brain/
← all contexts
project

stock-market

Stock-market dispatch — 2026-06-09

Auto-generated by the daily research routine. Summary of what changed in the wiki today; the wiki is the truth. Research output, not advice — brain never trades.

Recovery note. The automated 05:00 ET dispatch ran the ingest steps (0.5–4a, 4e) and committed nothing — it stalled at the step 4b price fetch (the headless run yielded its turn to a background wait-loop, which never resumes in claude -p, so steps 4b→7 never executed). This dispatch is the hand-completed finish of that run: quotes refreshed, CEG re-dated, feed authored. The runner prompt was hardened so the quote fetch blocks synchronously next time.

Top of mind

CEG's headline overhang partially reversed itself — the one material thesis change today. Yesterday cut CEG 0.70→0.65 on the FERC RM26-4 "likely-NOPR + behind-the-meter (BTM) co-location blocked" read. Today's research (2026-06-09-autoresearch-ceg-pjm-co-location-front-of-meter-shift) is a framing reversal: CEO Joe Dominguez has pivoted Constellation to front-of-the-meter"we do not need behind-the-meter for compelling pricing" — with the Microsoft/Three Mile Island PPA (restart 2027) as the proof point. So the BTM block is a drag on pure-BTM plays (the rejected AWS–Talen structure) more than on CEG's actual chain; the PJM co-location track (Docket EL25-20) is a distinct, further-along proceeding from RM26-4, whose end-June instrument is now read as genuinely undetermined, not "likely a NOPR." CEG conviction partially restored 0.65→0.68 (held below 0.70 — RM26-4 still unresolved, front-of-meter insulation only partial). Otherwise a podcast-heavy ingest day (5 transcripts): the strongest non-CEG signal is Nikesh Arora (PANW CEO) giving the seat-based-SaaS-disruption thesis its first named $100B-software-CEO corroboration — "analytical SaaS is dead... five people become one."

Strongest-conviction buys

Ranked shortlist from step 4c — chains, not advice. Prices 2026-06-09 (twelvedata).

#TickerCausal chain (mechanism)ConvictionValuation vs baseFundamentals (one line)Next catalyst
1LINras-laffan-halt-to-lin-helium-pricing-powerHigh (0.78)Marginal — $511, near 52w high $521High-margin industrial gas, net-cash, dividend grower — not a value trapHelium contract repricing Q2–Q3; APD Q3 (late July) read-through
2CCJai-capex-to-power-and-materials-cascade + kazatomprom-supply-cut-to-western-uranium-premiumMedium-high (0.72)Yes — $106, −21% from hiQ1 adj EBITDA $509M (+78%), Westinghouse stake; uranium-price beta is the riskUtility contracting at $90/lb term; India deal 2027
3MUhelium-cliff-to-hbm-supply-crunch + hbm-cowos-as-binding-bottleneckMedium-high (0.68)Yes — $969 vs 52w hi $1,089HBM sold out through 2027; memory-cyclical is the watchQ3 FY2026 earnings Jun 24
4CEGpjm-capacity-prices-to-nuclear-premiumMedium-high (0.68 ↑)Dislocated — $254, near 52w low, below entry window$11–12 EPS affirmed, 2027 TMI restart, FCF→$11.5–13B by 2028–29PJM EL25-20 / framework submission; Calpine lockup Jun 30; RM26-4 end-June
5TSMhbm-cowos-as-binding-bottleneckMedium-high (0.62)Marginal — $432 vs hi $450CoWoS/HBM the binding bottleneck downstream of foundry; 50%+ GM, backlogCoWoS capacity adds; HBM4 ramp

Watchlist

TickerThesisConvictionPriceUndervalued vs base?Next catalyst
LINras-laffan-halt-to-lin-helium-pricing-powerHigh$511.11MarginalContract repricing Q2–Q3
CCJkazatomprom-supply-cut-to-western-uranium-premiumMedium-high$106.16YesTerm price $90/lb; contracting restart
MUhelium-cliff-to-hbm-supply-crunchMedium-high$969.12YesQ3 earnings Jun 24
CEGpjm-capacity-prices-to-nuclear-premiumMedium-high ↑$253.75Dislocated — see belowEL25-20 / framework; Calpine lockup Jun 30
CFhormuz-nitrogen-supply-shock-to-cf-risk-premiumMedium$109.55YesH2 nitrogen season; Q2 earnings
PWRpwr-transformer-moat-to-eps-doublingMedium$707.26MarginalFERC end-June
MPdrone-warfare-demand-to-mp-hree-rerateMedium$58.21Yes (−42% from hi)HREE circuit commissioning Q2
NOWagentic-ai-seat-erosion-to-saas-rerateMedium$111.49Yes — −2.4% on the dayQ2 FY2026 earnings
INTCtaiwan-chokepoint-to-allied-reshoringExit review$111.80No — above ~$89 consensusVLSI gate

Undervalued candidates (today)

  • CEG at $253.75 — still dislocated near its 52w low and below the entry window, but the reason has weakened: today's front-of-meter pivot mitigates the BTM/RM26-4 overhang that drove the 06-08 cut. The structural chain (3 confirmed steps, $11–12 EPS, 2027 TMI restart) is intact; this is now a "buy the regulatory-overhang dislocation, watch EL25-20 / the framework submission" call rather than the broken-catalyst read it looked like yesterday.
  • MU at $969 — HBM sold-out-through-2027 intact; firming (+1.7%) into the Jun 24 earnings gate.

