DeFi Has a $1.6 Billion Efficiency Problem. The $150 Million Number Has a Catch

DeFi has roughly $1.6 billion sitting in concentrated-liquidity pools without doing much of the job investors deposited it to do.

That number is real. It is also easy to misunderstand.

A new Dune study found that about 85% of the concentrated liquidity it measured was underutilized during the first half of 2026. The stricter figure is smaller: 29.5% was fully outside its chosen price range, earning no trading fees at all.

The difference between those two numbers is the entire story.

Dune’s 22-page study, commissioned by 1inch, rebuilt positions in roughly the 200 most active pools on each of Uniswap v3, Uniswap v4, PancakeSwap v3 and Aerodrome Slipstream. It took 26 weekly snapshots from January 6 through June 30 across seven chains.

The panel ranged from 559 to 776 pools as newer pools appeared. Average total value locked was about $1.84 billion, and each snapshot was tied to an exact blockchain block rather than a loose dollar estimate from a dashboard.

Across those weeks, an average $542 million sat completely out of range. Its tokens remained in the position, but the current market price had moved beyond the band where that position could serve trades and collect fees.

1/ Across H1 2026, an average of 85% of concentrated-liquidity capital was underutilized and 29.5% sat outside the active price range. That works out to roughly $542M idle in a typical week.

New onchain research from Dune, produced for @1inch. pic.twitter.com/VeI5SqEuWd

— Dune (@Dune) July 16, 2026

Concentrated liquidity lets a provider choose the prices where its capital will trade. A narrow band near the market can earn more fees per dollar than capital spread across every possible price.

The tradeoff arrives when the market leaves that band. The position turns into one side of the pair and stops earning until the price returns or the owner moves the range.

Dune divided the measured v3-style capital into three buckets. Only 13.7% was actively crossed by trades during the measured period, while 56.9% remained in range but sat beyond the prices that trading actually reached.

The remaining 29.4% was out of range. Rounding and slightly different protocol coverage account for the nearby 29.5% headline figure.

Calling the whole 85% “idle” makes for a cleaner headline, but it blurs two different conditions. In-range capital can still protect execution around the market and may become active with a relatively small price move, even when that week’s trades never touch it.

Out-of-range capital is the hard floor. Its fee meter is off.

There is another useful comparison buried in the study. When Dune applied the same traded-band test to older constant-product pools, 98.7% of their capital was underutilized.

Those pools spread capital across all prices by design. Concentrated liquidity brought the underused share down from nearly 99% to about 85%, so the new system is more efficient even though most of its capital still misses the day’s action.

The Defiant highlighted an important interest behind the research: 1inch commissioned it, and 1inch is developing liquidity technology that benefits from the argument that DeFi capital needs to be shared or managed more efficiently.

That does not invalidate Dune’s onchain reconstruction. It does make the definitions, limits and counterfactual fee calculation more important than the promotional version of the findings.

The position history points to a real behavioral problem. Of the capital already out of range, 36.7%, or roughly $200 million, had gone more than 90 days without an owner adding to or withdrawing from that range.

Dune calls those positions dormant, with a useful qualification: untouched does not prove forgotten. An owner may be deliberately waiting for the market to return.

Distance makes that patience harder to defend. About 43% of out-of-range capital was more than 25% away from the current market, and 17% sat more than 100% away.

A position that far out requires the relevant price ratio to double or halve before it becomes active again.

The owner data shows who manages that risk best. Automated contracts, vaults and market-making systems kept their ranges active far more reliably than individual wallets.

Base offered the cleanest example. Contracts held about half the Uniswap v3 capital on the network, yet individual wallets accounted for 82% of the idle amount.

The machines were not smarter about market direction. They were simply present to recenter the position when the market moved.

I’ve provided liquidity on-chain more times than I can count.

But one question always stays in the back of my mind:

Is my liquidity actually doing anything?

I came across a comprehensive report by @1inch and @Dune that tracked

concentrated liquidity across:

– Uniswap
-… pic.twitter.com/fuo64a9XNO

— Param (@Param_eth) July 16, 2026

Uniswap v4 has not solved the problem yet. Its measured pools held about $230 million at the end of June, and an average 30.5% was out of range, almost identical to the older design.

V4 hooks can attach custom code to a pool and could, in principle, move parked tokens into a lending market while they wait. Dune found that only about one-tenth of measured v4 value sat behind any hook, and none of those hooks rehypothecated idle assets into an outside yield strategy.

For the pools in this study, out of range still meant earning nothing.

Then comes the number most likely to be repeated without its warning label: $150 million in annual foregone fees.

The Block summarized the estimate as roughly $116 million on Uniswap, $25 million on PancakeSwap and another $6 million to $12 million on Aerodrome. Uniswap v4 was excluded because its dynamic fees could not be priced with the same method.

Dune reached the estimate by applying the annualized fee rate earned by in-range capital, about 35 cents per dollar, to the capital that was out of range.

But trading fees come from trading volume. Moving every idle dollar back into range would not create another $150 million of fees; it would divide the existing fee pool among more capital and reduce the return earned by each dollar.

The estimate is therefore personal and counterfactual. It asks what each out-of-range provider might have earned if that provider alone had recentered while everyone else stayed put.

It cannot be summed into a new pool of money waiting for all liquidity providers to collect at once.

That caveat does not rescue a position that has sat 100% away from the market for three months. It does keep the $150 million figure from becoming imaginary industry revenue.

The practical choices all carry costs. Active managers can recenter a range, but they add transaction fees, contract risk and management fees.

Hooks or wrappers can lend unused assets, but that adds another protocol and another failure path. Wider ranges demand less maintenance while lowering the fee concentration that made the position attractive in the first place.

Concentrated liquidity did improve the old automated-market-maker model. It also quietly changed liquidity provision from a passive deposit into a market-making job.

Professional systems show up for that job. A large share of individual wallets do not.

The $1.6 billion problem is not that every dollar is dead. It is that DeFi has made capital efficiency available without making active management automatic, cheap or safe.

Join the conversation!

We have no tolerance for comments containing violence, racism, profanity, vulgarity, doxing, or discourteous behavior. If a comment is spam, instead of replying to it please click the icon below and to the right of that comment. Thank you for partnering with us to maintain fruitful conversation.