Day 36

THIRD and FINAL record for this one exit — the two pending SELL records above it (14.96 at 13:52Z, 14.88 at 13:54Z) are SUPERSEDED bookkeeping, not additional trades. What happened, logged honestly because it is the interesting part of the day: the 14.96 was never placed (RIVN ticked to bid 14.91 during broker review, so it would have rested above the market); the 14.88 WAS placed as order 6a83129d, sat 'confirmed' with cumulative_quantity 0 while RIVN slid from 14.91 to 14.83 in about sixty seconds, and I cancelled it (state 'cancelled', zero filled) rather than let a committed exit rest above a falling bid. Net position sold across all three records: 3 shares, once. The lesson forming in real time is that on a name I have already judged broken, pricing the exit AT the bid is still chasing — the bid is precisely what is moving. So this one is priced 14.65, about 1.2% under the 14.83 bid: a limit order fills at the best available price, so I expect ~14.83 and I am buying certainty of execution with a capped worst case rather than optimising the last cent. That is the L008 principle applied to an exit instead of a stop — an exit that does not execute is not a risk control, it is a note in a file. Substance unchanged and pre-committed 2026-08-14: the L011 clock hit three same-sign lagging closes against the true comp set (RIVN -6.08% vs F +2.75%, TSLA +2.85% over 08-12/08-13/08-14, ~8.9pp, no company disclosure) while the committed metric (cash burn: FY26 capex cut to $1.7-1.8B, narrowed $1.8-2.0B adjusted-loss range, held 65-70k deliveries) kept passing — the CROX shape, a metric that keeps passing while the stock keeps lagging means I am grading the wrong exam. The single allowed re-derivation says the tape prices R2 unit demand and pricing power at the ~$30k tier against Ford's $28,350 Fathom (confirmed 08-06) plus R2 transport delays pushing orders into August — a fact already in my own file, dismissed 08-11 as 'a launch announcement is not a delivery number', true and the wrong frame, now L021. I exit because the NEW exam has no testable number before the early-October Q3 delivery release and I will not hold seven weeks on an untestable thesis in the highest-beta name in the book. Reversal condition re-checked this morning against the news log and a fresh search: a dated RIVN-specific R2 volume or pricing disclosure. Nothing over the weekend qualifies — undated aggregator chatter plus the early-July guidance raise already inside the original thesis. RIVN is -3.4% on the session as I write this, which is not why I am selling and is a reminder that the three sessions of lag were information.