Rainbow staking

1

Barnabé Monnot

Robust Incentives Group (RIG), Ethereum Foundation

ZuBerlin Protocol Day — 12/06/2024

h/t S. Reid

Validator services

2

Network of validators

Consensus service

Block construction service

Censorship�resistance service

We want a decentralised set of protocol participants because they express a wider set of preferences!

• When making a block, include txs that others dislike!
• When participating in consensus, decorrelation => resilience!

Consensus services

4

Validator services (consensus)

5

Consensus service

Finalise chain�(Casper FFG)

Propose blocks (Consensus data)

Figure out chain head�(LMD-GHOST)

The world according to eth

6

Why stake to provide FFG?

Credible commitment to good service provision, the protocol can make statements such as� “If X happens, Y billions of dollars are lost”.

Commitment requires capital, capital wants to be free,� enter Liquid Staking Protocols…

Note that the statement above doesn’t specify whose dollars…� The attacker’s? Protocol doesn’t see delegations!

7

The world out there

8

🙈

Operators

9

Protocol service providers

Two classes of providers:

• Solo operators: A priori untrusted, think living room validators, solo stakers (operators + own capital)
• Professional operators: A priori trusted, think�registered companies, big staking providers

Not a binary distinction!� More like credible signal, to be learned over time.

(Liquid) Staking Protocols may employ a mixture of both types.

10

Take Rocket Pool, operator can be solo staker,� put up some ETH as collateral, delegators fill the remainder.

Still, capital efficiency + cost pressure mean� LSPs generally rely on professional operators.

11

Towards Rainbow staking

12

• Premise 1: We can make only the people who actually perform validation (the operators) liable.
• Premise 2: People want to stake (“do something with ETH”).

Strawman:�“Trustless Rocket Pool”

13

• Issue: People want yield, so they’ll want to be slashable�Will recreate an operator—delegator separation (heavy).

You could still have light delegators getting yield for choosing heavy delegators well, or doing something useful… but what?

14

15

V. Buterin

If light stakers do stuff, then they are operators!� And if they are operators… can they take delegated “stake”?

🥱 People are lazy…

16

Idea: Let light delegators delegate to a distinct set of operators.

Light delegators “back” light operators, give them weight.

None of the light stack is slashable, no pressure to rebuild the LSP stack with centralising forces => Solo operator-friendly!

17

A great light service: Inclusion lists

Don’t need economic security!�Just need good Sybil-resistance/leader-selection.

18

h/t T. Thiery

Economic organisation revisited

Validator services

20

Network of validators

Consensus service

Block construction service

Censorship�resistance service

Post-APS world

21

Network of�beacon proposers

Consensus service

Block construction service

Censorship�resistance service

Network of execution proposers

🥴

A nice economic organisation

22

Network of validators

Consensus service

Block construction service

Censorship�resistance service

Network of proposers

Heavy!

Light!

23

A way out for issuance?

Grotesquely simplified version of the argument�High issuance => Lots of people stake� => Real yields fall� => Lots of stake under LSPs => LSTs are money� => You heard the pitch before :)

Posts on revising the current reward curve (h/t Caspar+Ansgar):

• Electra proposal: Moderate rewards
• Case for targeting: Tune issuance, reach range of staked ETH

24

Solo operator sustainability?

Rainbow staking to increase solo operator sustainability?

25

Thank you!

https://rig.ethereum.org