Outlier Ventures: Exploring 5 Improvement Proposals for the VE Token Model
"We are still in the early stages."Author: Robert Mullins, Outlier Ventures
Compiled by: Peng SUN, Foresight News
In the article "Pros and Cons of ve Token Models and Improvement Ideas," we introduced the current state of vote-escrowed token design, covering the initial goals of token design and its current flaws, including the meta-governance protocols that accumulate all underlying veTokens, which obtain rewards from veToken holders and significantly dilute their positions.
This article will present mechanisms proposed to address these issues and restore vote-escrow to fulfill its original intention of rewarding its loyal supporters, thereby promoting governance and better decentralization.
The next-generation vote-escrow model ve_v2 aims to achieve this goal by utilizing some of the following mechanisms.
RageQuit:
A core feature of the ve_v2 token design pioneered by MolochDAO is the RageQuit function. RageQuit allows locked users to cancel their lock-up early if they are dissatisfied or if their long-term evaluation of the project changes, albeit with some penalties. This feature allows disgruntled lockers to exit, who might undermine governance decisions or contribute less than before. RageQuit will also reduce the opportunity cost risk for those who might decide to commit to extending their lock-up period initially but know they will exit in the future.
Some thoughts on the RageQuit feature include:
The longer a user locks their tokens, the smaller the penalty for RageQuit. This penalizes those who want to leave early during the lock-up period; it is not a penalty for those who stick with the protocol but may now have other priorities.
When a user RageQuits, the deducted tokens will be distributed to the remaining lockers based on their "ve" share. This rewards the more loyal savers with the tokens of those who exit early, reducing dilution rates and rewarding their loyalty without incurring additional costs to the protocol.
Incorporating the RageQuit feature into the "ve" design can address some of the apparent flaws in the current system, specifically:
Lockers can exit if they wish, eliminating less loyal members and shaping a loyal community.
Loyal lockers will receive token rewards deducted from those who RageQuit, giving the most loyal individuals more governance power.
For those who wish to lock for the long term, the perceived opportunity cost risk is lower. They know that if they wish, they can exit early using RageQuit.
While implementing the RageQuit feature seems like a positive decision, there are also some drawbacks:
As prices decline, RageQuit may seem like a more attractive option for disgruntled lockers, which could increase sell-off pressure when the token price drops.
The RageQuit mechanism still does not address the issue of meta-governance protocols accumulating all voting rights and placing them in the hands of those with shorter lock-up periods.
If RageQuit becomes a common occurrence each cycle, and a few lockers receive all the deducted token rewards, the token distribution may become very centralized over its lifecycle.
Base Protocol Liquid Staking
In addition to implementing the RageQuit feature, there should also be a liquid staking option for those who are not prepared to lock long-term but want to govern and earn rewards in the short term (in a reduced manner). The main advantage and business model of the meta-governance protocol is to provide a wrapped derivative for "ve" tokens that offer liquid staking. By removing this feature, the demand for accumulating voting rights and concentrating governance power in meta-governance protocols with shorter lock-up periods is eliminated.
The base protocol can adopt a two-tier staking model for governance and rewards:
Liquid Staking: This option will reduce holders' governance power and protocol rewards while providing them with liquidity positions. The protocol can use the income generated to buy back tokens from the market and distribute them to stakers. Similar to the xSUSHI mechanism, these holders will still have governance rights over the protocol but with lower weight than locked position holders. By providing a liquidity option, stakers are less likely to deposit into the meta-governance protocol and cause voting power concentration.
"ve"_v2 Token Design: The secondary token staking will be an extended locked token design for vote-escrow that incorporates some of the suggestions outlined in this section.
Some advantages of base protocol liquid staking include:
Completely eliminating the main advantage of the meta-governance protocol, returning value to the governance tokens of the protocol.
Providing investors with an option that does not seem as permanent as multi-year locks.
Eliminating risks associated with maintaining liquidity-wrapped alternatives, as the asset can be redeemed directly from the protocol, including all accumulated rewards.
Further decentralizing the protocol, placing more governance power in the hands of "ve" lockers, while also allowing short-term supporters to have more say.
Allowing any type of investor to receive a portion of the rewards generated by the protocol's income, which in turn incentivizes the use of the product as more users benefit from its success.
