BCR Forecast Scorecard

Triggers (set ~Feb-Mar 2026, all due Feb-Apr 2026): Feb payrolls sub-150K = confirmed / >200K = trouble; Feb retail <0.2% = confirmed / >0.5% = trouble; Feb core CPI 0.3%+ = strengthens; Black unemployment >8% w/ white flat = bifurcation accelerating. Source's own kill-switch: "wrong by end of Q1 2026."

Resolution (window closed ~May-Jun 2026): NO headline recession: May payrolls +172K, unemployment 4.2%, NY Fed Q1 2026 "delinquency transition rates hold steady" (aggregate 4.8%). Bifurcation leg CONFIRMED: card 90+ delinquency 13.12% Q1 2026 (15-yr high), TransUnion adopting literal "K-shaped" framing.

Lag: ~3 months from trigger window to definitive resolution.

Winning path: the COUNTER-RISK the source rated 15% (labor market holds): the LOWEST-probability branch on the page.

Lessons: (a) bottom-half stress was real but aggregate-blind: in a K-shaped economy the top-half consumer carries the aggregates, which breaks recession-TIMING calls specifically; (b) a low ex-ante probability on a benign branch is not evidence against it: n=1 here, but the ledger exists to count exactly this. (c) The structural read survived the tactical call's death: resolution can be PARTIAL, and the durable leaf pages (bnpl credit stress, subprime auto credit stress) are where the surviving substance lives.

Trigger fired: the pre-announced §122 150-day expiry date, 2026-07-24 (legal/calendar-date class). Interim state going in (2026-07-04 adjudication): CIT struck §122 2026-05-07 (counter-risk 2, 30%, fired); Fed Circuit stayed it 2026-06-11 (tariffs still collected); no Congressional extension (counter-risk 1, 20%, did not fire).

Resolution (2026-07-24, web-verified against Jul 23-24 press/primary sources): §122 lapsed by operation of law at 12:01 a.m. ET: but did NOT lapse into a vacuum. USTR re-imposed essentially the same wall under Section 301 (forced-labor investigation basis), two-tier 10% / 12.5% on ~60 economies (~99.4% of US imports), no rate cap, no expiry, effective the instant §122 died. Same-day legal challenge filed (Liberty Justice Center, on behalf of Burlap & Barrel + Collective Horology) arguing the administration cannot preserve a predetermined tariff policy by swapping statutes.

Lag: ~2.5 months from the CIT-strike interim trigger (2026-05-07) to date-resolution (2026-07-24); 0 days from the pre-announced expiry date to resolution (date-class triggers resolve ON the date by construction).

Winning path: the shipped video script's "Outcome three": administration pivots to §301/232, rated ~60% (the MODAL call): WON, including its explicitly named follow-on ("restarts the legal challenge clock"). NOT the thesis page's benign lapse-without-replacement counter-field (which would have removed the tariff-inflation leg). The three page counter-risks: Congress-extend (20%) did not fire; courts-strike-§122 (30%) fired-but-stayed; bilateral-deals (15%) did not fire.

Lessons: (a) the video's scenario tree was the sharper instrument than the wiki page's counter-risk list: the teleprompter gave the actual outcome (§301 pivot) ~60% and named the legal-challenge follow-on, while the thesis page carried only the three benign memo counter-risks + the lapse steelman and so under-represented the modal path. Fixed same session: Outcome-three added to the page's counter-risk section. Rule: a thesis page's counter-risk/scenario section must carry the SCRIPT's outcome probabilities, not just the memo's benign counter-risks. (b) legal-date triggers can resolve as continuity-by-statute-swap: "the cliff" is not binary lapse-vs-extend; a third mode is the same policy re-based on a different statute, which defeats the benign lapse path while CONFIRMING the core "fiscal dominance persists / no clean resolution" framing. (c) first legal-date-class pair (n=1 for this class): no lag prior transfers from pair #1 (data-print class).

Trigger fired: reassessment trigger 4, "demonstrated operational constraint": US munitions stocks visibly constraining a live operation (press-reporting class, a new trigger class for this ledger: no data print, no calendar date). Bar: two independent instances in 6 months, or one on-the-record DoD/allied-MoD confirmation. Fire date 2026-07-29, on three instances inside nine days: (1) 2026-07-20 WaPo, anonymous US official on THAAD: "we do not have enough to safely sustain operations"; (2) 2026-07-26 NYT: Iran escalation paused over dwindling interceptor stocks; (3) 2026-07-28/29 CNN: US forces deliberately allowing some Iranian missiles through depleted Patriot/THAAD defences. Supporting receipts (CSIS, late Jul 2026): <827 Patriot and <278 THAAD remaining, Patriot −~65% across the war, 3+ years to rebuild, no THAAD deliveries forecast in 2026.

Resolution status: INTERIM: the trigger resolved, the thesis did not: Trigger 4 is a reassessment trigger on a structural thesis (deadline: null, review_by: 2027-07-04). What it establishes is that the thesis's least-evidenced and most load-bearing link: depletion actually constraining policy rather than merely appearing in procurement arithmetic: is now demonstrated in the live world. What resolves the thesis is whether a genuinely multipolar arms supply materialises on the memo's 12–36-month clock from May 2026. Status unchanged at live; conviction on the premise rises, the conclusion stays open.

Lag: trigger 4 was created 2026-07-27 and fired 2026-07-29: 2 days, which is a misleading number and is recorded as a warning rather than a data point. The mechanism's first instance predates the trigger's existence (2026-07-20), and the trigger was written because the mechanism had already been observed firing through a channel triggers 1–3 could not see. True mechanism-to-recognition lag: ~9 days (first instance 07-20 → adjudication 07-29). ⚠️ Do not enter retro-fitted triggers into lag statistics: a trigger authored after its own mechanism starts firing will always show a near-zero lag and will corrupt the class prior.

Winning path: CONFIRM, against the counter-field. Notably it is a confirm on the mechanism/premise leg only: the three memo counter-risks are all untouched and stand at their ex-ante probabilities (US production catches up 35%, allied capacity overstated 30%, US export dominance stickier 30%), and the page's strongest contrary evidence (SIPRI: US arms-export share RISING 35%→43%) bears on the conclusion, which this trigger does not address. The counter-field sensor (US production catching up) remains live.

Lessons: (a) ⭐ first pair where the winning path IS the page's headline framing: breaking the 2-of-2 streak in which it was not (see Pattern observations). Note the asymmetry that makes it a weak break: this confirmed a mechanism, and mechanisms are easier to confirm than outcomes; pairs #1 and #2 both died on outcome/timing while their structural reads survived. Watch whether "mechanism confirms, conclusion still fails" becomes the recurring shape. (b) On-the-record authority can point the wrong way, and behaviour beats testimony. The trigger privileged on-the-record DoD confirmation; DoD was on the record denying the constraint (SecDef Hegseth: enough munitions "as long as we need to") while forces that had been intercepting everything began choosing not to. The ruling discounted the denial as expected wartime institutional posture: but the contradiction was kept on the page, not resolved away, and the citation caution stands (never cite as DoD-confirmed). Rule extracted: when a trigger names an authority as its evidence standard, pre-state what happens if that authority denies the thing. (c) A trigger's enumerated forms will under-cover its own intent. Clause (b) listed inter-theatre and US-vs-allied rationing; the winning instance was within-theatre triage: the purest case of scarcity dictating allocation, and not on the list. Amended at adjudication as a wording repair. Check enumerated-list triggers for missing forms at every adjudication. (d) Anonymous sourcing counted toward the multiplicity leg because the text barred only a single anonymous report: read trigger text as written, and write it knowing it will be read that way.