Nearly a third of concentrated DEX liquidity sat idle in H1
DeFi & Yields ·
Dune research commissioned by 1inch finds concentrated liquidity pools on major DEXs still leave significant capital earning nothing, despite design upgrades meant to fix that problem.
The study, commissioned by 1inch, examined top pools on Uniswap, PancakeSwap, and Aerodrome and found that 29.5% of capital sat out of range through H1, meaning it was parked outside the active trading band and earned no fees. In an average week, that amounted to roughly $542M in idle capital. Annualized, the forgone fees for liquidity providers came to approximately $150M.
The mechanics behind concentrated liquidity are meant to let LPs concentrate capital in a specific price range to earn more fees per dollar deposited than under the old model. But the research suggests that benefit only materializes when positions stay within range as prices move. When prices drift outside the chosen band, that capital stops earning anything until the LP manually adjusts the range or the price moves back.
The idle capital problem also appears sticky. Over a third of the idle money has not been touched in 90 or more days, and the research attributes this largely to individual wallets that set a range once and did not return to manage it. That points to a passive-user behavior pattern rather than active market-making, with retail-style depositors bearing the brunt of the forgone yield.
The research frames this as still an improvement over the prior liquidity model, where 99% of v2-style liquidity went unused entirely. But it singles out Uniswap v4 for scrutiny, finding its newer design performs no better than v3 on the idle-capital metric, despite architectural changes intended to improve efficiency.
What remains unresolved is the exact reconciliation between figures circulating in the wider research cluster, some citing 85% underutilization, $1.6B to $1.84B in total idle liquidity, and annual fee losses ranging from $150M to $195M, depending on the report and dataset scope. It is not yet clear whether these are different snapshots of the same dataset, expanded chain coverage, or methodological differences. Whether 1inch or other protocols will propose specific fixes for range-management automation, and whether Uniswap addresses the v4 finding, are open questions.