ETH Exit Queue: Record Backlog

ETH Exit Queue: Record Backlog

D

TL;DR

Between August 18 and 20, 2025, the validator exit queue grew to a little over 910k ETH, an all-time high, and the wait towards full unlock was approximately 15 to 16 days, though this increments in real time. Under today's throughput limitations, both exits and activations are capped at about 256 ETH each epoch, or about 57,600 ETH each day. There is a queue of nearly 910k ETH, and simple capacity math implies a wait of about sixteen days if new requests do not accelerate. A long queue raises carry costs for arbitrage and has the consequence of forcing markets to demand a larger discount for liquid staking tokens (LSTs) compared to ETH, which elevates depeg risk and can cause forced deleveraging in DeFi loops such as those built on Aave.

How the queue works and why it grew

Exiting is a two-stage process. First, a validator initiates exit and waits in the protocol's exit queue. Once the exit is processed, withdrawals finalize: partial withdrawals, i.e., rewards above the effective balance, are streamed automatically, but full withdrawals only once the validator has finished exiting. Post-upgrade constraints restrict total exit throughput to around 256 ETH per epoch, or around 57.6k ETH per day. As requests queue, total wait time rises roughly in proportion to the size of the queue. Recent queue growth has been driven by the unwind of leveraged "LST to borrow ETH to buy more LST" loops as variable borrow rates rose, along with some operator rotation and profit taking. Those dynamics pushed more LST into spot markets and more validators towards protocol redemptions.

Pressure channels for LST pegs and derivatives

The main avenue is spot trading of LST vs ETH on AMMs and centralized exchanges. Longer wait times increase the discount that market makers have to warehouse LST vs ETH, which widens spreads and pulls LST below one to one. Imbalances in core pools such as stETH/ETH on Curve can intensify a depeg in the moment and extend the round trip required for arbitrage to restore parity. The second channel is leverage on top of LST collateral. When ETH borrow rates surge on Aave or Morpho, the classic loop is no longer profitable, positions are closed, LST is sold on spot, the discount gets worse, and new exit requests follow. A third channel is yield derivatives and forwards. The longer the line, the larger the waiting premium arbitrageurs require for the trade of buying discounted LST and waiting it out for redemption. Since funding costs and waiting time rise together, the equilibrium discount tends to expand. Eventually, stress in underlying LST markets overflows into liquid restaking tokens that employ LSTs as collateral, increasing slippage, rising leverage costs, and undermining secondary pegs.

Where to look for vulnerabilities

Start with network level trackers that give queue size and estimated time to exit, along with validator analytics that estimate when a new exit would actually finish. For markets, start with the large LST to ETH pools such as stETH/ETH and frxETH/ETH on Curve, and wstETH/ETH on Uniswap or Balancer. Deterioration will generally express itself as a rising percentage of ETH in the pool, a persistent discount of the LST to ETH pair, and range liquidity contracting. For credit markets, track variable ETH borrow rates and utilization on Aave and Morpho. Sudden rises in borrow APRs and increasing utilization have a tendency to precede loop unwinds. On the news flow side, be on the lookout for announcements from major staking operators like decentralized protocols and centralized staking desks, as concentrated waves of exits and liquidity management changes can contribute to ongoing stresses.

Implications for ETH and DeFi liquidity

Because throughput is limited at approximately 57.6k ETH per day, a flow of new exit requests at or near that rate will maintain the queue at high levels or boost it higher, which consequently maintains LST discounts wide. The main stress channel remains leverage unwind: expensive ETH borrowing rates accelerate unwinds of recursive positions, force LST selling on spot, drains balanced liquidity out of pools, and raises arbitrage costs. It is also possible for ETH spot price to decouple from on-chain stress if external demand such as ETF flows hold price, but the on-chain liquidity layer can still tighten even if the headline price is range-bound or higher.

Scenarios

In the base case, new exit demands arrive at approximately the same pace as protocol throughput, the queue remains between ten and sixteen days, LST discounts stay moderate, and credit markets adapt to the new equilibrium. In the bear scenario, another bout of borrow rate surges or intense operator rotation stretches the queue beyond sixteen to twenty days, depegs are larger and more common, and deleveraging accelerates. In the bull scenario, as borrowing costs return to normal and validator flows stabilize, the queue shrinks towards one week, pegs narrow, and LST market depth increases.

Practical triggers and an operating checklist

Use the queue size and its ETA as regime markers. Under ten or so days, pegs are neutral; between ten and sixteen days, vigilance is heightened; over sixteen days, markets tend to require a plumper LST discount to ETH as a recompense for time and risk. Track the live LST discount on CEX and DEX, as a sustained spread in the area of half a percent to one percent in the face of a long queue has a tendency to imply structural rather than ephemeral stress. Keep an eye on ETH borrow rates on Aave and Morpho, as spikes into the low teens and beyond have a tendency to coincide with loop unwinds and new waves of exits. Finally, monitor AMM liquidity structure; declining depth near the mid price and, within LST pairs, a rising ETH proportion are early signs pressure is building.

Report Page