BuidlerDAO: A Comprehensive Analysis of the Mechanisms and Innovations of the veToken Economic Model

Article: @Buidler DAO

Authors: @Jane @Niels

Table of Contents:

01/Introduction to ve

01/01 How ve works

01/02 Curve ------ Pioneer of the ve model

02/ Innovations of ve and different protocols from the perspective of core mechanisms

02/01 Balance of liquidity

02/02 Setting of incentive mechanisms

02/03 Trade-offs in governance mechanisms

02/04 Attempts at separation/layering of rights and interests

03/ Summary In the early DeFi projects, liquidity mining rewards for early users were a common means to facilitate cold starts. However, while this approach contributed some initial liquidity, it could not support the healthy development of a protocol in the long term. Users' more intuitive reactions still leaned towards speculation and cashing out, leading to an inevitable rapid decline of the protocol, aptly depicted by the death spiral of defidigital. Furthermore, when a protocol's token has only governance functions and the governance rights are quite limited in practical scenarios, users will also lack sufficient motivation to hold on.

If we summarize these two issues, it boils down to how tokens can provide sufficient holding motivation, and how the interests of users and the protocol can remain aligned over a relatively long period, thus motivating contributions to the long-term development of the protocol.

The emergence of ve (vote escrow) is a quite good answer to the aforementioned dilemma. After Curve first proposed ve, many other protocols have incorporated it into their economic models and made their own iterations and innovations based on Curve's ve.

This article will first explore the operational model of ve, using Curve as an example to analyze the advantages and disadvantages of ve, and then we will look at the iterative development of several core dimensions of ve in conjunction with the innovations of different protocols at various levels. Finally, we will provide some suggestions for the practical application of ve based on the above analysis. Introduction to ve How ve works The core mechanism of ve is that users obtain veToken by locking tokens. veToken is a non-transferable and non-circulating governance token; the longer the locking period chosen (usually with a maximum locking time), the more veToken can be obtained. Based on the weight of their veToken, users can acquire a corresponding proportion of voting rights. The voting rights are reflected in the ability to determine the allocation of liquidity pool rewards for newly minted tokens, which can have a substantial impact on users' personal gains and enhance their motivation to hold tokens.

For the protocol, locking can effectively reduce circulation, alleviate selling pressure, and thus stabilize the token price. At the same time, once users lock their tokens, their personal interests are more aligned with the protocol during the locking period, which also helps them make better governance choices. Those willing to lock for the long term can earn more rewards and governance rights, which is also very fair from a moral standpoint. Next, we will further understand the real application of ve through Curve's model. Curve ------ Pioneer of the ve model Users on the Curve platform primarily earn two types of rewards: trading fee rewards and CRV minting rewards.

Trading fees are the source of revenue for the Curve platform. The platform charges this fee during swaps (0.04%, charged in the token of the output party), deposits, and withdrawals (0-0.02%). Of this, 50% of the trading fees are given to veCRV holders (used to purchase 3CRV, which is then distributed to holders, redeemable once a week), and 50% goes to LP providers (i.e., Base APY).

CRV is the native token of Curve, used to reward liquidity providers who stake LP tokens (the tokens received by users who contribute liquidity to the pool) into the Gauge. The Gauge controller is a core component of Curve, controlled by Curve DAO. The Gauge controller records all parameters related to voting governance, such as "the current slope for each user in each Gauge, the powers available to users, and the end time for user voting locks."

The initial supply of CRV is approximately 1.273 billion, with a total supply of around 3 billion, and the initial daily minting amount is about 2 million. These newly minted CRV tokens will be distributed proportionally to users who stake LP tokens in the Gauge based on the measurements recorded in the Gauge controller, i.e., users with weight in the Gauge. https://github.com/curvefi/curve-dao-contracts/blob/master/doc/README.md When staking users further lock their CRV, they can obtain veCRV (veCRV is non-transferable, and locking is the only way to acquire it). Users can choose a locking duration ranging from 1 week to 4 years. The longer the locking period, the higher the exchange rate for veCRV. Locking 1 CRV for 4 years yields 1 veCRV, while locking for 1 year yields 0.25 veCRV. During the locking period, the number of veCRV held by users will linearly decrease as the remaining locking time decreases. If users lock in multiple transactions, they must choose a unified expiration time.

