veTokenomics Models Compared: Curve, Frax, and Next Gen

By DeFi Intel Research · Reviewed 2026-07-02 · 11 min read · intermediate

The veTokenomics model comparison is essential for any DeFi builder or investor choosing a governance and incentive framework. Vote-escrowed tokens (veTokens) lock tokens in exchange for voting power, fee sharing, and amplified rewards, but each implementation trades off lock duration, liquidity, and governance control differently. Understanding these trade-offs separates successful protocol design from flawed incentives.

This guide dissects the pioneering Curve veCRV model, Frax Finance’s veFXS variation, and emerging next-generation designs from Paladin, Tokemak, and others. By comparing lock mechanisms, bribe efficiency, yield dynamics, and centralization risks, you’ll learn which model suits specific protocol goals and anticipate where the trend is heading.

What Is veTokenomics and Why Does It Matter?

veTokenomics, short for vote-escrowed tokenomics, is a design pattern where users lock their tokens for a chosen period (typically weeks to years) and receive a non-transferable or transferable veToken representing their locked voting power and rights to protocol rewards. Introduced by Curve Finance in 2020, the model solved two problems: reducing circulating token supply and aligning long-term stakeholders with protocol governance. Since then, it has been adapted by Frax, Convex, Balancer, and many others, each tweaking parameters to fit their ecosystem. The core mechanic is simple: longer locks grant more voting power and a larger share of protocol fees or trading rewards. This creates incentive alignment but also introduces liquidity and opportunity costs. Understanding veTokenomics is crucial because it directly affects a protocol’s token price stability, governance decentralization, and bribe markets like Votium and Hidden Hand.

The Original: Curve’s veCRV Model

Curve Finance launched veCRV in August 2020, setting the standard. Users lock CRV for up to four years to receive veCRV, which is non-transferable and non-sellable. The voting power decays linearly as the lock approaches expiry, so a holder must extend or renew the lock to maintain maximum weight. Key features include:

• Lock duration: 1 week to 4 years; longer lock gives proportionally more veCRV.
• Rewards: veCRV holders earn a portion of Curve trading fees and can boost CRV emissions on pools where they provide liquidity.
• Bribe market: Curve’s gauge voting is a target for bribes via protocols like Convex (vlCVX) and Votium. Bribes incentivize veCRV holders to vote for specific pool gauges.
• Non-transferability: veCRV cannot be traded; it’s locked and can only be withdrawn as CRV after lock expiry. This reduces circulating supply but creates a secondary market in the form of “vote selling” via delegation.

Example: A user locks 10,000 CRV for 4 years gets 10,000 veCRV (equal to max voting power). They receive 50% of Curve trading fees proportionally. However, if they want to exit early, they cannot trade veCRV; they must wait or use over-the-counter deals with counterparty risk. Curve’s model became the gold standard but also inspired competing models that prioritize liquidity.

Frax’s veFXS: A Governance-First Variant

Frax Finance introduced veFXS in 2021, closely modeled on Curve’s design. Users lock FXS (the governance token) for up to four years to receive veFXS, which — like veCRV — is non-transferable and cannot be sold or moved. The main distinction is that the system is applied to Frax’s broader stablecoin ecosystem rather than a single AMM. Key differences from Curve:

• Non-transferable veFXS: Like veCRV, veFXS cannot be traded, sold, or transferred; it exists only in the locker’s wallet and converts back to FXS when the lock expires. Locking therefore carries the same illiquidity trade-off as Curve’s model.
• Lock duration: Variable, up to four years, with longer locks granting more veFXS (up to a 4x weight at the maximum). Voting power decays linearly toward 1x as the lock approaches expiry, and users can re-lock to reset the decay.
• Rewards: veFXS holders vote on Frax ecosystem decisions and earn a share of protocol revenue, including fees from FraxLend, FraxSwap, and the FRAX stablecoin seigniorage.
• Bribe dynamics: Because veFXS is non-transferable, voting power cannot be bought outright; projects must instead incentivize existing lockers to direct gauge votes, similar to Curve. Accumulating influence requires acquiring and locking FXS over time.

“Transferable or tradeable vote-escrow designs lower the barrier to participation but introduce continuous price discovery for voting power, which can be both a feature and a risk.”

