F Opal dragon bolts (e) — patient 4h buy-limit spread flip vetoed

Opal dragon bolts (e) (enchanced, members) sustain a post-tax spread of ~110–150gp on ~442k units/day of real volume. Spread persists in the 15-min block history across 48h: 80%+ of 15-min blocks show avg_high - avg_low >= 80gp. The 30d daily history is range-bound (avg_high 2522→2956 then back to ~2930; no secular trend). Seasonal avg-margin is biased POSITIVE across most hour-of-week buckets (positive bursts at bucket 87, 95, 96 of 100+ gp), unlike Dragonstone (which shows a strongly negative bias and was rejected). The market consists of PvM/range-grind sellers dumping at market and PvM buyers paying the insta-buy price - neither side places patient offers, leaving a flippable spread. Capital (11000 x 2803 = 30.83M) fits the 50M research budget on its own.

Entry: Place a buy offer at 2803 gp (current low) at any time, up to full 11000-unit buy limit per 4h window. Use the live quote refresh every 5 min; if both legs remain fresh (high_age and low_age each < 10 min) and the spread is >= 100gp, refresh the buy at the new low.
Exit: Once the buy fills, place a sell offer at 2979 gp (current high, post-tax). If the high softens, walk the sell down by 5-10gp increments. Don't hold overnight if the 15-min block spread (avg_high - avg_low) drops below 60gp.

Kill switch: If the 15-min block spread drops below 60gp for 6 consecutive hours, stand down. If the 7-day avg_low drifts down by more than 5%, the trend-confound has eaten the spread - exit any open position. If the live quote drops margin below 60gp, treat the thesis as dead.

From run 455 · F-opal-dragon-bolts-e-20260801