Holders of veCRV have governance rights over Curve. To encourage users to participate in governance (i.e., to have the motivation to acquire more veCRV), they enjoy certain rights: The intention behind limiting the reward multiplier for locked users is that while the ve mechanism aims to encourage users to lock, rewarding loyalty and stickiness, it needs to avoid allowing veCRV whales to endlessly reap more rewards. For example, to achieve a 2.5x reward by depositing 100 USDC stablecoin, one would need 2840 veCRV, equivalent to staking 2840 CRV for 4 years, assuming CRV = 1 USDC, which means holding LP tokens worth 28 times the value of CRV.

Ordinary users find it difficult to obtain such a multiple of CRV in a short time, hence their reward multipliers are often limited. The reason for settling on 2.5x is likely because the team believes this boundary achieves a relatively balanced incentive. Additionally, if a user provides liquidity to multiple pools, their reward multipliers may vary based on the locking conditions of the LPs in different pools.

The following diagram summarizes the different rights and incentives of holding CRV and veCRV within the Curve system: https://resources.curve.fi/crv-token/understanding-crv Interpretation:

1) Liq in pool ------> earns lending & trading fees: Providing liquidity (deposit), earning fees

2) Has veCRV ------> vote on DAO proposals & vote on gauge weights & earn governance fees: Holding veCRV: having governance rights, voting to determine gauge reward weights

3) Liq in gauge ------> earn CRV: Staking LP tokens in the gauge earns CRV

4) Liq in gauge & has veCRV ------> boost: Staking LP tokens in the gauge while holding veCRV can receive boosted rewards

From the current operational results, this mechanism has been quite successful. 45% of CRV is locked for voting, with an average locking duration of 3.56 years. https://dao.curve.fi/releaseschedule If we compare the distribution of locking durations, the long-term locking effect of ve becomes clearer. https://members.delphidigital.io/reports/an-alternative-implementation-of-vetoken-economics Advantages As mentioned earlier, here is a summary of the advantages of ve:

1) Reduced liquidity after locking decreases selling pressure, helping stabilize the token price.

2) Better governance possibilities: Governance rights are directly tied to reward distribution, enhancing users' motivation to hold tokens and increasing governance participation. At the same time, users who stake for the long term have greater governance weight, and they are also motivated to make better governance decisions. The time and quantity-based weighting reflected in governance weight currently seems quite reasonable.

3) Long-term interests of all parties are relatively aligned: During the locking period, users cannot transfer their tokens, and due to their vested interests, rationally, they are more likely to ignore short-term gains and make decisions more aligned with the long-term interests of the protocol.

The locking mechanism also increases the short-term manipulation costs for whales. Once they choose to lock to increase their influence, they are likely to make rational votes that align with their own interests, significantly reducing the likelihood of malicious decisions.

Additionally, veCRV holders also enjoy a share of the transaction fees, meaning the interests of liquidity providers, traders, token holders, and the protocol are coordinated together. (The interests of traders are reflected in benefiting from the pool's strong liquidity and lower slippage.)

If the cold start period is successfully navigated, Curve forms an excellent positive feedback loop, with VE playing an indispensable role. It can be said that ve is a typical mechanism-level constraint on human nature, guiding positive behavior. Disadvantages No model is perfect, and ve has its points of criticism.

1) Rigid "locking time" is not friendly to investors:

The locking duration is both a sweet spot and a point that deters many investors. Some jokingly say that 4 years in the crypto industry is equivalent to 100 years in other industries. A significant number of investors do not want or cannot lock for such a long time. If it wants to continue to develop, how to increase its appeal to a broader range of investors and enhance the flexibility of locking is a topic.

2) Centralization of governance:

Currently, over half of the governance rights on Curve are held by Convex (53.65%), leading to a concentration of governance power.

Note: Convex is a liquidity staking and mining platform based on Curve, where CRV holders can stake CRV on Convex to receive cvxCRV. The Convex platform automatically locks the obtained CRV tokens on Curve to acquire veCRV tokens controlled by the protocol, which can be understood as cvxCRV tokens being tokenized and tradable veCRV. In other words, Convex achieves maximum reward incentives by betting on Curve through certain rule designs and risk controls, leveraging "economies of scale." https://dao.curve.fi/locks/1667977964 https://dune.com/queries/56575/112408 This is related to the locking time mentioned in 1), as non-whales find it difficult to have sufficient motivation and ability to continue locking. It is also related to Curve's whitelist mechanism. This mechanism stipulates that smart contracts cannot participate in the DAO unless they receive voting permission (51% approval rate, 30% participation rate). Its origin is to maintain the stability of the protocol. Historically, only three protocols have obtained whitelist permissions: Yearn, Staking DAO, and Convex.

