Ethereum

🧵 Rainbow roles & incentives: ABPS + FOCILR + AS

This post explores the limits of a maximally unbundled staking set, examining potential rainbow roles and the incentive designs required to coordinate them. It presents three unbundling mechanisms:

🧵 Rainbow roles & incentives: ABPS + FOCILR + AS

This post explores the limits of a maximally unbundled staking set, examining potential rainbow roles and the incentive designs required to coordinate them. It presents three unbundling mechanisms:

https://ethresear.ch/t/rainbow-roles-incentives-abps-focilr-as/21826

Developing transparent finance for @ethereum

FOCIL with ranked transactions (FOCILR)

Rewards FOCIL includers and ensures proposers include appropriate txs from the IL aggregate when the block is full, strengthening censorship resistance. Advanced block stuffing strategies are prevented through collective includer incentives.

Attester separation (AS)

Attesters to the available chain and the finality gadget can be split into separate roles, and the post analyzes the consensus implications. It then explores the use of separate reward curves or relative reward curves to incentivize the two roles.

Two possible rainbow constellations from unbundling are shown here. The left pane features ABPS + FOCILR, with proposer and includer fees determined by a DPA (see below). The right pane incorporates AS and utilizes multiple reward curves.

The dynamic pricing auction (DPA) for ABPS works similarly to EIP-1559, adjusting the fee to maintain a target proportion of beacon proposers. The plot illustrates how quickly the fee changes under a moderate setting, depending on the deviation from the target.

If Ethereum adopts a missed proposal penalty under ABPS and the MEV falls, it may be necessary to allow for negative fees (a reward). A t-bounded DPA can then be adopted, which converges at some threshold t that differs from 0.

Multiple reward curves can also be used for incentivizing roles. The first example is to use separate reward curves. As illustrated, given that the supply of stakers willing to assume the different roles is unknown and may vary, relative proportions may diverge under equilibrium.

If the relative proportion of stakers in two roles is important, the protocol can instead use a relative reward curve (RRC). For AS, a standard reward curve can be used for finality attesters, while an RRC determines the yield for stakers who are only available-chain attesters.

A theme in the post is that micro incentives and overall yield must be unbundled. Micro rewards tied to specific duties determine the incentive that someone in a role has to fulfill its duty, whereas the expected yield determines the incentive of assuming the role. Read the post for further details, or ask here!

Attester–beacon proposer separation (ABPS)

Enforces a target proportion of beacon proposers relative to attesters. A deviation from the target causes a proposer fee adjustment, incentivizing stakers to join or leave until balance is restored. The MEV is burned at the macro level.