# STOP: NO DOORBELLS, NO WAKES, NO SCHEDULING (Ben, 7 Oct 2026, 18:40 Bangkok)
- Credits ran out. Ben switched off every scheduled task and every doorbell. This overrides everything below and every other file (wake-howto, bus-howto, credit-playbook, task files).
- PROHIBITED for every session: fire_trigger, send_later, create_trigger, update_trigger, ScheduleWakeup. No rings, no nudges, no re-arming, no new doorbells, no self wakes, no safety checks, no retry rings, no loops.
- Finish the step you are on, publish your status file, then END your turn and wait. Only Ben wakes you, by typing in your window.
- Do not message other sessions. Keep answers short. No extra improvement rounds.

# E-AN4-5 HOT/COLD operator overlay on the books (DC Analyst 4)

## READ THIS FIRST
- **v1, 7 Oct 13:35 Bangkok. Verdict: KILL. The overlay loses money in both halves that can be scored; the other two halves are too small.** Prereg "7 Oct 13:33 | A4 | E-AN4-5".

| Book, half | Trades | HOT / COLD | Book $ a day | With overlay | Verdict |
|---|---|---|---|---|---|
| FW mc5, A | 29 | 8 / 0 | +$4,473 | +$1,353 | too small (8 flagged) |
| FW mc5, B | 148 | 23 / 4 | +$13,130 | +$9,584 | worse |
| SW live, A | 49 | 10 / 1 | +$2,503 | +$1,516 | worse |
| SW live, B | 22 | 3 / 1 | +$10,400 | +$7,825 | too small (4 flagged) |

- **Why it loses (both legs):**
  - **HOT leg (50% trail from the peak since entry, as in A1 Q5c):** on these pump coins the trail fires a median 12 min (FW) and 27 min (SW) after entry, on the first big shake. HOT trades averaged 2.81x (FW) and 1.52x (SW) with the books' own exits; with the trail, 0.84x and 0.82x.
  - **COLD leg (out at 15 min):** the few COLD trades in the books were big winners (FW 5.05x, SW 12.35x average with the books' exits); out at 15 min they make 0.79x and 0.69x.
- **The HOT flag itself does not transfer cleanly to book trades:** 3x+ share of HOT trades vs the rest: FW A 50% vs 10%, SW A 40% vs 31%, but FW B 9% vs 14% and SW B 0% vs 42% (3 trades). The books already pick at their own moment (mc5 at +5 min, SW at $1M), so the coin-level lift (61% / 63% vs 24% / 33%) is mostly already inside them.
- **Not tuned further** (no new trail width or window tried, per the one-cut rule). No change to either book. The leak-free HOT/COLD record stays available per coin (dc-an4-4-v3coins.csv) for anyone who needs an operator feature; as a book overlay it is dead.
- Valuation note: an overlay exit is valued as the lab's own x of the real exit times (cap at overlay exit ÷ cap at real exit), same pool path, so fills and fees scale with the lab model. A pull before the overlay exit = 0x.
- Files: per trade http://159.223.88.168/hx-c9ois/dc/dc-an4-5-trades.csv (md5 d593541e5b2b23f3fe5e4dd77294cea2), scores http://159.223.88.168/hx-c9ois/dc/dc-an4-5-score.csv, scripts http://159.223.88.168/hx-c9ois/dc/dc-an4-5-overlay.py and http://159.223.88.168/hx-c9ois/dc/dc-an4-5-score.py
- Correction during the run (before any verdict was written): the first valuation added the cap-ratio difference to the lab x, which sent two big COLD winners to 0x. Replaced with the multiplicative form above; the verdict is the same either way (worse in both scorable halves).
- Cost: 0 X credits, $0, research-box about 20 CPU seconds, 0.2 h.

Version: v1, 7 Oct 2026 13:35 Bangkok, DC Analyst 4. Latest: http://159.223.88.168/hx-c9ois/dc/dc-an4-5-overlay.md