Currently, voting rights are concentrated in a few protocols, which intuitively have no motivation to agree to let more protocols participate, leading to fierce competition. If new innovative gameplay emerges, the Curve war may be pushed to a new stage.

However, the whitelist is somewhat contrary to the fairness and openness that crypto has always advocated. If the whitelist played its role in the early stages, whether an open ecosystem without permission would be more beneficial for Curve, or even Convex, in the long run is debatable. This proposal (Remove Curve DAO Whitelist) is discussing whether the whitelist should be removed, sparking widespread discussion. Users who insist on retaining the whitelist believe that it is still more beneficial than harmful, and if new protocols believe their innovation is sufficient, they should not worry about being blocked from participation due to voting. Looking at ve from the perspective of core mechanisms

And the innovations of different protocols The DeFi world is changing rapidly, and economic models are continually iterating. Curve is by no means the pinnacle of the ve mechanism; different protocols have made various improvements based on Curve's ve. The following will categorize these innovations from different levels of core mechanisms, analyze the problems they attempt to solve, and see what points we can weigh and optimize when designing the ve model. Balance of liquidity As mentioned earlier, the long locking period in Curve ve poses obstacles for many investors. The most straightforward approach may be to shorten the maximum locking period, such as Balancer's maximum locking period set to 1 year. In subsequent models, based on Curve's own veCRV, some protocols directly provided liquidity solutions. Additionally, compared to Curve's hard locking, many relatively soft locking methods have begun to emerge, attempting to strike a balance between long-term locking and liquidity. Tokenization of ve tokens Convex's answer to the low liquidity of ve is to tokenize ve tokens. Users who store CRV on the Convex platform can obtain cvxCRV (which only supports one-way conversion within Convex). After staking cvxCRV, in addition to receiving 3CRV trading fee rewards (like veCRV holders), they will also receive additional CRV (10% of Convex's CRV mining rewards) and CVX rewards (the token of the Convex platform). Users who lock CVX (i.e., vlCVX holders, vlCVX = The Vote Locked CVX) can gain governance rights over Convex, voting to determine the weights of various gauges.

As mentioned in section 1.2.2, Convex controls over half of the governance rights on Curve, so obtaining governance rights from Convex (vlCVX holders) effectively allows for strong indirect control over Curve. The difference is that the locking period on Convex "only" requires 16 weeks + 7 days, significantly increasing flexibility compared to Curve's 4 years.

Moreover, as the meaning of the word "Convex" suggests (the slope of a convex function increases), Convex effectively aggregates the power of "retail" investors to achieve higher reward incentives. The increased annualized yield further enhances Convex's appeal. These two factors also partly explain why Convex quickly gained acclaim after its launch (in May 2021).

However, Convex somewhat reduces the long-term consistency between veCRV holders and the protocol, which may be an unavoidable consequence of increased liquidity. Based on Convex, platforms like Votium have also emerged, further increasing governance hierarchy and complexity. (This article refers to Convex as a functional L2 of Curve, and Votium as a Convex L2, providing an insightful analysis of various Curve L2 protocols.)

Additionally, attention must be paid to the decoupling risk between cvxCRV and CRV. Currently, cvxCRV/CRV is around 0.83, having deviated from 1 for a considerable time. To address this, Convex is also launching new countermeasures, primarily aimed at providing more rewards to cvxCRV holders. https://www.defiwars.xyz/wars/curve NFTization of ve tokens From the perspective of locking, veCRV has non-transferable ownership. In the VE(3,3) model proposed by AC (the founder of Yearn), one significant mechanism innovation is to make ve tokens transferable. He designed the locked ve tokens as NFTs, allowing an account to form multiple veNFTs during multiple locks, with each NFT cumulatively contributing to the total amount in the account. After NFTization, ve tokens can not only be traded in the secondary market but can also further develop a lending market, significantly enhancing liquidity. Since ve is linked to governance, liquidity providers and governance users (veNFT holders) may be separated.

