
If you've gone through several "what is CRV" articles and still think something is missing, you’re not alone. Because many of these posts provide information on what Curve is, but not who operates it, it's history, or the current situation.
Curve Finance is a decentralized exchange, meaning code runs it rather than a company or bank. It focuses on swapping coins meant to hold the same value, like the stablecoins USDC and USDT, or pegged assets like stETH, which tracks ETH's price. Physicist and cryptographer Michael Egorov launched it in January 2020.
Most exchanges allocate their available funds evenly across all possible price levels. That's a waste when two coins are already supposed to be worth the same thing. Curve's pricing model, called StableSwap, concentrates the money tightly around the point where the two coins are equal in value, as detailed in Curve's own technical documentation. In practice, that means big trades between similar coins barely move the price. Curve also runs Cryptoswap pools, which apply a similar idea to coins that do swing in value, so it now handles more than just stablecoins.
As of late July 2026, Curve holds about $1.25 billion in total value locked and earns roughly $50.6 million a year in fees, with about $35.6 million of that counted as revenue, according to DefiLlama's live protocol tracker. That's well down from its 2022 peak above $24 billion, but the protocol has kept running through real damage: a $575,000 DNS-hijacking attack in August 2022, and a much bigger $61.7 million exploit from a bug in the Vyper programming language in July 2023 (both incidents are logged in DefiLlama's own hack records). It's still active today on Ethereum, as well as on Arbitrum, Optimism, Polygon, and a few newer chains.
CRV is Curve's governance token. I’ll get into what it actually means, because "governance token" gets used loosely all over the web3 industry.
A DAO - a decentralized autonomous organization - is a group that makes decisions by voting rather than a single executive calling the shots. Curve DAO is that group for Curve Finance, and CRV is what gives you a voice in it. The DAO runs on a modified version of Aragon, an established Ethereum governance framework, and controls every admin function of the protocol across all chains Curve runs on.
If you simply own CRV, it does not provide any meaningful actual voting power. To have all advantages that come with CRV ownership, it must be locked, turning it into veCRV. Locking leads to removal of those coins from the market, so it reduces selling pressure, and rewards long-term investors over short-term traders.
Back in August 2020, right after launch, Egorov locked a large personal stash of CRV specifically to counter a governance push from another protocol, yearn.finance, temporarily giving himself around 71% of Curve's voting power.
Unlocked, CRV mostly sits idle. Locked, it earns a share of protocol fees, a boosted cut of liquidity rewards, and a real vote in what Curve does next.
Most guides claiming to be "updated for 2026" miss out on all this:
None of this changes what CRV fundamentally does, but a genuinely current guide needs to reflect where things stood by summer 2026, not just the first quarter.
CRV does three connected jobs.
Governance: Locked CRV holders vote on which liquidity pools get weekly CRV rewards, on fee changes, and on protocol upgrades.
Liquidity rewards: If you deposit funds into a Curve pool, you earn a share of that pool's trading fees, plus CRV rewards on top. The more veCRV you hold, the bigger your reward boost, up to 2.5x the base rate.
Fee sharing: Half of all protocol fees, from every stablecoin swap to every crvUSD loan, gets paid out to veCRV holders. That's the closest thing Curve has to a dividend, and it's tied to actual usage rather than just token emissions.
veCRV, short for vote-escrowed CRV, is what you receive when you lock CRV in Curve’s contracts for anywhere from one week to four years. You cannot sell or transfer veCRV, and its voting power declines linearly toward zero as the lock approaches its end. Here’s the data:
The calculation is linear:
Initial veCRV = CRV locked × (lock time remaining ÷ 4 years)
So, locking 1,000 CRV for one year initially gives you 250 veCRV, while locking the same amount for four years gives you 1,000 veCRV. The 1-week, 6-month, 2-year, and 3-year figures follow the same formula.
The amount shown is your initial veCRV balance. It starts declining after the lock is created and reaches zero when the lock expires. Curve also rounds lock expiry times down to whole weeks, so a displayed amount can vary very slightly from the simplified examples above.
A longer lock can help you obtain a larger CRV-reward boost when you provide liquidity, but it does not guarantee a particular boost. Curve calculates the multiplier using your veCRV balance, your supplied liquidity, total gauge liquidity, and total veCRV. The maximum possible boost is 2.5x.
