Introduction
DeFi protocols go through many stages of growth and iteration with the ultimate goal being an in-demand product that can scale and generate revenue. Gearbox is entering the next stage of growth with the upcoming V3 launch. Given the magnitude of the changes, V3 is also a chance to look for ways to improve the tokenomics to further align the governance token with the success of the protocol. The purpose of this pre-proposal is to begin to spec out a use case for GEAR in the V3 Quota system, as well as re-establish some additional long term token ideas that have been considered in previous discussions.
Motivation
The V3 Teaser article introduced the potential for the GEAR token to govern quotas for long-tail risk assets. The purpose of this pre-proposal is to attempt to flesh out the details of how this model could be implemented, as well as to synthesize the status of previous tokenomics discussions from past forum posts and discord messages.
Intro to the Quota System
Very Quick Summary
To understand the proposed tokenomics, one has to first understand the changes V3 introduces. This is a very short summary of the system described in the teaser article above.
V3 enables adding assets with less liquidity as collateral, by introducing quotas and the total limits on those quotas. An assetās quota is a parameter set by the user for their own account, and determines how much of the asset is counted towards the health of that account. The total limit on quotas determines the maximal sum of quotas on all accounts for a particular asset - if the limit is reached, users canāt increase their quotas until one of the existing Credit Accounts reduces theirs.
The borrower pays interest on the size of their quota, both to adequately compensate LPs for risk and meter quotas, as they are a limited resource.
Quota System - Who Decides?
With more exotic assets, there will be a wide range of strategies that all borrow from the same pool - ie, should a newer degen token with 20% APR pay more quota interest than a more established yield strategy with 10% APR? Because the quota is set for each strategy, the balance of strategies is not a security term. What it does offer is a way for the protocol to charge users higher rates for using lending pools for high risk/reward farms, while avoiding rendering established lower APR farms as useless.
In this regard, the interest rate on quotas is subjective - which tokens should have high rates relative to others? What is a fair quota rate for borrowers to pay lenders? Who decides? That will be the GEAR holders! That is where a possible use case for GEAR comes into play.
Possible Updated Tokenomics
V3 Quota System
In the quota system outlined in the V3 teaser, it is noted that GEAR governance and committees will be required to set minimum and maximum viable interest rates for each supported quota asset. From there, each week GEAR holders can distribute their votes among assets to shift the interest rate lower (better for Credit accounts/lenders) or higher (better for lenders). If a GEAR holder does not have a position they may simply wish to shift the rates in favor of the user type that is most in need, to maximize users and/or protocol revenue.
The following is an outline of how this could function:
- GEAR tokens are staked in the GearStaking contract. They are locked for 4 weeks. The contract is a general use voting contract, of which the first use case would be the quota system.
- GEAR stakers can at any point vote for the borrower side or the lender side for any asset with enabled quotas (note in step 4 that only the vote status at the end of the epoch matters). Each asset with a quota will have its own gauge that ties into the GearStaking voting interface.
- GEAR stakers may vote for any asset/side, receiving one vote per GEAR token held (ie, they can use all their voting power on a single side for a single asset, or spread it across many assets. One can even vote for both sides in the same asset, to reduce rate volatility without actually changing the rate).
- At the end of each voting epoch (which is week-long), the rates are recalculated based on votes at that moment. All existing votes are rolled over into the next epoch, and the new voting epoch begins.
GearStaking - more info
The GearStaking contract will likely be the main voting instrument across all voting initiatives in the Gearbox ecosystem (which will start with voting for quota interest rates, but will later be expanded to on-chain governance, and possibly other things).
The GEAR stakers would be able to use their staked GEAR for voting immediately on deposit, in all connected voting contracts. The staking contract keeps track of how much GEAR from the stakerās staked balance is already used in some voting initiative - so that users canāt ādouble-voteā. The staker can withdraw their staked GEAR with a 4-week delay, provided that the withdrawn amount isnāt used for voting.
In practice, this means that stakers would be able to reallocate their staked GEAR among voting initiatives freely and gas-efficiently, with any voting-based changes applied at epoch end. To exit the staking module and get access to their GEAR, the user would have to remove their votes and wait for 4 weeks, which prevents āflash-votingā.
Other Tokenomics Ideas
These are longer term ideas. Unless the DAO makes a significant push to work out the details, these would not be intended to be implemented with V3 (GearStaking would be first priority).