The question raised here is how to price veNFTs, and if locked ve tokens can be traded directly (even at discounted prices), what motivation do users have to lock ve? This is a point that needs further refinement.

Note: The VE(3,3) model combines Curve's ve model with OlympusDAO's (3,3) game theory model. (3,3) refers to the game results of investors under different behavioral choices. The simplest Olympus model involves two investors who can choose to stake, bond, or sell. As shown in the table below, when both investors choose to stake, the joint yield is maximized, reaching (3,3), aimed at encouraging cooperation and staking. https://olympusdao.medium.com/the-game-theory-of-olympus-e4c5f19a77df Soft locking * exit penalties Compared to Curve's mechanism where users cannot exit midway after locking, many protocols have extended to allow mid-exit mechanisms, albeit with added exit penalties. Penalties are usually related to the locking reward coefficient, meaning that while users can exit midway, they will also lose their locking rewards, increasing the opportunity cost of exiting. Thus, in certain scenarios (such as short-term price fluctuations), users may choose to hold firm.

Taking Platypus (a new type of stablecoin AMM) as an example, staking PTP (Platypus's native token) can yield vePTP and additional PTP minting rewards. One staked PTP can generate 0.014 vePTP/hour, with a generation cap of 180 times the deposited PTP (approximately 18 months). Users can withdraw their stake at any time, but if they withdraw any amount, all vePTP accumulated over time will be completely cleared.

This effectively gives users the choice to either continue holding for the results they have already accumulated or to easily choose to give up. It can be imagined that, due to the added judgments and loss aversion psychology, users' choices will be more nuanced and diverse.

In contrast to Platypus's complete zeroing rule, Yearn (a DeFi yield aggregator) is much gentler. In its upcoming ve model (based on the YIP65 passed at the end of 2021), the locking period for veYFI ranges from one week to four years. If holders choose to exit midway, they will face a variable penalty related to their remaining locking time. If the remaining locking period is greater than 3 years, the penalty is 75% of the locked YFI; otherwise, it is calculated based on the ratio of the remaining locking period to the maximum locking period (4 years). This penalty fee will be rewarded to the remaining holders of veYFI.

GMX (a DEX) also adopts a proportional cancellation of rewards.

In addition to penalty rewards, some protocols have chosen to extend the withdrawal time, such as Prism (a derivatives protocol), which stipulates a 21-day withdrawal period.

Soft locking achieves a balance between liquidity and long-term holding incentives to some extent. Redistributing penalty fees to locked holders further encourages long-term holding behavior. However, specific parameter settings (such as maximum locking time, exit penalties, etc.) should be determined based on the priorities desired in specific scenarios. Setting of incentive mechanisms In this section, we will explore some details of the incentive distribution mechanism.

The rewards for holding ve tokens can generally be divided into:

1) Token minting rewards

2) Enhanced reward multipliers

3) Revenue sharing (such as transaction fees)

4) Penalty fees and other earnings distributed proportionally (if any) Business revenue sharing incentives First, let's look at the sharing of business revenue, i.e., whether ve holders enjoy a share and the specific sharing ratio. Theoretically, if users are optimistic about the long-term development of the protocol, and the rewards for ve holders can be continuously tied to long-term income, it should increase users' motivation to hold long-term. Taking Curve as an example, veCRV holders enjoy 50% of the trading fees from all pools, distributed in the form of 3CRV. This means that when the price of CRV drops, investors are motivated to buy CRV to purchase a larger share of the revenue distribution at a cheaper price.

On this basis, the ve(3,3) model further refines this, meaning that users only enjoy the transaction fees generated by the pools they vote for. This new restriction makes users inclined to vote for the pools with the strongest liquidity that generate the most transaction fees.

As for how to balance the earnings of veToken holders and liquidity providers, the specific sharing ratio depends on the protocol designers' considerations. Curve adopts a 50-50 split, while in the GMX protocol, the revenue sharing ratio between stakers and liquidity providers is 3-7.

Additionally, if business revenue sharing is involved, tokens will carry a certain degree of stock-like nature. In U.S. law, the Howey Test is used to determine whether something qualifies as a stock; if classified as a stock, it will be subject to additional regulatory constraints. However, since it is case law, specific determinations must be made based on individual cases. Different countries also have different methods of determination. Token minting incentives In issuing token rewards, Platypus (a new type of stablecoin AMM) has implemented a three-pool design: https://medium.com/platypus-finance/platypus-liquidity-mining-design-eli5-part-i-52fd6b8bed1d The minting revenue shares of the three pools are 20%, 30%, and 50% respectively (this weight can be adjusted later).

