spends between 32 ETH and 747 ETH is unknown.

Lido's bond curve follows the number of keys, so a 0x02 key needs no extra collateral as its balance grows ^[8]. The operator's capital goes in at the start, and the stake that repays it arrives on the protocol's schedule ^[6][8]. Penalties scaled to balance ^[14] work against the operator here, because the full balance that fixes the fee comparison is also the balance a penalty is scaled to.

If the cap is generous and top-ups come quickly, a key filled to 2,048 ETH earns about 1.28 of yield per ETH bonded against about 0.83 for the 23-key default spread, roughly 1.5 times as much ^[6]. If the cap binds and the queue is long, keys sit near their activation balance at a small fraction of default-route fee income ^[3]. Verified Independent Community Stakers face a third case ^[15]. Lido's table gives them a 1.5 ETH first bond and a 6% reward share ^[15]. If that share applies to the first key, the new route only matches it at exactly 2,048 ETH, the module maximum ^[7].

I think queue position and the cap will decide this route's returns more than the bond will, because the bond is fixed at 32 ETH and the stake is not ^[5][6]. That view is wrong if the router vote funds new keys past 1,330 ETH soon after activation. In that case the bond is the only number left, and the 0x02 route beats a spread of default keys on fees per ETH bonded ^[12].