Blast exit wave: where do ~15,000 queued ETH and $22M of DeFi TVL go before Oct 26?

Object of investigation

Blast L2 exit capital on Ethereum: ETHYieldManager 0x9807...C8FE, OptimismPortal 0x0Ec6...b6Cb, L1BlastBridge 0x3a05...9115, USDYieldManager 0xa230...8438, and the wallets that receive withdrawals

Observation

Sourced facts (times in evidence). Blast is shutting down: users must withdraw to Ethereum by 2026-10-26; withdrawal delay cut to 24h with a roughly one-week pause while Lido assets are withdrawn (Phemex, 2026-10-02). Blast docs list the L1 contracts above. Etherscan tags 0x9807...C8FE as 'Blast: ETH Yield Manager Proxy'; its cached page (snapshot 2026-09-09 23:14 UTC) showed 16,581.61 stETH and 1,609.34 ETH. L2BEAT showed Blast TVS $67.25M (read 2026-10-13 ~13:00 UTC). DefiLlama Blast chain TVL: $23.02M on 2026-10-01, $22.20M on 2026-10-13. Etherscan tags 0xf70d...dbef, a large recipient of Blast bridge withdrawals, as 'Relay: Solver' (read 2026-10-13 ~13:00 UTC).

Hypothesis

Interpretation, not established fact. The author's indexed Ethereum transfer data (non-public, verifiable on Etherscan) for 2026-09-25 to 2026-10-13 shows: about 16,427 ETH returned from the Lido withdrawal queue to the ETH Yield Manager; only about 1,429 ETH paid out to users via the canonical bridge (weekly 329, 818, then 282 ETH partial); about $1.18M of DAI/USDC out of the USD Yield Manager. Roughly 15,000 ETH (~$37M) and most Blast DeFi TVL therefore have not exited, so the 13 days before 2026-10-26 should concentrate a large exit wave. Of the 38 largest non-hub ETH recipients, 29 (about 77% of their ETH) had not moved it onward. Relay's solver took about 20% of ETH exits and sent 2,608 ETH into the Base bridge, hinting Base receives fast-bridged capital. If exits keep parking on L1, the wave is not sell pressure; if it routes to exchanges or one destination chain, that is a measurable rotation.

What would invalidate it

Invalidated if: (1) public explorer data shows the ETH Yield Manager already paid out most of the roughly 16,400 ETH before 2026-10-13, so no wave remains; (2) by 2026-10-27 less than 30% of the remaining balance has exited, showing users are not racing the deadline; (3) labelled exchange deposit addresses receive more than 50% of exited ETH within 7 days, meaning the flow is sell-side rather than parking or rotation; or (4) fast-bridge data shows no single destination chain receiving more than 25% of Blast-origin volume.

Published observations provide a starting point, not a verified conclusion.