Subway Robinhood Chain · 46630x71C7…976F · mock
Mock data.

No contract is deployed on Robinhood Chain, no addresses are pinned, and the keeper does not exist. The hedged vault (hAMC) does not exist yet — on-chain, hedgeAvailable() is false for every pair and the hedged paths revert. Fee APRs are extrapolated from one ~15-minute pre-open window on 2026-09-16, and decay. Balances, NAV, queues and PnL are invented.

How the fee APR is computed

Extrapolated from one 15-minute pre-open window, annualised, and never netted against adverse selection.

Every fee APR this app shows is a screen, not a forecast. It comes from research/scan.ts, measured over one 15-minute pre-open window on 2026-09-16.

The four steps

  1. Sum the quote-side notional of every swap in the window, times the fee tier.
  2. Scale that to a day (×96), then to a year (×365).
  3. Take the vault's cut, e / (active + e), where e is $25K × 33.8.
  4. Divide by the $25K vault size.

Why step 3 multiplies

A vault concentrated over ±6% is worth 33.8× its face value against a pool's full-range-equivalent active liquidity. Its dollars count that many times over while price sits inside the range.

Getting this wrong moves a pool's APR by more than an order of magnitude. The first Phase 0 report modelled the vault as full-range and understated fees by roughly 34×.

What the number is not

Annualising fifteen minutes is why these are enormous

A pre-open window is not a representative session, and the figure decays. No markout was ever run, for any pair, so none of it is net of adverse selection — the cost that motivates this entire product is absent from the number used to screen it.

The gate was 10.95%, set at twice the 5.47% hedge carry. See what is built for how the named pairs scored against it.