For example, if you lock 10,000 CRV:
Going from a one-year lock to a four-year lock gives you four times the initial voting power and fee-share weight for the same amount of CRV. But it also means your CRV remains inaccessible for four times as long: Curve’s lock contracts do not offer an early exit.
Curve Wars is the nickname for an ongoing competition among DeFi protocols to accumulate veCRV, since whoever holds it determines which pools receive Curve's weekly reward emissions. More emissions attract more liquidity, and more liquidity means better prices and lower costs for that protocol's own users. So protocols that depend on deep Curve liquidity have every reason to grab as much veCRV as they can.
Most of them don't do that by locking their own CRV for four years. They go through Convex Finance, which pools user CRV, locks it permanently on their behalf, and hands back a token called cvxCRV that carries boosted rewards without requiring the four-year commitment. Convex has become the single largest holder of veCRV as a result, which means a meaningful chunk of Curve's gauge-vote power and a meaningful share of the economic value the Curve Wars generate sit with Convex and its own CVX token holders rather than flowing straight to individual CRV lockers.
This results in certain practical implications. If you choose to directly entrap CRV, you will have full governance rights and a 50% fee share. However, if you deposit CRV tokens in Convex, you lose direct governance, but gain on the operational side, since Convex acts as a middleman between you and Curve regarding incentives. Also, thanks to bribe marketplaces such as Votium and Hidden Hand, protocols win the ability to pay veCRV and cvxCRV holders for their votes.
CrvUSD is an autonomous stablecoin created by Curve in 2023 and employs LLAMMA (Lending-Liquidating AMM Algorithm) for liquidation, which makes it different from DAI by allowing gradual liquidation of collateral as opposed to liquidation at a set ratio.
A portion of the interest CRVUSD borrowers pay flows into protocol revenue, which is part of what veCRV holders are claiming a share of. CRVUSD's growth directly affects CRV's fee-sharing yield.
The competitive moat isn't static: Curve's edge has always come from being purpose-built for a specific kind of trade. Uniswap V4 introduced a "hooks" system that lets developers build custom pricing curves on top of Uniswap's own infrastructure, including curves that can approximate Curve's low-slippage stablecoin behavior. Curve still holds a real head start in liquidity depth, though Uniswap V4's hooks are a genuine, current competitive pressure on the exact thing that made Curve different from everyone else in the first place.
You can't get your CRV back early: There's no exit button inside Curve's contracts for a lock-in in progress.
Founder concentration risk: In July 2023, a hack on Curve's own pools crashed CRV's price and put Michael Egorov's heavily leveraged personal positions at risk of liquidation. At that time, he managed to evade it by selling 106 million CRV in over-the-counter transactions, equating to $46 million in August 2023, and later on in the next month, he finished paying off his remaining loans on Aave.
He wasn't as lucky the second time. On June 13, 2024, a sharp CRV price drop actually triggered forced liquidations across five lending protocols, wiping out around $140 million of his collateral and leaving over $1 million in bad debt on Curve's own LlamaLend, which he covered by selling 30 million CRV for $6 million. This has happened more than once, involves the protocol's founder, and has directly caused bad debt in Curve's lending product.
Governance power isn't evenly spread. As covered above, a large share of veCRV sits with Convex, not individual holders, so a lock doesn't hand you the influence the marketing language around CRV tends to imply.
If you don't want to commit to a multi-year lock but still want boosted yield, depositing into Convex for cvxCRV is the common middle ground, at the cost of some direct control, as covered above.
Ultimately, the choice will come to this. If you are actively involved in providing liquidity or desire, some say in Curve governance, and you are okay with long lock-in periods, veCRV is your best bet as it will actually reward you for that. If you want some exposure with a degree of flexibility, Convex is the usual choice, with an established trade-off. If you are simply a holder hoping for short-term price growth, CRV is designed to reward those who lock their tokens, rather than just hold and wait.
Answer: Curve DAO Token is related to the usage of the protocol and governance participation rather than purely speculative reasons. It works better for active DeFi participants than for passive holders, and it carries real concentration and competition risks.
Answer: Convex owns the largest part of veCRV, meaning it has the largest control over the share of weekly gauge votes. When locking up CRV tokens,, you keep all governance rights and a full fee share; by using Convex, you trade off some of that control for advantages.
Answer: Yes, it has actually happened two times according to news reports.