SCALING — PLAN OF RECORD (am. 8/13)
HOLD 10%/window · milestone ② storm: MET (S1 8/10 — flat by its own gates)
① —/50 gated windows (— trades) → TIER-1-MINI: external capital, flat $1–1.5k windows, ~50 trades — fill quality at 2× size (capital go = Jake, at the crossing)
at 100 → full TIER STAIRCASE — flat tiers, step on receipts, find the capacity top.
AM pilot stepped $45→$150 (8/13, bar met 42–6). Amended 8/13; not calendar-bound.
MAIN STRATEGY — LIVE $
12m favYES · FV edge ≥ 0 · 10%/window · since 0.00 flip
AUTO-RATCHET — SIZING LADDER
pre-registered 8/20 · session cohort only · steps need hit ≥88% · de-risk −1 rung on 3 losses/10 · beyond $2,500 = human review
LANE PILOT 7–9M — LIVE $
main's FV rule in the 7–9.8m zone main never sees · $20 clips · weekday day-session · regime-gated · observed cohort counts toward the bar (Jake 8/18) · bar n≥50
SHADOW — LANE COHORT · STRICT (BTC+ETH)
first FV edge ≥ 0 look in 7–12.5m · weekdays · day-session · executor-open windows only · net-new vs main · the lane's evidence record
SHADOW — FV NIGHT 75–89¢ (BTC+ETH)
what the CURRENT gate would trade 21–06 ET · edge ≥ 0 · hypothetical $45/leg · the nights-off jury
MARGINAL HARVEST — LIVE $
model-declined favourites, edge −20..−5¢ · 06–15 ET weekdays · auto-ratchet legs (now $350) · $1,000/day cap · rolling-30 breakeven kill · era cohort 114-16, ~+10%/$ (n=132)
BOUNCE — SHADOW (live retired 8/19)
wreckage thesis, retired to observation at 26/40 fills: 8-18, +$42.93 (+23.8%/$, 31% hit vs ~27% BE) · rule still scored offline from the 1Hz ask-path · revivable on evidence
STORMWATCH
liquidations (OKX, 5 coins) · options fear (30d IV + 25Δ skew) ·
ⓘ how to read
What this panel watches: two kinds of market fear.
1. Forced selling (the "liq" line). Many crypto traders bet with borrowed money.
When price moves against them hard enough, the exchange force-closes their position —
a liquidation. A trickle of these is normal background noise. The line shows the
last 5 minutes of forced selling vs. what's typical ("norm"), as a ratio. The danger is a
chain reaction: forced sellers push the price down, which forces MORE sellers, and
so on — a cascade. Our tripwire (S1 CASCADE):
forced selling over 10× normal with ≥70% of it being longs (people
betting up getting flushed). When that fires, the next few 15-min windows have
historically been dangerous for favorites — it's the market's fire alarm.
2. Crash insurance prices (the IV / skew lines). Options are insurance contracts on
price moves. IV is how bumpy the pros expect the next 30 days to be — higher =
rougher roads expected. Skew compares the price of crash insurance vs rally
insurance: positive means people are paying extra to protect against a fall. The level
matters less than the change (the "Δ vs 7d" number): when skew jumps ≥3pts
richer AND IV spikes together (S2 FEAR),
the professionals are bracing for something.
How to read it in 2 seconds: everything small and grey = calm ocean. A red pill =
a storm signature is live right now — expect the main book to go quiet and treat any
open position's loss as the storm tax, not a malfunction.
Note: this gauge doesn't control trading yet. It's collecting evidence —
its three storm hypotheses get scored in September, and only rules that predict correctly
get wired into the book.