Case Study 2: Curve Wars — The Battle for Governance Power and What It Reveals About DeFi Politics

Background: Why Curve Governance Matters

To understand the Curve Wars, you first need to understand why Curve Finance is not just another DEX — it is the liquidity backbone of DeFi. Curve's automated market maker is specifically designed for trading between assets that are supposed to be worth approximately the same amount: stablecoin-to-stablecoin (USDC/USDT/DAI), or wrapped-to-unwrapped (stETH/ETH, WBTC/renBTC). By using a specialized bonding curve (the "stableswap invariant"), Curve offers dramatically lower slippage for these trades than general-purpose AMMs like Uniswap.

This makes Curve the default venue for the largest and most important trades in DeFi. When a protocol needs deep, efficient stablecoin liquidity — and almost every DeFi protocol eventually does — it needs a Curve pool. A stablecoin without a deep Curve pool cannot be traded efficiently, and a stablecoin that cannot be traded efficiently will lose its peg. Curve liquidity is, in a very real sense, the oxygen supply of the DeFi ecosystem.

Curve's governance token, CRV, controls a critical lever: the allocation of CRV emissions (newly minted CRV tokens) to different Curve pools. Each pool has a "gauge" that determines what share of CRV emissions its liquidity providers receive. More CRV emissions mean higher yield for LPs, which attracts more liquidity, which makes the pool more efficient, which attracts more trading volume, which generates more fees. The allocation of CRV emissions is determined by weekly gauge weight votes, and the voting power is proportional to a user's veCRV (vote-escrowed CRV) — CRV tokens that have been locked for up to four years in exchange for governance power and protocol fee sharing.

This creates a direct economic relationship between governance power and liquidity: if you control veCRV votes, you control where CRV emissions flow, which means you control where liquidity concentrates. For any protocol that depends on Curve liquidity — which includes every stablecoin issuer, liquid staking provider, and synthetic asset protocol — controlling Curve governance is not a political abstraction. It is an existential necessity.

The Rise of Convex Finance

Convex Finance launched in May 2021 with a simple but powerful proposition: it would aggregate CRV governance power on behalf of depositors, providing them with boosted Curve yields without requiring them to individually lock CRV for four years. Users could deposit CRV into Convex and receive cvxCRV, a liquid token representing their share of Convex's locked veCRV position. Convex would handle the voting, the locking, and the yield optimization.

The incentive was compelling. Individual CRV holders faced a dilemma: locking CRV for four years to earn boosted yields and governance power, or keeping CRV liquid but earning lower yields. Convex offered a third option — deposit CRV, receive a liquid token (cvxCRV), and earn boosted yields without the lockup. The tradeoff was that CRV deposited into Convex was locked permanently (Convex locks CRV for the maximum four-year period and re-locks it upon expiry), but depositors could sell their cvxCRV on secondary markets if they wanted to exit.

Convex's flywheel worked. Within six months of launch, Convex had accumulated more veCRV than any other single entity. By early 2022, Convex controlled approximately 50% of all veCRV — giving it effective control over the majority of Curve gauge weight votes. The protocol had become the single most powerful governance participant in one of DeFi's most important infrastructure protocols.

Votium and the Bribery Market

With Convex controlling half of all veCRV, the question became: who controls Convex's votes? Convex's own governance token, CVX, grants voting power over how Convex's veCRV is used in Curve gauge votes. This created a second layer of governance — to influence Curve emissions, a protocol could either acquire veCRV directly or acquire CVX to influence Convex's Curve votes.

Votium, launched in August 2021, formalized this into an explicit bribery marketplace. On Votium, protocols deposit "incentives" (bribes) for CVX holders who delegate their Curve gauge votes toward the bribing protocol's pool. The mechanism is transparent and on-chain: Protocol X wants more CRV emissions directed to its Curve pool, so it posts a bribe of $Y worth of tokens on Votium. CVX holders vote for Protocol X's gauge and claim the bribe.

The term "bribe" is used without irony in DeFi. Votium's documentation refers to "incentives," but the community universally calls them bribes, and the economic structure is unambiguous: payments in exchange for governance votes. By late 2021, protocols were spending millions of dollars per week on Votium bribes to direct Curve emissions toward their pools.

The economic logic was sound, if circular: a protocol bribes CVX holders $1 million to direct $1.5 million in CRV emissions to its pool. The $1.5 million in emissions attracts liquidity providers who otherwise would not have deposited. The deeper liquidity improves trading efficiency, supports the protocol's peg or token price, and attracts more users. The return on bribery exceeded the cost, making it rational for any protocol that depended on Curve liquidity.

The Major Combatants

Frax Finance

Frax, the issuer of the FRAX stablecoin, was among the most aggressive Curve Wars participants. FRAX's stability depended on deep Curve liquidity (so users could efficiently swap FRAX for other stablecoins), and Frax's leadership (principally founder Sam Kazemian) publicly articulated a strategy of accumulating CVX and veCRV as a core business objective. Frax accumulated CVX tokens aggressively, consistently spent large sums on Votium bribes, and created financial products (like the Frax Price Index stablecoin, FPI) that further deepened its Curve integration. By mid-2022, Frax was one of the top three CVX holders and one of the largest Votium bribe payers.

