Uniswap v4: beforeAddLiquidity/beforeRemoveLiquidity gate the vault, afterSwap tracks exact delta, and beforeSwap returns a dynamic-fee override (Nezlobin directional fee) on a dynamic-fee pool. Delta math cross-checked against v4's own getAmount0Delta.
Reactive Network: A Reactive Smart Contract (AbstractReactive) subscribes to the hook's HedgeRequested event on Unichain and fires a cross-chain callback (AbstractCallback) to the hedger — no off-chain bot. Nonce-ordered replay protection on both legs. Verified live on Reactive Lasna → Unichain Sepolia.
LVR ("loss-versus-rebalancing") is the single biggest reason providing liquidity is harder to make money at than it looks. A liquidity provider deposits two assets into a pool and earns a small fee on every trade. The catch: when ETH's price rises, the pool automatically sells ETH to traders; when it falls, the pool buys ETH. The pool sells the thing going up and buys the thing going down — the opposite of what any investor would choose.
The people on the other side of those trades are arbitrageurs — bots that buy the cheap side and sell the expensive side. Every time the price moves, they take a small certain profit from the pool. The LP pays for it.
LVR is widely cited as the single biggest unsolved problem for Uniswap LPs — it's the reason sophisticated capital hesitates to provide liquidity, which caps how deep and competitive pools can get. Lambda attacks the LVR leak from two independent sides at once: the on-chain directional fee makes toxic flow pay the LP, while the off-chain hedge neutralizes the residual price risk.
By making a delta-neutral, yield-positive LP position possible, Lambda pulls risk-averse capital into v4 pools that would otherwise sit on the sidelines, and brings Hyperliquid funding income on-chain as a new, reusable v4 yield primitive.
The Hyperliquid perp leg (LambdaHedger + CoreWriterLib) is fully written, unit-tested, and fork-proven against real HyperEVM mainnet state — it is not on testnet for one external reason only: Reactive's testnet (Lasna) doesn't route callbacks to HyperEVM testnet; HyperEVM is a Reactive destination on mainnet (999) only.
The sequencing is deliberate: keep the free fork proof now (validates the real hedger against real mainnet state), and spend money once — on the audited mainnet deployment. A testnet detour would be paying real money for a strictly weaker artifact in between.