KoinlyticsDune published a study on July 18 that quantified something every honest LP already suspected. Roughly $1.6 billion in DeFi concentrated liquidity failed to earn returns in the first half of 2026. Weekly, about $542 million worth of positions sat completely outside their trading range. Zero fees. Zero market depth. Just capital rented to the counterparty for free.
Why the APR on the pool page is not your APR
Every concentrated liquidity dashboard quotes an aggregate pool APR. That number divides fees earned by the pool by the total liquidity in the pool. It assumes uniform utilization. Real positions are not uniform. Some positions are dead center on the current tick and earn every trade. Some drifted 30 ticks below the tick a week ago and have not caught a single swap since. The pool APR reports the average. Your position is either well above or well below it. The average is almost never what you get.
The number that matters is your position's realized APR, which is your accumulated fees divided by the time-weighted capital you actually had at risk in the active range. That number is invisible on 90 percent of tracker dashboards because most trackers just show your current position value versus initial deposit. That comparison sweeps two errors together: fee income and impermanent loss.
Where the idle capital lives
Dune's decomposition is worth reading in full. The concentration of idle capital is not evenly spread. Three patterns dominate. First, positions from the March and April volatility spikes that never got re-ranged after price mean-reverted. Second, stablecoin-stablecoin pools that used tight ranges around a peg that then depegged briefly. Third, LST/ETH pools where the LST slowly drifts against ETH and slowly walks positions out of range without triggering any alarm.
The tracker angle
This is the exact problem Koinlytics was built to surface. Portfolio truth for an LP is not the pool APR quoted on the DEX. It is: current position value, fees accrued while in range, time spent out of range, impermanent loss versus a hold baseline, and the deviation from a hypothetical uniform position. We publish all five for every Uniswap V3, V4, Orca Whirlpool and Raydium CLMM position by wallet. If your capital has been idle for two weeks, we tell you now, not when you next log in.
The pattern to watch
Expect more research pieces along the same lines through August. The idle capital number is easy to compute and impossible to hide once you look at it. That pressure is going to push protocol teams toward auto-rebalancing hooks and away from pure LP-managed ranges. V4 hooks will take share. Passive V3 will continue to bleed. If you are running the older model, this week is a good time to audit which of your positions are earning and which are just parked.