Terra (Before the Collapse)

Before its May 2022 collapse, Terraform Labs also entered the Curve Wars to support UST's peg. In January 2022, Terraform Labs funded the creation of a new Curve pool — the UST-3pool (also called the "Wormhole pool" because it used Wormhole-bridged UST) — and directed significant CRV emissions toward it through bribery. The 4pool initiative, announced in April 2022, was an ambitious plan to create a new base pool on Curve consisting of UST, FRAX, USDC, and USDT — effectively replacing DAI in Curve's core infrastructure. The plan never materialized because UST collapsed in May 2022.

The Terra Curve Wars episode is instructive because it shows how governance influence can be weaponized: Terraform Labs was spending millions to direct emissions toward UST pools, creating the appearance of organic demand and deep liquidity. When UST depegged, the Curve pool was one of the first battlegrounds — the UST-3pool became severely imbalanced as users swapped UST for real stablecoins, and the CRV emissions that had attracted liquidity providers became worthless bait as LPs fled with losses.

Stake DAO and Yearn

Stake DAO and Yearn Finance also competed in the Curve Wars, though with different strategies. Yearn focused on using its Curve positions for yield optimization across its vault products. Stake DAO positioned itself as a more decentralized alternative to Convex. Both accumulated veCRV and CVX, both participated in Votium bribery, and both offered competing products for CRV depositors.

Game Theory of the Curve Wars

The Curve Wars exhibit several classic game-theoretic dynamics:

Accumulation as defensive necessity. For a stablecoin or liquid staking protocol, not participating in the Curve Wars means ceding governance influence to competitors. If Frax accumulates CVX and directs emissions toward FRAX pools but MakerDAO does not, then FRAX pools will have deeper liquidity than DAI pools, making FRAX more efficient to trade and potentially threatening DAI's market position. The rational strategy is to participate, even if you would prefer a world where no one spent resources on governance accumulation.

The bribery equilibrium. Votium bribes converge toward an equilibrium where the cost of bribery approximately equals the value of the directed emissions. If bribing costs $1 million and directs $2 million in emissions, more protocols will bribe until the competition drives bribe costs up to near $2 million. In practice, the equilibrium is complicated by the secondary effects (deeper liquidity supporting token pegs, attracting users, etc.), but the basic dynamic holds.

Vote locking as commitment device. CRV's four-year lock creates a credible commitment: veCRV holders cannot sell their governance power in response to short-term price movements, aligning their incentives with the protocol's long-term health. Convex's permanent locking amplifies this — CRV deposited into Convex is locked forever, making Convex's governance influence persistent and growing.

Governance capture vs. governance engagement. The Curve Wars can be read as governance capture (external protocols accumulating disproportionate influence over Curve's emissions) or as governance engagement (protocols with a genuine stake in Curve's liquidity exercising legitimate governance power). The interpretation depends on whether you believe governance power should be proportional to economic stake (in which case the Curve Wars are working as designed) or distributed more broadly (in which case the concentration of veCRV in Convex and a few protocols is a governance failure).

What the Curve Wars Reveal About DeFi Governance

Governance Is Economic Warfare

The Curve Wars demonstrate that DeFi governance is not a neutral democratic process — it is an economic competition where governance power is a scarce resource with direct financial value. Protocols that control governance power can direct economic flows (CRV emissions) in their favor, creating a feedback loop where governance influence begets economic value begets more governance influence.

This is not fundamentally different from traditional corporate governance, where large shareholders use their voting power to influence board composition, executive compensation, and strategic direction. But DeFi makes the process more transparent (all votes are on-chain), more liquid (governance tokens can be traded), and more composable (governance power can be aggregated, delegated, and bribed through purpose-built protocols like Convex and Votium).

The Meta-Governance Layer

Convex introduced a concept that has since become widespread in DeFi: meta-governance. Rather than governing Curve directly, protocols govern Convex, which governs Curve. This creates a governance stack: CVX holders influence Convex's veCRV votes, which determine Curve's gauge weights, which control CRV emissions, which determine where liquidity flows, which affects every protocol that depends on Curve.

Meta-governance is efficient (it aggregates governance power and reduces the coordination cost of participating in votes) but it also concentrates power. When 50% of veCRV is controlled by a single entity (Convex), the governance of one of DeFi's most important protocols effectively reduces to the governance of Convex. If Convex's own governance were compromised — through a token accumulation attack, a smart contract exploit, or a malicious upgrade — the cascading effects on Curve and every protocol downstream would be severe.

The Plutocracy Problem

Token-weighted governance is inherently plutocratic: one token, one vote means that governance power is proportional to wealth, not to usage, expertise, or commitment (beyond the willingness to lock tokens). The Curve Wars make this plutocracy explicit: the protocols with the most resources to spend on CVX accumulation and Votium bribes have the most governance influence, regardless of whether their interests align with Curve's broader user base.