While this option brings several significant advantages, there are also some drawbacks:
It does not guarantee that it will prevent the meta-governance protocol from permanently locking "ve" tokens and distributing enhanced rewards to their token holders.
It introduces another token, and if they want to allow trading of this xToken, the protocol must incentivize liquidity.
It creates a governance attack-related issue, as the liquid staking alternative has some governance power (though less than locked tokens).
It increases complexity for new market participants, as there are now two phases of staking.
Loyalty Rewards
Loyalty enhancement rewards have been utilized by protocols such as GMX and Platypus to encourage user retention and leverage game theory to engage them in the protocol, or else they will lose additional rewards for their loyalty.
While these two protocols use different variables, they share similar fundamental principles. Once stakers participate in the network by staking their tokens, they receive corresponding rewards equivalent to the staked tokens. This additional enhancement is earned as long as users maintain their stake or compound their rewards. If they unstake or begin to release their rewards, this loyalty enhancement disappears.
Some advantages of using this mechanism include:
Stakers feel obligated to continue staking or compounding to avoid losing the opportunity for loyalty enhancement.
In the case of short-term volatility or reduced protocol rewards, the opportunity cost of unstaking may be perceived as too high; if stakers still believe in the medium to long-term prospects of the protocol, they will continue to stake.
The drawbacks of this mechanism are:
- Stakers with a large number of loyalty enhancement rewards may be more inclined to hedge their positions using perpetual futures contracts (if available).
Compared to "ve" locking, loyalty rewards are a forward-looking metric indicating that past behavior may predict future behavior. While this is not a perfect assumption, we can say that the longer they are staked, the greater the opportunity cost of unstaking. Therefore, there will be greater motivation to maintain staking and reap the rewards from loyalty enhancement.
Reward Release
Another interesting model used by GMX is that governance token rewards are distributed in the form of esGMX to stakers. esGMX is a custodial token that functions exactly like a regular governance token, but it is non-transferable for a period. In the case of GMX, this release period is one year under certain conditions. If these conditions are not met, the release will stop.
For these esGMX tokens, stakers have two options:
Automatically compound these tokens (and all other rewards, including loyalty enhancements) and earn rewards as if they were regular GMX staked tokens.
Release these tokens, during which time they will not earn any rewards. To release these tokens, they must stake an average amount of tokens used to earn esTokens.
This model brings several advantages:
Theoretically, stakers are more likely to compound their rewards, continuing to earn protocol rewards rather than releasing these tokens, as these tokens do not earn rewards during the release period.
In GMX's specific case, to release esTokens, users must have staked an equivalent amount of tokens to earn esTokens. This mechanism ensures that if stakers want to unstake and leave, they will need a long time (in this case, one year) to release all their rewards, leading to liquidity (GLP) or stakers remaining in the system, or esTokens still unreleased and out of circulation.
It reduces immediate sell-off pressure. When issued, these esTokens cannot be sold, although they can be released over time, during which the protocol is expected to handle sell-off pressure more resiliently.
It rewards long-term stakers with more protocol rewards. This approach is particularly effective for protocols that generate income and wish to distribute it to committed participants.
When stakers decide to unstake and sell their tokens, their esTokens continue to earn protocol rewards, which is fair to them as they earned these rewards while contributing to the protocol. This also creates a connection between the protocol and stakers, leading to a higher likelihood of stakers re-staking to release these tokens or earn more.
Unlike the previously mentioned loyalty enhancements, esTokens can be sold on the market after the release period. If not released, they will remain with the stakers who earned them for life and continue to earn protocol rewards, which differs from loyalty rewards that are lost when stakers unstake from the protocol.
The drawbacks of this model are:
This model is not applicable to all protocols; for example, Curve cannot have a successful metering voting mechanism with a release mechanism.
The protocol's rewards are still accumulating for those who have already unstaked and left the protocol, extracting rewards from the remaining participants.
Locking Without Dilution - ve(3,3)
Andre Cronje recognized the imperfections of the standard "ve" token design and attempted to improve it with the ve(3,3) token he designed for Solidly. However, his implementation was poorly executed, coupled with a bad UI/UX, ultimately leading to the implosion of the Fantom DeFi ecosystem, resulting in nearly $7 billion in TVL loss.
