🏛️Governance

The past and future of token voting

Optimistic governance was pioneered by Fei Protocol among others to solve the token voter apathy problem. Legacy token governed markets tend toward either very low voter participation such that it is difficult to mobilize an adequate quorum, or toward dependence on paid delegates. The latter have a valuable role to play, but a majority-rules system still limits the throughput of decisionmaking and introduces the risk of a 51% attack.

The core premise of optimistic governance is vetocracy -- it should be easy to propose changes in the protocol, but also easy to reject them. Uncontroversial changes can be implemented without needing a majority of tokenholders to show up and vote, meaning these changes can be made with much greater frequency.

It follows that distinct types of protocol changes should have different rules. In our case, it's much easier for an honest minority to evaluate whether a collateral asset is acceptable than for them to determine whether a smart contract upgrade has been implemented safely. We would also collateral assets and interest rates to be adjusted much more frequently than new smart contracts like a custom implementation of a lending term that supports natively staking ETH.

There are two veto-able governance flows in the Credit Guild, the DAO Governor which, as the name suggests, is the highest authority over the protocol, and the Onboard Governor, which can be used exclusively to onboard lending terms. The latter is faster and has a smaller quorum than the former. The parameters below may be adjusted for the live deployment or thereafter. Both of these flows are subject to veto by either GUILD holders or lenders in the pool.

Besides the two Governors depicted above, there is also a distinct process to offboard lending terms. This process cannot be vetoed, to ensure the protocol can safely halt lending activity and wind down even if a majority of GUILD tokenholders are dishonest. This power cannot be extended to lenders, because then anyone could deposit in the protocol and offboard existing lending terms without substantial cost to themselves, which is a griefing vector.

The goal of the Electric Development Co is to spin out more granular powers for the Governor and ultimately remove its ability to make arbitrary smart contract upgrades. For example, we can create a separate Interest Rate Split Governor that only controls updating the interest split between GUILD stakers, lenders, and the surplus buffer. This presents a distinct profile of desired rules to those discussed above, as it should only be vetoable by GUILD holders but not lenders. Specifying all the possible governance actions and defining their parameters was judged to introduce unnecessary complexity prior to launch.