Dune research finds 85% of concentrated DeFi liquidity is underutilized, with $150M in annual fees foregone

Quick Take

• About 85% of concentrated liquidity across protocols Dune analyzed was underutilized at any given moment, with roughly $542 million sitting fully outside its fee-earning range in an average week.
• Out-of-range liquidity providers give up an estimated $150 million a year in fees, and more than a third of that idle capital has gone untouched for over 90 days, Dune said.

About 85% of concentrated liquidity across decentralized exchanges is not doing the work it was deposited to do, according to research by onchain analytics platform Dune commissioned by 1inch.

The study rebuilt every liquidity position in roughly the top 200 pools by activity on Uniswap v3, Uniswap v4, PancakeSwap v3, and Aerodrome Slipstream, taking 26 weekly snapshots across seven chains between Jan. 6 and June 30. It covered about $1.84 billion in liquidity on average per week, approximately $1.6 billion of which was underutilized at any given time.

Concentrated liquidity, introduced with Uniswap v3, lets a liquidity provider set the price band their capital covers rather than spreading it across every possible price. It was designed to offer an improvement on the constant-product venues it replaced.

As a comparison, Dune found roughly 98.7% of liquidity on constant-product AMMs sits outside the day's traded band.

Out of range is only the floor

The strictest measure is out of range, since the market price sits outside a position's band, so the capital earns no fees and adds no depth. Averaged over the 26 weeks, 29.5% of tracked capital was in that state, or roughly $542 million in an average week.

According to Dune, the figure barely moved. Apart from an early-February spike that touched 41.4%, the out-of-range share held inside a 25-35% band for the entire period.

Being in range is not the same as efficiently working, the report argued. A position counts as in range the moment the price falls anywhere inside its band, however wide that band is drawn.

Rebuilding the broader calculation from 1inch's Aqua whitepaper, Dune found v3-family capital splits into 13.7% actively used, 56.9% in range but never touched by the week's trades, and 29.4% out of range.

"Due to structural inefficiencies in DeFi, liquidity providers are leaving billions of dollars in underutilized capital and millions of dollars in fees on the table," said Sergej Kunz, 1inch co-founder. Kunz added that 1inch is preparing to launch Aqua, a shared liquidity product aimed at the same problem.

Share of DeFi capital out of range | Image: Dune.

A third of the idle money has not moved in 90 days

Out-of-range capital is not automatically abandoned.

Some liquidity providers deliberately park ranges to one side, where they function as standing limit orders that convert when the price crosses them. Others run bots that constantly rotate positions.

Considering this dynamic, Dune checked when each idle position was last touched. About 43.8% had seen a deposit or withdrawal within 30 days. Another 19.5% was 30 to 90 days old. The remaining 36.7%, roughly $200 million, had gone more than 90 days without a single adjustment.

The venues also age their idle capital very differently.

On Uniswap v3, 44.5% of out-of-range capital is dormant by the 90-day test, the largest single block in the study. Aerodrome sits at the other end, at about 20% dormant, with 58% of its idle touched within the past month.

Distance tells the same story as time. Only about 35% of idle capital sits within 5% of the market price. Roughly 43% is more than 25% away, and 17% is more than 100% away, where the market would have to double or halve for those positions to earn again.

Composition of idle capital by protocol | Image: Dune.

The idle rate falls with size; the idle dollars do not

Smaller positions go idle more often, per Dune's report. The rate ranges from 53% for positions under $1,000 to 26% for those above $1 million, and is monotonic across all intermediate buckets.

Yet, the dollars run the other way. Positions over $1 million hold about 47% of all idle capital, roughly $260 million, and those above $100,000 hold about 76% of it. Sub-$1,000 positions dominate the count but account for 1.6% of idle dollars.

Individuals, not bots, hold the idle capital

Dune traced each position's liquidity NFT to whoever holds it now, attributing staked capital to its farm or gauge contract.