The AVAX-PTP two-pool (20%) is the liquidity pool for PTP;

The base pool (30%) rewards liquidity providers, with reward weight proportional to their storage ratio;

The boosting pool (50%) rewards ve holders, with reward weight depending on the amount stored and the number of vePTP. Anti-inflation Since protocol tokens are typically continuously minted, if no measures are taken, locked tokens will also face inflationary pressure. If foreseeable inflationary pressure is too great and users cannot sell midway, they will clearly be unwilling to lock. To address this, the ve(3,3) model proposed its optimization plan, specifically reflected in:

1) Weekly minting amounts will be dynamically adjusted based on circulation

Assuming the original weekly minting amount is 2M, if currently 0% of tokens are locked as ve, the weekly minting amount will still be 2M, i.e., 2M(1-0%); if 50% of tokens are locked as ve, the weekly minting amount will be 1M, i.e., 2M(1-50%); if currently 100% of tokens are locked, the weekly minting amount will be 0, i.e., 2M*(1-100%). In other words, the more tokens locked, the less the minting amount, reducing the impact of minting on locked users by dynamically adjusting the minting amount.

2) ve lockers will receive proportional compensation

Assuming the current total supply is 20M and the ve locked amount is 10M, when the weekly minting amount for that week is 1M, it means a 5% increase in supply, i.e., 1M/20M. To ensure that ve holders are not diluted, under the (3,3) model, their holdings will also increase by 5%, i.e., 0.5M=10M*5%. The remaining 0.5M of the weekly minting amount will be released as rewards.

Looking at 1) and 2) together, designers aim to enhance the locking motivation by protecting ve users from dilution.

Inflation acts like a tax, increasing users' holding costs. Therefore, if token minting is involved, we need to pay attention to the current circulation, maximum issuance, and minting speed when deciding to lock. Protocol designers can refer to ve(3,3) for more detailed regulations on minting and compensation.

However, on another level, the balance of token incentives between early and late entrants must also be considered. If minting rewards overly favor existing ve holders, it may lack appeal for potential new investors. The new DeFi Thena.fi has made further improvements based on ve(3,3), limiting the increase ratio to 30%, providing only partial dilution protection for ve holders, thus preventing tokens from being overly concentrated in the hands of early ve holders. The main consideration here is how token issuance can continuously provide motivation for the long-term development of the protocol. Sustainability of Real Yield Using token rewards to promote cold starts and holding has become a common means for project parties, but after the aggressive campaigns of 2021 where DeFi protocols released large amounts of tokens to attract liquidity, investors have begun to calmly examine the sustainability of APY. The introduction of Real Yield reflects this pursuit of stable returns.

The calculation method for Real Yield: Net income = Protocol income - Market value of minted tokens

The intention of this formula is that we should consider the tokens issued to users as a real cost and deduct them from the protocol's income to calculate net income, using net income to assess whether the protocol has the potential to achieve positive returns.

Defiman conducted a study on Real Yield, revealing that most leading protocols implementing token incentives fail to achieve positive real yields, let alone lesser-known or early-stage protocols that often adopt more aggressive token incentive policies, raising doubts about their long-term sustainability.

In the ve model, to provide users with sufficient long-term holding motivation, while the protocol income rewards are limited, designers often choose to primarily rely on token rewards. Therefore, when setting incentives, it is worth considering how to reasonably set the incentive level to provide users with sufficient motivation while ensuring the protocol's sustainability, meaning that business income must be enough to cover this portion of reward expenditure, and the business itself must remain competitive (with reasonable but not excessively high fee income). However, balancing short-term and long-term is indeed challenging and tests the designers' ability to maintain rhythm. Trade-offs in governance mechanisms Fluid governance rights One of the core purposes of the ve model is to achieve good and stable governance, with the key being who holds governance rights and how the governance weights are distributed among different individuals. Ideally, the main governance rights should be allocated to investors whose long-term interests align with the protocol. Additionally, from a temporal dynamic perspective, it is best to achieve a certain fluidity of governance rights.