1. Revenue/GEAR rewards for GEAR staking and/or locking, token buybacks, token burns
Giving GEAR to GEAR stakers does nothing for the protocol. There is also very little revenue and a lot of growth potential at the moment, meaning sharing revenue (as well as token buy backs and/or burns) also doesnāt make sense. There is also regulatory uncertainty around sharing revenue. Lastly, decentralization requires effort from GEAR token holders, so if revenue were to be shared it should be done in a way that incentivizes users to put effort into taking actions that decentralize the protocol and create more active participants. Ie, it may or may not be justified to consider providing marginal GEAR rewards for those that stake in GearStaking and vote in the Quota system gauges to incentivize GEAR holders that arenāt CA or LP users to participate - but given the
2. veNomics
VE systems often involve locking a token for a period of time, with a multiplier on the relative voting power of the position based on the lock time. Typically this voting power is used for gauges - weekly or bi-weekly votes to direct protocol emissions to incentivize various aspects of the protocol. If your protocol needs liquidity mining, it is a somewhat elegant way to decentralize the decision making around how to allocate the liquidity mining. However, it is also a complex system and many users try to force veNomics onto protocols that do not need users to participate in these types of decisions. DAO contributors are split on the general idea of ve systems, but one certainty is that a general proposal to ādo veGEARā will not work. veGEAR should only exist if specific goals of the DAO lead to a solution that is best solved by veGEAR, NOT by assuming veGEAR needs to exist or that GEAR needs more utility and then trying to force a system to exist.
A basic idea for veNomics is to do gauges for lending pools. However, this assumes liquidity mining will always be necessary for lending pools, which may not be the case. If it is needed, a ve system is a somewhat proven way to decentralize the decision making around distribution.
Another idea is to do gauges on credit accounts. No one wants to pay lenders except the DAO itself, and borrowers. However, if a DAO gets an aspect of their protocol incorporated into Gearbox, then that aspect of their protocol should in theory be more efficient. This creates an arb where the protocol can bribe more TVL into their pool, saving money + getting more users. However, this could create oddities where strategies yield negative base interest and only go positive because of GEAR rewards, thus rendering the borrowing rates high and pushing other strategies into negative interest rates. Additionally, bribe multipliers mean protocols are emitting more token than the value of the bribes they recieve - so it may be better to simply offer the option for protocols to bribe their own tokens to integrated strategies, rather than build an entire gauge system for it. Again - liquidity mining / GEAR distribution is likely not necessary on the CA side, so it would be odd to force a veSystem on it.
One aspect that veSystems solve is ensuring that revenue is distributed in return for a service being provided by token holders. One approach to doing this is to use protocol revenue from each gauge strategy to bribe votes for that strategy. This creates a baseline of incentivization protocol profitability that other bribes have to overcome if they wish for the protocol to be used inefficiently.
3. Liquid wrappers
There are many limited-locked governance tokens like vlCVX and sdCRV for veCRV, and vlAura and sdBAL for Balancer. There are liquid lockers for these tokens, including many by Redacted (such as pirexCVX). In theory quotas could make these assets feasible to integrate to allow leveraged bribes. However, they do not have oracles. One potential workaround is to allow people to lock GEAR as cross-collateral for their locked positions. Liquidators receive the locked tokens and some quantity of GEAR based on the time to unlock. Or, give the liquidator the GEAR and then the user gets their locked tokens back instead of the GEAR. The user could re-buy GEAR when their locked tokens unlock. This solution would require deeper GEAR liquidity as well as an oracle for GEAR - but that would be the only oracle required in order to integrate any number of locked bribe governance tokens.
4. Aave style safety module
A safety module would entail GEAR or GEAR/WETH liquidity being staked in return for protocol revenue. If there were to be an liquidation event or an exploit causing bad debt, the staked GEAR would be slashed to cover the debt. Note that AAVE is currently discussing changes and upgrades to their staking module due to various shortcomings, including inefficiency of the TVL staked and the fact that if there were bad debt or an exploit, the cost of AAVE may suffer, which reduces the effectiveness of the module.
5. Staking for Enhanced Features
The goal of the token is to govern and enhance the protocol. The goal of the protocol is not to enhance the token. In general, limiting certain features of the protocol to token holders increases the value of the token at the expense of the protocol - which long term, likely means it is also at the expense of the token. Any ideas along this line of thinking should be case-by-case, ideally looking at a feature that would need to be limited to only a small group of users anyway. Quotas is NOT a good example of this, because anyone can use quotas, and the more users there are the more demand/revenue there will be for the lending. A better example would be if perhaps a bot needed to pay keepers for it to function. Certain bots could be offered to GEAR holders where keeper fees are covered, while non-GEAR holders could use version of the bots that require the user to pay keeper fees as part of their transaction costs.
