All research
Risk Analysisvaultsfeesauto-compoundingyield

Vault Economics: Who Actually Earns on an Auto-Compounder

A walk through the unit economics of LP vaults — manager fees, rebalance cadence, performance shares — and the conditions under which a vault beats running a position yourself.

2026-04-153 min readby Manu0x51

A vault sells you two things: discipline you wouldn't apply yourself, and gas pooled across depositors. Strip those away and most LP vaults are a worse version of holding the underlying position — because the manager has to be paid.

Useful frame: a vault is profitable for you exactly when its operational savings exceed its take rate.

The fee stack

A typical L2 LP vault charges some mix of:

Fee typeTypical rangeWho it hurts most
Management0–2% / yrLong-duration depositors
Performance5–20% of feesHigh-APR depositors
Withdrawal0–0.5%Short-duration tourists
Harvest gasPass-throughSmall depositors

The performance fee is the honest one — the manager only earns when you do — but it's also where stack-up surprises live. A 20% performance fee on a 30% APR strategy is 6 percentage points off the top, before you net IL.

The break-even depositor

For a vault to beat a self-managed position, the saved gas must exceed the manager's take. Concretely, if R is the rebalance count per year and g is the gas cost per rebalance:

const savedGas      = R * g
const managerTake   = perf * grossFees + mgmt * tvl
return savedGas > managerTake

Since g is roughly fixed per rebalance and amortized across all depositors, vaults pay off more for small depositors than for large ones — the big depositor would have been able to absorb the gas anyway. This is the opposite of what most marketing pages imply.

Where vaults outperform

Three real edges, ranked:

  1. Cross-pool routing — vaults that move liquidity between pools as fee tiers shift. Hard to do manually without an indexer.
  2. Rebalance discipline — most retail LPs procrastinate rebalancing; the vault doesn't.
  3. MEV protection — pooled txs through private mempools or batch routers, especially on Ethereum mainnet.
Read the strategy, not the APR.

A 40% APR vault that rebalances weekly into the same one-sided trend is just a leveraged directional bet sold as yield. The APR is a function of regime, not skill.

What to check before depositing

  • Last 30 rebalances on-chain — are they reactive or scheduled? Reactive wins.
  • Manager-side concentration — if one address holds >40% of vault tokens, you're an LP for the manager.
  • Withdrawal liquidity — a vault that takes hours to exit during volatility is a tax on your worst day.

The vaults worth using are the ones whose strategy you could replicate, but won't.