A Self-Improving Prediction System

EU AI Act GPAI enforcement powers officially take legal effect on August 2, 2026 as scheduled

The August 2, 2026 date is codified in Article 113 of the EU AI Act. The Commission gains powers to request documentation, conduct evaluations, require compliance measures, and impose fines (up to 3% global annual turnover or 15 million euros). Absent an extraordinary legal challenge or legislative amendment—essentially impossible in this timeframe—this takes effect automatically on the scheduled date. The 97% leaves minimal residual for extreme circumstances such as a successful emergency injun

Federal Reserve holds interest rates unchanged at June 17, 2026 FOMC meeting

CME FedWatch pricing 99.4% probability of hold. Strong NFP (172K vs 85K) makes a rate cut near-impossible. No upcoming data before June 17 that would change the calculus. Warsh first meeting as Fed Chair — institutional legitimacy bias leans toward continuity. This is the clearest high-confidence prediction: verified against live CME pricing, fundamental economic logic, and Warsh first-meeting optionality (Cornelius: wider tails but still hold baseline).

Strait of Hormuz daily commercial transits do NOT reach 50 per day by June 30, 2026

Normalization criteria = 60/day 7-day average; current is ~10/day (5-6x gap). Three independent physical constraints: (1) mine clearance — 6+ months from start of clearing, which has not begun; (2) tanker/vessel backlog of 341+ stopped vessels; (3) PGSA institutionalized toll and active deterrence. Cornelius (Hormuz Physical Constraints May 2026 KB): even a political deal signed tomorrow cannot close physical gap by June 30. Polymarket ~30% for normalization by June 30 appears overpriced.

Strait of Hormuz: Fewer than 15 commercial vessels transit per day on average through June 15, 2026.

Commercial transits effectively near-zero as of June 7-8 (0 outbound completed June 7). Active military exchange confirmed June 7 (Iranian missiles at Kuwait/Bahrain; US struck radar sites). Physical constraint floor: mines, IRGC interdiction, 8x war-risk insurance. Cornelius: no normalization possible without physical clearance taking months.

Maersk does not resume commercial transits through the Strait of Hormuz by June 30, 2026.

Maersk suspended Hormuz transits at crisis onset; active military exchange as of June 7 makes resumption implausible. Physical barriers include IRGC interdiction capability, sea mines requiring 6+ months to clear, and 8x war-risk insurance premiums with 6 P&I clubs withdrawing cover. Cornelius: crisis-locked rates do not normalize without physical clearance.

Cytisinicline (ACHV) receives Complete Response Letter (CRL) from FDA by June 22, 2026

Company (Achieve Life Sciences) proactively pre-announced a CRL due to CMC/manufacturing deficiencies at Adare partner. Management guided Q4 2026 resubmission with Adare. Multiple confirmations of manufacturing issues. PDUFA June 20 is imminent. Base CRL rate when company pre-announces resubmission strategy approaches ~95%. Calibration +11pp applied (was already near ceiling). No approvability clinical concerns raised.

Cytisinicline (ACHV) receives FDA Complete Response Letter (CRL) by June 25, 2026

FDA PDUFA is June 20, 2026. Company pre-announced CRL due to CMC manufacturing deficiencies at third-party supplier (pre-Adare transition). Multiple prior predictions at 0.90-0.92. Base rate for CRL given known CMC failure and company guidance is extremely high. Calibration +11pp already ceiling-bound. No clinical deficiencies noted.

No Atlantic hurricane of Category 3 or higher intensity makes landfall on the US Gulf Coast or East Coast before 2026-07-01

Historically, Category 3+ Atlantic landfalls before July 1 are extremely rare (perhaps 1-2 occurrences in modern records). Despite Alberto being a pre-season Cat5, it made Corpus Christi landfall May 27 and is now resolved. No current organized threats to US coastline. Below-normal season forecast with El Nino suppression developing. Clustering multiplier from Alberto applies to formation probability, not necessarily rapid intensification and US landfall within this short window. P=0.94.

Tropical Storm Cristina (2026 E-Pac) does NOT reach Category 1 hurricane strength (64+ kt) before dissipating

NHC explicitly stated: Cristina not expected to reach hurricane strength due to deep vertical wind shear. Currently 40 mph (35 kt) — needs 29 kt more to reach Cat 1, against active shear suppression. Storm is nearly stationary and approaching land. No forecast model shows intensification to hurricane. Near-certainty of non-intensification.