Next-Gen Models: Paladin, Tokemak, and Beyond

Emerging protocols are iterating on veTokenomics to address liquidity, composability, and decentralization. Paladin Finance, for example, created vePAL – a vote-escrowed token that can be delegated to specific NFT positions, enabling finer-grained voting control. Tokemak’s TOKE uses a “Dual Token” model where voters lock TOKE to wield governance and direct liquidity, but Tokemak does not use a classic veToken; instead, it uses a staking + delegation system. Other notable iterations include:

• Redacted Cartel’s Hidden Hand: A bribe marketplace that aggregates veToken voting power across multiple protocols (Curve, Balancer, Convex) and allows bribers to target specific gauges efficiently.
• Thena’s veTHE: A BNB Chain fork of Curve’s ve(3,3) mechanism (Solidly-inspired) with boosted rewards for locking, but with built-in “reset” of voting power every epoch to prevent dominance.
• Wombat Exchange’s veWOM: Uses a “boosted” voting model where veWOM holders get amplified trading fee revenue, similar to Curve but with additional mechanisms for cross-chain expansion.

These next-gen models often incorporate ve(3,3) mechanics (combining vote-escrow with the (3,3) game theory from Olympus), where locking rewards are amplified by overall lock rates. They also explore programmable voting (e.g., Paladin’s NFT-based votes) and liquid staking derivatives for veTokens, such as yvToken (Yearn vaults wrapping veCRV). The trend is toward modularity: separating lock mechanics from voting strategies and bribe markets.

Key Trade-Offs: Lock Duration, Liquidity, and Bribes

Lock Duration: Longer locks (e.g., Curve’s 4 years) increase commitment and reduce circulating supply, boosting token price but locking liquidity. Shorter locks allow more frequent participation but may encourage shorter-term thinking.

Liquidity: Non-transferable veTokens (veCRV) create a severe liquidity discount; users cannot exit except through opaque OTC markets. Fully transferable or NFT-based veTokens can provide some exit liquidity but also introduce price volatility and potential governance attacks (buying voting power on the open market). Next-gen models like veNFTs (e.g., Paladin) offer partial liquidity through fractionalization.

Bribe Efficiency: Bribe markets such as Votium and Hidden Hand allow protocols to incentivize gauge voting. In non-transferable systems, bribes must go to individual veToken holders, often through intermediaries (e.g., Convex pools veCRV). Tradeable veTokens can streamline bribes where they exist: a market maker could buy voting power, vote, and sell, making the bribe cost transparent. However, this transparency also means that voters may demand higher premiums to offset the risk of holding volatile veTokens.

Yield Sharing: In Curve, veCRV holders claim a share of trading fees and boost their own liquidity rewards. Frax shares protocol revenue. Next-gen models may introduce additional yield sources like protocol-owned liquidity fees or AI-curated incentives.

Comparison Table: Curve vs Frax vs Next-Gen

How Bribes and Voting Markets Evolved

The rise of bribe markets transformed veTokenomics into a political economy. Curve’s gauge voting, where veCRV holders allocate CRV emissions to specific pools, created a ‘voting-as-a-service’ industry. Convex Finance pools veCRV from users and votes on their behalf, then distributes bribes. Votium, an on-chain bribe marketplace, allows projects to pay veCRV holders to vote for their pool.

In the Frax ecosystem, bribes happen both on-chain and off-chain. Because veFXS is non-transferable, a project cannot simply buy and resell voting power; instead it must incentivize existing veFXS lockers to vote for the desired gauge, much like Curve’s gauge bribes. Influence therefore accrues to those willing to lock FXS over long periods.

Next-generation models like Paladin enable bribes to be directed at specific NFT-based veTokens, offering more granularity. For instance, a protocol can bribe only vePAL holders who have a certain minimum lock time or who voted on a specific proposal. This reduces bribe waste and increases efficiency.

“The evolution of bribe markets shows a shift from packaged voting power (Convex) to liquid, composable voting tokens that can be traded or delegated atomically.”

The Yield Dilemma: Locked Tokens vs Liquid Staking Derivatives

veTokens lock capital, creating a yield opportunity for liquid staking derivatives (LSDs). For example, Yearn’s yvCRV vault accepts CRV, locks it, and issues yvCRV that earns yield from Curve fees and Convex bribes. Similarly, Convex offers vlCVX (vote-locked CVX). These LSDs allow users to earn veToken yield without committing to illiquidity, effectively creating a synthetic version of the veToken with additional risk of smart contract and de-pegging.

Frax’s veFXS faces the same dilemma as Curve: because veFXS is non-transferable, locked FXS is illiquid, which can invite third-party wrappers or liquid staking derivatives just as with veCRV. Next-gen projects like Paladin issue veTokens as NFTs, which can be fractionalized via platforms like NFTfi or sold on secondary markets, providing a different form of liquidity.

Critically, the existence of LSDs can reduce the scarcity effect of locking – if everyone can get yield without locking directly, the token’s supply dynamics change. Protocols must decide whether to encourage LSD adoption (increasing composability) or discourage it (preserving lock incentives).

