Why has Lido dominated on Ethereum, while liquid staking on Solana only makes up 3% of total staked SOL?

Liquid staking is incredibly strong on Ethereum, but hasn’t done well on Solana. I was there when the incumbents, Lido and Marinade, gave out millions of dollars of liquidity mining rewards weekly to secure a small amount of recurring revenue. Lido has now moved to shut down its Solana operations. And Marinade is cannibalising their own product in order to pivot to native staking. Why?

Because the value proposition of liquid staked SOL is significantly weaker than that of liquid staked ETH. For three reasons:

(almost) No lockup period. Because staked ETH is completely locked until the Shanghai upgrade (and even post-merge there is a withdrawal queue), people need a way to transact in that locked value. But staked SOL is different. You can get liquid SOL by waiting just ~2 days.

No stake minimum. There is a 32 ETH minimum stake amount. On Ethereum stake pools are an acceptable compromise for most users because it’s a necessity. Unless you run your own node and own 32 ETH, there’s no other way to stake your ETH. But Solana has a delegated proof-of-stake model, so individual stakers can stake any amount of SOL to any validator -- and they can do it in one click with any wallet.

No slashing. Validator selection is much more important in Ethereum than in Solana. In Ethereum, poorly-performing validators can be slashed, which means that you could lose your capital. There is no slashing on Solana -- the worst that can happen is you forfeit your staking rewards.

Furthermore, on Solana it's just not that hard to pick a validator. At Socean we tried Bayesian mean-variance optimisation as a delegation strategy, but found little risk-return tradeoff: the top validators stay high-performing. Unlike equities, the factors underlying validator performance are deterministic and well-understood. Therefore, delegating to a bunch of validators does not in practice give us better risk-adjusted returns than picking a single best validator.

Stake pools on Solana still have their uses. They tokenise your stake, converting non-transferrable stake accounts into fungible and composable SPL tokens. And they do monitor their delegation to validators, which might interest you if you don't know which validators to trust. But these features are being outcompeted by other primitives that perform these functions much better.

The moral of the story is that we should reassess the most obvious primitives we take for granted instead of blindly transplanting stuff from one chain to another. As my cofounder said, “California has plenty of roads, therefore you build petrol stations. You could build petrol stations in Venice, but should you?” But this is, of course, perspective gained from painful experience.