On Ethereum, wallets own 91% of Uniswap v3 capital and 94 cents of every idle dollar. On Arbitrum, 78% of capital and 92% of the idle. Base makes the divide starkest: contracts hold about half of the Uniswap v3 capital there, but only 6.5% of it sits out of range, compared with roughly 30% for wallets, so individuals carry 82% of the idle.

Rewarding in-range liquidity holds the number down without erasing it. Aerodrome's staked capital is the least idle measured anywhere at about 16% out of range, while PancakeSwap's staked capital runs markedly higher, particularly on BNB Chain, where sharp moves push pools out of range faster than LPs restore them.

Dune flags a caveat on reading the ownership split: a wallet-held position is individually held but not necessarily retail, since professional desks trade from ordinary wallets too.

Uniswap v4 changed the plumbing, not the idle

Uniswap v4's top 200 pools held about $230 million at the end of June, with 30.5% out of range on average --- effectively identical to v3.

The architecture kept v3's tick-based ranges and changed what sits underneath. Pools share a single contract rather than each deploying their own, which cuts costs without touching the idle question.

Hooks, the consequential addition, let a pool attach custom code around swaps and liquidity changes, and in principle could push parked tokens into a lending market to earn while out of range.

None of the hooked pools Dune measured does that.

Indeed, only about a tenth of v4 TVL sits behind any hook at all, and every one of those runs swap logic, dynamic fee, or accounting code that leaves the tokens in the pool, according to the report. Not one is a rehypothecation hook routing idle capital into Aave or Morpho.

What the idle capital costs

Each dollar of in-range capital on the Uniswap and PancakeSwap pools earned about 35 cents a year in fees over the period.

The report said applying that rate to the idle capital puts the forgone total at roughly $150 million a year.

Idle LPs forgo on the order of $150M a year | Image: Dune.

By venue, Dune noted about $116 million on Uniswap, $25 million on PancakeSwap, and an estimated $6 million to $12 million on Aerodrome, whose Slipstream fees are dynamic and can only be bounded. Uniswap v4 is excluded from that calculation for the same reason.

The number is not additive, Dune noted. Fees come from trading volume rather than from deposited capital, and only in-range liquidity collects them, so moving all the idle money into range would split the same pot more thinly rather than create more of it. The $150 million measures what each idle LP gave up individually, as if it alone had repositioned while everyone else stayed put.

That same math explains why the 35% rate looks so high in the first place. A fixed pot of fees spread over a thin sliver of working capital makes that sliver earn a lot.

Fragmentation sits underneath all of it

Ether/stablecoin volume trades across roughly 170 venues a week, and the largest never carries more than 40% of it, averaging 28.8%.

Bitcoin (BTC) pairs concentrate more, as the top venue peaked at 57.8% for BTC/stablecoin as flow moved to Aerodrome on Base, and 52.5% for ETH/BTC. Which venue looks idlest depends entirely on the pair. Uniswap runs highest on BTC/stable at 41.9% and ETH/BTC at 35.1%, while PancakeSwap is higher on ETH/stable at 43.7% against Uniswap's 32.8%. Aerodrome reads lowest on all three. Pool the pairs together, and the venues converge into a tight band in the high 20s to low 30s.

Even stablecoin pools are no exception, running about 30% idle. LPs concentrate stablecoin liquidity into ranges a few basis points wide, so the small wobble a peg permits is enough to push a position out.

"Decentralized exchanges have grown into one of the deepest, most liquid markets in crypto, and it is now competing with centralized exchanges and traditional trading venues," commented Filippo Armani, research lead at Dune. "What our research shows is that it has reached this scale even though much of its liquidity is not yet fully at work."

Dune discloses a survivorship tilt in the panel, which grows from 559 to 776 pools as newer ones come into existence, and reconciles its reconstructed positions against onchain balances to within about 97% by value. Solana's concentrated-liquidity venues are excluded, as is Curve, whose LPs do not set price ranges at all.