Skip to main content

Extraction timing traces (L2→L1→exchange)

This evidence pack attempts to strengthen (or falsify) the on-chain leg of the yield-extraction thesis by linking a tighter, repeatable sequence than a one-off exit:

  • Arbitrum WithdrawStake (liquid LPT leaves BondingManager)
  • Arbitrum bridge-out (LPT burn via gateway router)
  • Ethereum L1 escrow release (LPT transfer from the Livepeer L1 escrow)
  • Tight-window routing into labeled exchange endpoints (best-effort; often via a repeatable second-hop EOA)

This is still not proof of delta-neutral hedging (the hedge is mostly off-chain). It is, however, a measurable “cashout routing + timing” fingerprint that is harder to explain as a single discretionary exit.

Summary

  • Generated: 2026-01-22T00:46:05.137369+00:00

  • Arbitrum RPC: https://arb1.arbitrum.io/rpc

  • Ethereum RPC: https://rpc.flashbots.net

  • L1 token: 0x58b6a8a3302369daec383334672404ee733ab239

  • L1 escrow: 0x6a23f4940bd5ba117da261f98aae51a8bffa210a

  • L2 window: 5856381423863641

  • L1 window: 1460000024286807

  • Senders analyzed: 10

  • Burn (bridge-out) events: 85

  • Matched WithdrawStake→burn (≤ 72.0h): 75

  • Matched burn→L1 escrow receipt (≤ 60.0d): 84

  • Matched L1 receipt→labeled exchange (≤ 72.0h forward, then ≤ 72.0h to exchange): 68

Sender table

Columns: Arbitrum bonded stake now (snapshot), number of bridge-outs, and how many cycles can be followed all the way to a labeled exchange endpoint with tight timing windows.

Sender (L2)Bonded now (LPT)BurnsMatched withdraw→burnMatched burn→L1 receiptMatched receipt→exchangeMedian burn→receipt (d)Median receipt→exchange (h)
0x3d6182c59dbbbbc648570762da316ac8404816ae0.000161616118.080.19
0xc5519fd1129d6d22744e0ac491401fff45d265282,322,141.979181617167.342.85
0x962b029508b1054e2af4184bbaeb5d0c796f75260.0001110
0x60b8eb0947d25194afa248b81a087214beec5cfe0.00020227.010.18
0xef83273cbd014c4ae7998467c422275a8b37827e0.0001110
0x543df23b9a881fbdbdec7ab3e90f3ff7b905068b0.0001110
0x86abf78ac7ef44423873dabff35fe3e462b1ff6e283,800.2791281277.060.27
0x8d37a11a65b4f2c541af1312bc44f74a078160c70.000111110.680.39
0xadc6e5cbde4fbac7baf58a336edeab8590625baf0.0002220
0xa3bd517dcbdc063c4c24f0d9837bbc5ce869d092592,264.609312931317.040.19

Notes + limitations

  • This report relies on the canonical Arbitrum bridge-out signature captured in research/arbitrum-bridge-out-decode.json.
  • “Exchange” routing is label-set based (best-effort). The absence of a labeled exchange does not imply the tokens weren’t sold.
  • Many flows route via one or more EOAs on L1 before an exchange deposit; this report only follows one intermediate hop.

Raw output: see research/extraction-timing-traces.json.