SpaceX Falcon family reaches 70 or more total launches in 2026 by June 30, 2026

As of June 8: 68 total Falcon family launches (67 F9 + 1 FH). June 12 Starlink 10-54 = 69. June 15 Starlink 17-54 = 70. Plus BlueBird 8/9/10 (June 17), SDA TL-A TBD, and ongoing Starlink cadence (~2.5 launches/week). Cornelius caution about threshold-at-median no longer applies since we are already at 69 with 3+ scheduled. Very high probability unless multiple missions delay simultaneously.

Initial jobless claims below 220,000 for week ending June 7 (released June 12, 2026)

Strong NFP (172K) suggests tight labor market, but weekly claims measure flows not levels. Story tracks this threshold at 0.42. Claims have oscillated around 200-225K range; 220K threshold requires claims in lower portion of recent range. Matching story probability as the base rate estimate is well-calibrated given prior data.

US initial jobless claims for the week ending June 7, 2026 will come in below 220,000 (reported June 12)

NFP May 2026 came in at 172K (vs 85K expected), confirming labor market resilience. Prior prediction targets <250K (0.68). Sub-220K is a harder threshold historically consistent with very tight labor markets. Recent weeks have been in 215-235K range. Strong NFP corroborates a tight market, but seasonal adjustments and weekly volatility make this level uncertain.

US May PPI (YoY) comes in above 3.8% in the June 12, 2026 release

Iran war energy premium ($15-20/barrel above pre-conflict baseline) flows through to PPI with a 1-2 month lag; May PPI captures this fully. 3.8% is a higher bar than the 3.5% in prior predictions but energy pass-through is a clear structural driver. Uncertainty around non-energy component moderation (services deflation) creates genuine probability split near 50-50.

US initial jobless claims for week ending June 6, 2026 will print below 250,000

Strong NFP print of 172K for May suggests underlying labor market resilience. Previous stories context: claims<240K June 12 (0.67 already predicted for <240K threshold). Below 250K is a more conservative threshold. US labor market has shown consistent strength in 2026; claims above 250K would signal sudden deterioration inconsistent with the trend. Historical base rate for claims <250K in a strong labor market: ~70-75%.

Sun Pharma-Organon merger proxy (PREM14A) filed with SEC by July 31, 2026

Sun Pharma-Organon deal announced ~March 2026. Proxy preparation typically takes 60-90 days post-signing + HSR clearance. July 31 = ~4 months post-signing; standard timeline for initial proxy filing with SEC. Company has been actively preparing shareholder materials. Multiple prior preds at 0.80-0.85 in agreement. Calibration: base 0.70 + 11pp = 0.80.

TSMC May 2026 monthly revenue report (released ~Jun 10) shows revenue > NT$500B

April 2026 was NT$410.73B. Q2 guidance NT$1.4-1.45T implies May+June averaging ~NT$507B each. NT$500B requires +21.7% vs April, consistent with Vera Rubin NVL72 ramp driving N3/N2 demand surge. Cornelius: AI capex = structural floor, TSMC beats analyst consensus during active AI infra transitions. +13pp calibration applied to raw 0.52.

TSMC May 2026 monthly revenue exceeds NT$510 billion, reported by June 12 2026

Q2 guidance NT$1.4-1.45T implies May+June avg ~NT$507B each. April was NT$410.73B. May should step up significantly as AI chip production accelerates (N2/CoWoS ramp). Cornelius: 20-25% uplift from consensus during AI transition; consensus may be ~NT$495B. NT$510B sits just above implied average, consistent with AI demand outperformance. Existing >NT$500B at P=0.65 validates this as incremental probability. +13pp calibration applied.

TSMC May 2026 monthly revenue exceeds NT$520 billion, reported by June 12 2026

Higher bar at NT$520B requires ~2.6% above Q2 implied average of ~NT$507B. With N2/A16 ramp and HBM CoWoS demand firmly committed, May could outperform significantly. However, this is toward the upper end of realistic consensus beats without evidence of specific surge. AI step-function demand creates possibility. Calibration: gap between >NT$510B (0.65) and >NT$540B (0.56) suggests ~0.50 appropriate here. +13pp applied from gut 37%.