This program utilizes a mixture of professional and semi-professional (solos-stakers) operators and has proven track record or performance against centralised staking pools.
A custodian such as Bitgo could still be used, or a Multisig by foundation in control of the underlying funds, however I strongly recommend outsourcing the validator operation to decentralised set of node operators and not centralised exchanges such as Coinbase or Kraken.
There is a growing problem of staking concentration & net operator penetration, the top 48 operators control over 50% of staked ETH. The situation becomes even more pronounced when we factor in centralized exchanges (CEXs) like Kraken, Coinbase, and Binance, of which only the top 5 CEXs control an additional ~20% of the staked ETH. (source ‘rated.network’)
Obol can offer security and resilience for staking operations as well as competitive rates, with the added benefit of taping into a decentralised node operator set, this is much better aligned with Ethereum and Optimism.
After clarifying with the BitGo team, the expected rate of return is 3.24%, lower than the 4.5% initially cited in the proposal. This brings total expected one year net yield on the initial 20% staked to 62 ETH. Yield is highly dependent on market conditions and so actual yields may differ from estimates at different periods in time.
Aligned with the comments made by the organization I’m part of (@SEEDGov) and other contributors, I encourage the Foundation to publish the staking terms of Coinbase and Kraken with sufficient advance notice. Echoing this, the proposal should be treated as it is—a quick way to move forward and generate some yield—and an RFP process should be implemented to allow for a public assessment of the risk, reward, and terms.
I support staking the eth, but what’s the rationale for doing it through custodial pools rather than running validators / using DVT / staking in a decentralized pool? Increasing the staking power of non-custodial pools seems more aligned with Optimism & Ethereum values.
I understand the responsibility of good treasury management, but more concerned on the how these sequencer fees are staked on Ethereum. As @Gonna.eth already laid out in his post, step-by-step decisions are important because voices like mine can be heard rather than just presenting options that are pretty much exactly the same in terms of staking diversity. I think there are some things to think about when you’re allocating ETH for staking if you prefer not to do it yourselves.
As a protocol L1 dev, I can’t reiterate the importance of not centralizing stake. By going with centralized exchanges and centralized custodians, you actually inherit more risk for protocol health that isn’t seen on the surface when things are going well. Higher APR (generally from centralizing stake) comes with higher risks - the risk that the staking providers are using majority clients and are exposed to larger penalties should things like client bugs get deployed and parts of the network suffer from inactivity penalties. Luckily we’ve only had one non-finality incident on mainnet which didn’t amount to huge losses, but running majority clients Prysm and Teku at the time incurred some penalties.
I think there needs to be more conversation or more options before continuing to just hand over large amounts of stake to large entities and further centralizing it.
The Foundation has staked the initial 20% ETH approved in the proposal here. The Foundation will continue to stake incrementally generated ETH to maintain the 20% staked.
Summary of execution
Bitgo staking requires multiples of 32 ETH per validator.
20% - or 2,111 ETH - was approved for immediate staking with BitGo, which rounded down to nearest multiple of 32 = 2,080 ETH