If investors holding a large amount of veCRV control the protocol through voting rights and thus further hinder the entry of new investors (refer to Curve's whitelist mechanism), it will be difficult for the entire protocol to change. Once governance rights are completely solidified, to some extent, the concept of governance rights ceases to exist (as small voters cannot have a substantial impact on the outcome).

This involves several points (not considering other specific restrictions and requirements):

1) No whitelist restrictions, allowing interaction with the protocol without permission (to prevent monopolies and promote openness)

2) ve designed to be non-transferable to avoid direct bribery

3) The voting weight of ve decreases over time, allowing governance rights to flow naturally

Taking vcDORA (a veToken for open funding community governance) as an example, the governance capacity of a certain amount of ve over a period can be intuitively viewed in the diagram below (the shape of the curve is an example, and there is no need to focus on why it has this shape); the area under the curve from the current time to the end time represents its total governance capacity. When the staking time expires, the user's governance rights become zero. The curve g(t) indicates that g(t) is integrable (i.e., the area of the purple part in the diagram can be calculated), and it does not necessarily have to be a linear decay line. This nonlinear governance power curve is also an exploration direction. https://doraresear.ch/2022/09/16/vcdora/ Loyalty measurement

If we view weight as a quantifiable way to measure ve users' loyalty to the protocol (weight determines voting/reward distribution, etc.), we can see that there are currently two measurement methods: one is based on the remaining locking time (looking forward, veToken linearly decreases over the remaining locking period), similar to Curve, and the other looks at the historical staking duration of users (looking back, the longer the staking time, the more veToken accumulated, but with a cap), as seen in Platypus. These two measurement methods represent a balance between new and old users to some extent. If the second backward-looking method dominates, investors with longer staking times will find it very difficult to be surpassed. However, whether users who have staked longer are more likely to continue stable staking is subjective. Nevertheless, since the second method often combines mid-exit penalty mechanisms (see 2.1.3), investors will face certain opportunity costs, which may incentivize them to continue stable staking.

In contrast, the forward-looking method (primarily based on remaining locking time) led by Curve may result in a more fluid distribution of weights. Any new user who chooses to lock for a long time will receive sufficient ve, while old users who do not choose to continue locking once their remaining locking time expires will no longer hold ve. Therefore, if users want to continue enjoying good reward multipliers, they need to update their locking time to keep their remaining locking time at a longer level. This also partly explains why Curve's average locking duration can reach 3.5 years. Pros and cons of bribery Bribery is a controversial topic. We tend to view bribery as a neutral term, depending on the main demands of users in specific scenarios. After all, in real life, it has a similar term: lobbying.

In some scenarios, if users primarily care about yield, a convenient bribery mechanism can enhance user experience, and it may even be considered to directly integrate bribery reward mechanisms into the protocol to increase the attractiveness of ve rewards and boost participation motivation. For small ve holders, if there is no convenient way to participate in bribery protocols, they may find it difficult to commercialize their governance rights or face high communication costs. For whales, if they conduct transactions privately through OTC, they may also face the risk of breach of contract. An embedded bribery mechanism resolves these trust issues.

In fact, in Velodrome Finance (currently the DEX with the highest locked volume on Optimism), bribery income contributes significantly to the main ve rewards (as shown in the diagram below). Moreover, once the bribery mechanism is embedded, it will inevitably impact second-layer derivative protocols like Convex. https://twitter.com/VelodromeFi/status/1616489024268402711 In other scenarios, if there is a strong emphasis on the authenticity of voting and a desire for users to participate as much as possible while minimizing bribery occurrences, methods such as ZK-based MACI voting (hiding each person's vote but showing the final voting results) can be adopted. In this case, voters' privacy can be well protected, and bribers cannot know the true voting situation, thus lacking the motivation to bribe.

Thus, protocol design is merely a tool, and these methods can be combined based on specific scenarios. Attempts at separation/layering of rights and interests Based on ve, many protocols are also exploring new ways to distribute token rights, i.e., further breaking down veToken, where different tokens have different rights, allowing for more refined rights management.

As mentioned earlier, Convex essentially separates the rights to rewards and governance from veCRV, providing more flexibility: staking cvxCRV can earn 3CRV rewards, while locking CVX to obtain vlCVX can influence the veCRV controlled by Convex, thus voting on governance decisions for Curve pools. Separation of governance rights and reward enhancement rights Ref Finance (a DEX in the Near ecosystem) proposed a new scheme in its version 2.0 token design. Instead of obtaining a single veToken, liquidity providers can receive two tokens, veLPT and Love(Ref), after locking (1:1).

