Composability · v1
Pickle Finance
pJar yield optimizer — auto-harvesting strategy vaults with DILL ve-governance.
TradFi →Fund-of-funds with auto-reinvestment
Prerequisites
01 · Concept — what problem does it solve?
Yield in DeFi Summer was fragmented across dozens of farms, and compounding it manually burned . Pickle pooled deposits into pJars — strategy vaults that harvest reward tokens, sell them, and reinvest automatically. Socialized gas costs, professionalized strategy management, passive depositors. The yield-aggregator pattern (with Yearn) that every "vault" product since descends from.
02 · Mechanics
- pJars: deposit tokens or single assets → strategy contract deploys them into Compound, Curve, Uniswap farms → harvests & compounds.
- Strategy pattern: jar (accounting) and strategy (capital deployment) are separate contracts — strategies are swappable by governance.
- DILL: vote-escrowed PICKLE (lock up to 4 years) boosts farm rewards and directs emissions — the pattern.
- Fees: performance fee on harvested yield funds the protocol and DILL lockers.
03 · Formulas
// jar share price (ratio) only grows with harvests
ratio = totalAssets / totalShares
// compounding APY from periodic harvest APR
APY = (1 + APR/n)ⁿ − 1 n = harvests/yr
// DILL boost (ve-model)
boost ∝ min(deposit, 0.4·deposit + 0.6·pool·share_DILL)
04 · Edge cases & risks
- Nov 2020 exploit — $19.7M DAI drained via a fake "evil jar" passed to a swap function that didn't validate jar addresses. Canonical lesson: COMPOSABILITY = INHERITED ATTACK SURFACE.
- Strategy risk stacking — a pJar on a Curve LP inherits Curve risk + underlying risk + Pickle contract risk, multiplicatively.
- Aftermath — merged into the Yearn ecosystem weeks after the hack; consolidation as a survival strategy.