New chains to investigate (hypothesis-stage)

  • beef-squeeze-poultry-substitution-margin-expansion — PPC, TSN. Cattle herd 75-yr low + beef +13% → chicken substitution while grain glut cuts feed cost → poultry double-margin tailwind. Drafted by prospect-chains (forcing function confirmed; PPC-margin leg the un-cited gap). Thin vertical (consumer/ag). → /explore-chain.
  • drug-pricing-reform-pbm-spread-compression — CVS, CI, UNH (risk side), AMZN (disruptor). 2026 drug-pricing policy stack → PBM spread compression. Healthcare bucket; PBM EBIT exposure not yet sized; slow drip, not a dated catalyst.

New theses (now active)

  • None today — the day's net-new chains landed as hypothesis-stage questions (above), not graduated theses.

Updated theses (existing active)

  • pjm-capacity-prices-to-nuclear-premium (CEG) — front-of-meter pivot mitigates the BTM/RM26-4 overhang; conviction restored 0.65→0.68; signal re-dated.
  • agentic-ai-seat-erosion-to-saas-rerate (NOW) — Arora (PANW CEO) adds independent operator-CEO corroboration of Steps 1 & 4 ("take away UI... five people become one"); conviction held 0.50 (Steps 2–3 still partial); signal re-dated + source folded in.
  • seat-based-saas-ai-disruption — upgraded from single-secondary-source to a named $100B+ software CEO ("analytical = dead / data-infra = winner / system-of-work = re-engineer" taxonomy); panw added as a potential winner on the AI/cyber wave.
  • copper-supercycle-ai-data-centers — added the TC/RC-benchmark-→$0/tonne miner-pricing-power leg (FCX floor-cap breaking, SCCO integrated-smelting edge), Grasberg/Kamoa disruptions, copper off its June-2 record.
  • mega-issuance-supply-wave — DataTrek's Rabe quantified the June-2 6-sigma tech-vs-market outperformance; Colas framed the SpaceX IPO first-day pop as a 1997-vs-1999 regime marker.
  • fomc-private-credit-outflows-alt-managers — now two-sided: retail-ization AUM tailwind (401k EO → ~$4T by 2030) vs. the FSB May-6 first-full-cycle-test risk (barbell: own fee-pure OWL, watch balance-sheet-levered APO).

Contradicted / weakened

  • None today. CEG moved the other way (overhang mitigated, not deepened). No /calibrate.

Open questions worth a human's eye

  • Does CEG's PJM EL25-20 proceeding / framework submission become the clean unlock now that front-of-meter routes around the BTM block — and does the RM26-4 end-June action resolve as a final rule or a NOPR?
  • beef-squeeze-poultry-substitution-margin-expansion — confirm PPC poultry-margin exposure (flock-expansion self-defeat + avian-flu risks) before graduating.

What I looked at

  • Run status: recovery completion of the stalled 05:00 ET automated dispatch (steps 0.5–4a, 4e ran headless and were left staged-uncommitted; steps 4b→7 + commit completed by hand this morning). Root cause: headless turn-yield on the paced quote fetch; runner prompt hardened.
  • Steps 0.5–0.7 (ingestion veins): podcast-ingest 5 transcripts (All-In/Arora, Compound/Colas+Rabe, Moonshots [zero-novelty], Invest Like the Best/Sacerdote, + prior); earnings/feed no-op.
  • Steps 1/2/2b: refreshes on copper (TC/RC) and CEG (front-of-meter); healthcare + financials bucket scans (drug-pricing, private-credit); prospect-chains drafted 1 (beef-squeeze), reported 3.
  • Breadth check (2a): ai-infrastructure still ~57% (⚠ over); today's net-new chains landed in thin clusters — consumer/ag (beef), healthcare (drug-pricing) — correct steering.
  • Sources promoted: 8; sources ingested: 8 → 5 new entities (Arora, PANW, Colas, Rabe, Sacerdote), 0 new mechanisms, 2 new hypothesis questions, 6 pages updated.
  • Valuation snapshots refreshed: 35 tickers (twelvedata); concept snapshot updated for nuclear-baseload (CEG/CCJ).
  • Live-trading feedback (4e): digest pulled (performance-2026-06-09.json); no high-conviction mechanism flagged for calibration this run.
  • Paper ledger (4d): weekly cadence — last full run 06-04, next ~06-11. Not rewritten today.
  • Calibration events: 0. Conviction buys ranked: 5. New hypothesis chains opened: 2.
  • Signal feed: 44 signals (full active/armed set, all valid); re-dated CEG (0.68) + NOW (0.50, new source). Synced to Supabase on completion.