Against set-and-forget

The promise of the first generation of automated market makers was that liquidity could be passive. Deposit, walk away, collect. Concentrated liquidity broke that promise and, in exchange, offered something better: far more fee income per dollar, for anyone paying attention.

The trouble is the exchange rate. Attention is not free, and it is not something most depositors can supply. Watching a position across a weekend, deciding whether a 3% drift warrants a move, pricing gas against expected fees at two in the morning — this is a job. Done badly, it is worse than not doing it: a panicked reposition into a trend can cost more than a month of fees.

The three ways this usually goes

• 01The wide band. Set once, never touched, always technically in range, earning a fraction of what the capital could. Safe, and quietly expensive.
• 02The tight band, abandoned. Set on a good day, left behind by the first real move, out of range for weeks while the depositor believes they are earning.
• 03The over-managed band. Moved on every wobble, paying the full cost of a rebalance for a marginal improvement, several times a week.

All three are failures of attention rather than of intelligence. The first never spends any, the second spent it once, the third spends it without a budget. What is needed is not more cleverness but a rule applied consistently, cheaply, and without sentiment — which is a thing software is genuinely good at.

So: if you can pay attention, provide liquidity yourself. You will keep the whole fee. If you cannot, pay something that can, and read its ledger to check that it is.