Attack Vectors and Centralization Risks

Governance centralization is the top risk in veTokenomics. Curve has been criticized because Convex holds a large portion of veCRV voting power, effectively controlling gauge emissions. Frax’s veFXS is non-transferable, so votes cannot be bought outright, but a well-capitalized entity could still accumulate and lock large amounts of FXS to sway votes. Next-gen models with NFT-based veTokens might be less prone to accumulation by whales because votes are attached to specific lock durations, but they can still be aggregated.

Liquidity attacks: In tradeable or transferable veToken systems, a flashloan or flash purchase could temporarily acquire voting power to execute a malicious vote. Non-transferable designs like veCRV and veFXS resist this because voting power comes only from time-locked tokens, and protocols can add “delayed execution” of votes for further protection.

Bribe market exploitation: Bribing markets can be gamed. For example, a briber could offer a large bribe and immediately dump the token after the vote passes. Some protocols (e.g., Thena) implement anti-whale measures or require bribers to stake their tokens.

Locked token distribution: If a large percentage of tokens are locked by a small group, the protocol becomes an oligarchy. Curve attempted to avoid this with gradual lock release, but Convex’s dominance shows it’s not enough. Frax’s veFXS, being non-transferable, concentrates influence among long-term FXS lockers, so large holders willing to lock still retain an advantage.

Future Outlook: Modular veTokenomics and ‘Vote-Escrowed Everything’

The next wave of veTokenomics likely moves toward modularity. Protocols may decouple the lock contract from the voting contract, allowing users to lock tokens in one place and delegate voting power through other mechanisms. For example, the ERC-4626 standard for tokenized vaults could be used to create veToken wrappers that are composable across DeFi.

‘Vote-escrowed everything’ is a trend where not only governance tokens but also stablecoins and LP tokens can be locked for voting power. Frax is already experimenting with staked FRAX (sFRAX) that has voting rights. The idea is to allow any asset that contributes to protocol health to have a say.

We also anticipate cross-chain veTokenomics, where voting power from one chain influences emissions on another. LayerZero and Wormhole bridges could enable “wormhole” locking – users lock tokens on Ethereum and vote on BNB Chain gauges. Projects like WePiggy and Plenty are exploring this, but it’s still nascent.

Finally, artificial intelligence may draft voting strategies based on proof-of-activity, but that’s speculative. For now, builders should focus on aligning lock incentives with real utility, not just speculation.

Choosing the Right Model for Your Protocol

When comparing veTokenomics models, consider these questions:

• What is your primary goal? Lower circulating supply? Vote Curve (veCRV). Maximize governance participation? Vote Frax (veFXS). Need composable NFT-based votes? Look at Paladin.
• What lock duration suits your token velocity? If your token is deflationary, a long lock may reduce supply and increase price. If your token is used frequently in transactions (e.g., a gas token), a short lock or fixed duration works better.
• How important is bribe efficiency? Transferable veTokens make bribes more efficient but also more susceptible to short-term vote buying. If bribes are a key part of your ecosystem, consider starting with a non-transferable model and migrating later.
• What is your community’s risk appetite? Large holders may prefer transferable veTokens to avoid lock-in. Retail may prefer non-transferable to discourage whales from dominating via buyouts.
• How will you handle upgrades? veTokenomics designs are often hard to change once launched. Consider including a governance process to adjust lock parameters or add new reward sources.

There is no one-size-fits-all. Some protocols even launch with both a veToken and a liquid staking derivative (like Convex’s cvxCRV), essentially offering two flavors of commitment. The best model is the one that best aligns incentives for your specific user base while resisting capture and manipulation.

Common mistakes to avoid

• Copying Curve's veCRV lock model without considering token supply and community size; a 4-year lock on a low-float token kills liquidity.
• Assuming transferable veTokens reduce centralization; actually, large buyers can accumulate veTokens on the open market to gain disproportionate influence.
• Ignoring the role of bribes: launching a veToken without a functioning bribe marketplace (like Votium or Hidden Hand) leads to low voter engagement.
• Choosing a lock duration that is too short (e.g., 1 week) – this eliminates the commitment benefit and turns voting into a rent-seeking mechanism.
• Failing to account for liquidity staking derivatives eroding the lock incentive: if wrapped veTokens trade near peg, the original lock is devalued.
• Underestimating the complexity of upgrading veTokenomics once deployed; many protocols get stuck with suboptimal parameters.

Frequently asked questions

Related reading

• Liquid Staking Derivatives Yield Comparison
• Liquid Staking
• Curve Finance Explained
• Liquid Staking Tokens LST
• Token Bonding Curve Design
• Bonding Curve
• Staking
• Tokenomics