Several alternative governance models have been proposed to address this: quadratic voting (where voting power is proportional to the square root of tokens held, making it cheaper for many small holders to outvote a single large holder), reputation-based voting (where governance power depends on on-chain activity and contribution history), and conviction voting (where governance power depends on how long tokens have been committed to a specific vote). None of these alternatives has been widely adopted, in part because they introduce their own vulnerabilities (Sybil attacks for quadratic voting, gaming for reputation systems) and in part because large token holders have no incentive to change a system that favors them.

Transparency of Bribery

Perhaps the most distinctive feature of the Curve Wars is the transparency of the bribery market. In traditional politics and corporate governance, influence-buying is typically hidden, indirect, or illegal. In DeFi, it is on-chain, priced in real time, and openly discussed. Votium publishes the bribes for each gauge vote cycle; anyone can see which protocols are paying how much for which votes.

This transparency has a clarifying effect. It makes explicit what is implicit in many governance systems: that governance power is for sale and that the price of influence is determined by supply and demand. Whether this transparency makes the system more or less corrupt than opaque systems is a philosophical question, but it undeniably makes the system more legible.

The Curve Finance Exploit of July 2023

The Curve Wars narrative took a dramatic turn on July 30, 2023, when a reentrancy vulnerability in the Vyper compiler (not in Curve's own code, but in the programming language used to write certain Curve pool contracts) was exploited across several Curve pools. The total losses exceeded $73 million.

The exploit had immediate governance implications. CRV token's price dropped sharply (from approximately $0.73 to $0.50), threatening the solvency of Curve founder Michael Egorov's massive CRV-collateralized loans on Aave, Fraxlend, and other lending protocols. Egorov had borrowed approximately $100 million in stablecoins against $170 million in CRV collateral. As CRV's price fell, his positions approached liquidation thresholds.

A CRV liquidation would have been catastrophic: dumping $170 million of CRV onto the market would have crashed the price further, potentially triggering a cascade of additional liquidations and threatening the stability of Curve itself. The DeFi community — including protocols that depended on Curve — rallied to prevent this outcome. Egorov conducted a series of over-the-counter (OTC) CRV sales to large buyers (including Justin Sun, DWF Labs, and several DeFi protocols) at a discount, raising enough stablecoins to partially repay his loans and move away from liquidation thresholds.

The episode demonstrated that the Curve Wars' concentration of governance power also concentrated risk. A single individual's leveraged position on the protocol's governance token became a systemic threat because that individual's holdings represented such a large share of CRV's circulating supply.

Aftermath and Evolution

The Curve Wars have evolved since their peak intensity in 2022. Several developments have shaped the landscape:

Convex's dominance stabilized. Convex's share of veCRV plateaued around 50% and has not grown significantly, partly because alternative lockup strategies emerged and partly because the marginal benefit of additional veCRV accumulation diminished.

Bribe efficiency declined. As more protocols competed for gauge weight, the cost of bribes rose relative to the value of directed emissions, reducing the return on bribery investment. Some protocols concluded that direct CRV purchases and locking were more cost-effective than continuous Votium bribes.

The model was replicated. The Curve Wars template — governance token accumulation, meta-governance protocols, bribery marketplaces — was replicated across other DeFi protocols. Balancer launched a similar gauge system with its veBAL token, and protocols like Aura Finance (Balancer's equivalent of Convex) emerged to aggregate governance power.

Regulatory attention. The explicit bribery market attracted regulatory scrutiny. While no enforcement action has specifically targeted Votium-style bribery, regulators have signaled interest in whether governance token voting constitutes a form of securities activity, and whether organized bribery campaigns constitute market manipulation.

Risk Stack Analysis

The Curve Wars case study touches every layer of the DeFi Risk Stack:

Discussion Questions

1. Is the Curve Wars bribery market a feature or a bug of DeFi governance? Compare the transparency of Votium bribes to the opaque lobbying and influence-buying in traditional corporate and political governance. Which system is more honest? Which is more efficient? Which is more fair?

2. Convex controls approximately 50% of all veCRV. Is this a governance capture problem, or is it simply the market-efficient aggregation of governance power by parties with the strongest economic interest in Curve's success? Should Curve's governance design be changed to prevent this level of concentration?

3. The Terra/4pool episode showed how governance influence could be used to promote an ultimately unsound stablecoin. What governance mechanisms could distinguish between legitimate liquidity promotion and manipulation by unsound protocols? Is this even possible in a permissionless system?

4. Michael Egorov's concentrated CRV position created a systemic risk for the entire DeFi ecosystem. Should DeFi lending protocols impose position limits — maximum amounts that can be borrowed against a single collateral type? How would this be implemented in a permissionless system?

5. The Curve Wars model (governance accumulation + meta-governance + bribery) has been replicated across DeFi. Is this the inevitable end-state of token-weighted governance, or are there alternative governance designs that could avoid the Curve Wars dynamic? What tradeoffs would those alternatives require?