Despite the poor execution, the idea is sound, allowing users to lock their supply ratio as NFTs. For example, if you purchase 1% of the token supply and lock it as a ve(3,3) position, you will receive 1% of the release amount, keeping your token supply ratio at 1%.
Some have also suggested that if all circulating SOLID is locked as veSOLID, then the protocol will not have any releases. "If all participants lock up, releases will drop to 0; if only 50% of participants lock up, there will be 50% releases, but lockers will proportionally increase their releases." This is clearly impossible, as liquidity must be provided, but the aim is to maximize the benefits for token holders while locking, thus avoiding releases.
Cronje's vision certainly has some advantages:
It allows lockers to "lock" a percentage of their locked supply, resulting in no dilution.
It allows lockers to exit their positions by selling their veNFTs on the secondary market.
The reduced variable emissions as more tokens are locked allow for initial guidance and potential price appreciation in the long term.
However, as shown by ve(3,3), it also has some drawbacks:
Extremely difficult to implement.
It does not address the issue of meta-governance protocols accumulating and locking all underlying protocol assets.
The design of ve(3,3) is very complex and extremely difficult to execute, but this idea is intriguing and should be explored further. In DeFi, combining non-fungible weights with protocol governance is a relatively undeveloped area that theoretically makes sense for long-term locking.
Improving Governance Participation Among "ve" Lockers: A Proposal
A current issue with "ve" governance is that if participants do not have economic incentives (i.e., bribed votes for evaluators), they are unlikely to participate in governance. This statement is true for veCRV holders. Among 10,203 historical holders, only 1,875 or 18.37% (as of October 19, 2022) voted on governance proposals. This lack of governance participation is not new and should not be surprising. However, for a token design aimed at attracting long-term thinkers and having substantial governance participation, these numbers do not reflect the original vision of "ve" design.
Conditional Delegation:
Similar to Optimism's recent governance structure after the airdrop, users must choose a delegate to represent them in governance decisions before claiming their OP airdrop. They can still choose to delegate to themselves, but the Optimism team knows that governance participation among token holders is typically low; thus, they require people to delegate to a designated person before applying for airdrop allocations.
However, not every delegate can make all decisions on various aspects of the protocol. It is also possible that if there is a significant decision regarding the product's fundamentals, users may want to vote themselves. Conditional delegation will take more time to authorize each vertical area of concern to the appropriate delegate. In the long run, this may lead to better governance decisions.
A conditional delegation structure like the one below can allow experts in their respective fields to make decisions suitable for them:
If the decision involves privacy or censorship, delegate to person (x).
If the decision relates to DAO partnerships, delegate to person (y).
If the decision concerns the economic structure of the protocol, delegate to the user themselves.
Such a conditional delegation structure will allow token holders to allocate their voting rights to experts in the field they deem appropriate, benefiting the overall governance of the protocol.
The ve-model is Still Evolving:
The vote-escrow model is undoubtedly a step in the right direction, especially compared to the previous purely liquid governance tokens. However, as market participants become more aware of the impact of token design and the risks of holding illiquid positions, no amount of passive protocol rewards can compensate for the massive downturn that lacks a clear recovery. If some of the ve_v2 suggestions above are implemented by protocols seeking to incentivize long-term participants in their networks, (hopefully) there will be a healthier market, better governance decisions, and more thorough decentralization. The protocol and community will develop better.
While the token design of each project is very specific, including who the token holders are, how income is generated, value flows, and regulatory structures, at the very least, RageQuit and a basic level of liquid staking options should be implemented to avoid the meta-governance protocol significantly weakening the weight of "ve" long-term lockers. This will also alleviate most of the incentive voting and income being transferred to meta-governance token holders. Additional options on a case-by-case basis are beneficial and can provide extra advantages to the protocol if they fit the design.
The "ve" token model will be a lasting token design in DeFi protocols, although not in its current form. It may differ from what is outlined above, but it will focus on greater governance participation, decentralization, and more value being captured and retained by the protocol. Rewarding those loyal token holders who participated early in the protocol and its governance should be a top priority for the protocol. The goal should be to reduce dilution, reward these participants as much as possible, and provide them with exit opportunities when they need them in the future.
But we must remember that this is a new technology attempting to implement complex monetary policies; over time, it will become clearer, as we all like to say:
"We are still in the early days."
