Curve's invariant made deep stablecoin liquidity economically viable, and the veCRV model turned liquidity direction into a market where protocols bid for depth. Both are genuine contributions, and both have complications.
The vote-escrow economy
Locking CRV for up to four years grants voting power over emissions and a share of trading fees. Protocols wanting deeper pools bribe those lockers, creating an entire secondary market — the Curve wars — in which Convex, Aura and others compete for influence. It is the most economically sophisticated incentive design in DeFi and it concentrates a great deal of power in whoever holds the most locked CRV.
Two incidents worth understanding
The July 2023 exploit came through a Vyper compiler bug affecting specific pool versions rather than Curve's own logic; roughly $70m was affected with much returned. The 2024 episode was different in kind: the founder's very large CRV-collateralised borrowing positions unwound across several lending markets, leaving bad debt and threatening protocols with no direct relationship to Curve. Neither was an AMM design failure; both were real risks to users.
Where it stands
Still the deepest venue for stablecoin and LST swaps, still central to how pegged assets trade, with a governance system that is genuinely decentralised and genuinely complex.
Who should care
Anyone providing stablecoin liquidity, anyone whose protocol depends on a peg holding, and anyone assessing how concentrated token positions can transmit stress across DeFi.