RERSKIPROTOCOLAll pools

What the GMX pools paid

2026-09-19 – 2026-09-25 · Arbitrum · 59 pools with at least $100K of liquidity

Liquidity$174.85M+3.4% on the week
Volume$426.83M+108.6% on the week
Fees paid$205.8K+89.3% on the week
Fee yield6.1%realised, annualised

The capital in the pools turned over 0.35 times a day. The net liquidity flow from 2026-09-19 to 2026-09-25 (00:00 UTC), cleaned of the price move, was -$6.05M.

Who paid the most: net for liquidity providers

Ranked by what the pool kept: its fees plus its result against traders over the same seven days (realised plus the change in unrealised). A minus under “Against traders” means the traders won from the pool. A pool whose result against traders is not known comes last and is marked “—”, not taken as zero.

PoolFeesAgainst tradersNetNet APRFee APRTurnoverLiquidityPhase
NEAR/USD [WETH-USDC]$19.0K———149.2%13.49×$662.8Kpromising
FIL/USD [BTC-USDC]$2.5K———90.5%8.00×$144.2Kliquidity draining
BRENTOIL/USD [WETH-USDC]$1.3K———58.8%6.63×$113.9Kpromising
ONDO/USD [WETH-USDC]$3.9K———54.8%2.25×$371.8Kstable
AVAX/USD [AVAX-USDC]$2.3K———51.0%1.20×$232.2Kstable
SUI/USD [WETH-USDC]$5.5K———40.5%2.23×$705.1Kstable
VVV/USD [WETH-USDC]$5.2K———37.9%0.40×$710.8Kstable
ADA/USD [BTC-USDC]$893———32.2%1.61×$144.6Kstable
kPEPE/USD [WETH-USDC]$916———30.8%2.11×$155.1Kstable
LTC/USD [WETH-USDC]$953———29.3%1.28×$169.6Kstable

Where the liquidity went

PoolFlowChangeVolume vs the week beforePhase
DOGE/USD [WETH-USDC]$1.08M+108.0%+1137%stable
NEAR/USD [WETH-USDC]$986.6K+364.8%+2839%promising
SUI/USD [WETH-USDC]$377.7K+76.3%+548%stable
DASH/USD [BTC-USDC]$327.5K+250.3%+275%stable
VVV/USD [WETH-USDC]$241.3K+44.0%-17%stable
ETH/USD [WETH-USDC]-$5.75M-10.2%-35%cooling
ETH/USD [WETH-WETH]-$974.0K-5.0%+15%activity collapsing
HYPE/USD [BTC-USDC]-$906.2K-30.3%+44%stable
BTC/USD [BTC-BTC]-$776.1K-7.1%+150%stable
BTC/USD [BTC-USDC]-$382.9K-0.6%+103%stable

Phase changes this week

What we said a month ago, and before

Every pool is classified again as of 16 past days, a week apart — from 2026-05-14 to 2026-08-27 — using only what was known then, and measured against its fee yield over the following 30 days, without its result against traders. One observation is one pool as of one day: 1128 observations of 101 distinct pools. The windows of neighbouring days overlap, so the observations are not independent. The test is relative: did the pool beat the median of all pools.

Phase thenObservationsDistinct poolsBeat the medianMedian fee yieldWeakest … strongest
promising442477%14.7%1.8% … 135.6%
accelerating342553%7.4%0.5% … 40.2%
stable3495663%8.3%0.3% … 164.5%
cooling1084763%6.8%0.0% … 23.2%
activity collapsing472657%6.6%1.9% … 91.5%
liquidity draining271759%11.7%0.7% … 29.6%
too little data5197735%4.5%0.0% … 46.5%