Among them, veLPT is non-transferable and corresponds to the voting governance rights of the ve token, allowing holders to vote on the allocation of incentive distribution. Love(Ref) corresponds to the liquidity provided by the user in the ve token (the locked liquidity share), and holders can enjoy enhanced earnings (decided by veLPT voting). Additionally, Love(Ref) is transferable, meaning that if a user only wishes to enjoy voting rights, they can transfer the earnings enhancement rights to others. However, when unlocking veLPT, the account must still hold a 1:1 ratio of veLPT and Love(Ref). https://ref-finance.medium.com/ref-tokenomics-2-0-vetokenomics-on-testnet-c2b6ea0e4f96 It can be seen that after token rights are subdivided, different investors can choose the parts they desire most in the protocol and focus their investments. The flexibility of all parties is enhanced. Layering of rights and interests Astroport (a DEX in the Terra ecosystem) strikes a balance between Curve's non-locking governance rights and long locking high yields. It innovatively adopts a three-token mechanism: ASTRO, xASTRO (earned by staking ASTRO), and vxASTRO (earned by locking xASTRO, linearly decaying over time). xASTRO is transferable and enjoys certain governance rights and transaction sharing (50%). vxASTRO, in addition to the 50% transaction sharing, enjoys more governance rights and yield enhancement (up to 2.5 times).

The advantage of this model is that it accommodates the demands of both short-term and long-term holders. By allowing non-locking users (xASTRO holders) to also enjoy certain governance rights, the governance process becomes more democratic, and participation friction is reduced. At the same time, by asymmetrically granting vxASTRO holders excess rewards, it fully encourages long-term holding behavior. In this way, small but committed investors can amplify their influence through long holding, creating a subtle power dynamic with large but short-term investors.

Through the layered structure of revenue and specific distribution ratios, protocols can effectively adjust the investor structure they wish to achieve and potentially attract a broader range of investors. If Curve is 100% ve (only locked users enjoy governance rights and yield enhancement), then through rights layering, we can set an incomplete ve mechanism (giving non-lockers certain governance rights and yields, such as 50% ve), with the specific switch position depending on the protocol designer's considerations. Summary The ve model aligns the interests of all parties well through locking and revenue sharing, motivating participants to contribute to the long-term healthy development of the protocol. From the perspective of economic models, ve represents a significant advancement compared to previous models and has achieved great success.

Based on ve, different protocols have continuously innovated, contributing iterative solutions from various dimensions. In this process, we can see that each protocol makes repeated adjustments at key levels of the mechanism according to its own needs and focuses. As system designers, understanding the potential adjustment space and designing based on one's own situation, the brilliance lies in these trade-offs. The permissionless and composable characteristics further encourage continuous innovation. Core settings dimensions of the ve model Moreover, in protocol design, embedded mechanisms are the best way to interact with participants. We cannot assume that participants are all rational beings who can adhere to the (3,3) principle or are willing to align with the interests of the protocol. In practice, we need to encourage participants to make reasonable decisions through mechanisms such as locking reward enhancements, exit penalties, and opportunity costs. By separating/layering rights, participants can choose the investment solutions that suit them best. A good mechanism can guide the orderly development of the protocol, provide long-term holding motivation and reasonable incentives, which is also the core theme of ve's continuous iteration.

Through ve, we may also extend towards better governance mechanisms. In the table above, we can obtain the final voting synergy results by assigning different weights to different factors. One measure of a successful governance mechanism design should be whether it can promote more value creation and lead to better benefit distribution.

However, in the process of continuously refining the protocol, it is also essential to exercise restraint. AC reflected on the design of Solidly in his article, mentioning that "Simplicity works decentralized, complexity doesn't." Indeed, sometimes complexity does not equate to effectiveness. Furthermore, complexity must be adapted to the scenario. Whether ve is necessary, whether ve tokens should be tradable, and whether non-ve holders can also enjoy certain governance rights—there are no standard answers to these questions.

Of course, despite the many innovations, the effectiveness of the ve model ultimately relies on the market and time for validation. We can design the granularity of the ve model as needed and continuously adjust the design based on feedback. We look forward to more exciting ve model designs.