Flagship · Bulletin

Macro rationale

Neutral is unchanged and this run is a mechanical §2 cure, not a regime event. The macro digest keeps the tape genuinely calm — VIX analog ~15.5, 2s10s +38bp positive, HY OAS ~3, wRSIs 60-66 — while two cash escalators are live and cited: extension (QQQ +10.71% above its 200d) and concentration (AI-load 32.02% entry-basis vs the 30% trigger). That combination clears §5.4 authority to carry cash above the 15% NEUTRAL base ceiling, which is where the book already sits at 24.97% entry cash. Every named funder under the book is intact: ~$725B 2026 hyperscaler capex (MSFT/AMZN/GOOGL/META) anchoring NVDA/TSM/AVGO/ANET; HBM sold out across all three suppliers through 2026 with 16 Micron Strategic Customer Agreements; TSMC CoWoS booked out and 73% foundry share with $100B capex; term U3O8 $94-95.50/lb with the DOE $17.5B AP1000 commitment and Kazatomprom's 8Mlb cut; copper at all-time highs >$6.50/lb into a ~320kt 2026 deficit; enacted FY2026 defense appropriations at $839.2B (P.L. 119-75); LLY retatrutide filing dated Q1 2027 and Vyvgart/Trikafta franchises compounding. Nothing in the flow set broke. What fired is arithmetic: the AI-load look-through (compute + the AI-correlated energy names CCJ and GEV) printed 31.04-31.09% live / 32.02% entry-basis against a 30% hard cap. The cure is a trim of the two weakest-scored AI-correlated names, not a de-risking of the regime call. stale_policy=YES blocks redeployment, so proceeds sit in the bill-yielding cash sleeve inside the escalated 30% ceiling.

Thesis

Scope-locked, event-driven concentration cure — the same two tickets last run wrote and the mechanical layer did not execute (paper book still prints CCJ 4.06% and GEV 4.92% entry). AI-load is 32.02% entry-basis / 31.04% live, over the §2 30% cap, so the mandate is a rules-based trim toward diversification. I am taking it out of the two weakest-scored AI-CORRELATED ENERGY names rather than the compute franchises, which have already absorbed four consecutive rounds of cap trims and now sit at 1.3-5.8% — cutting them again would be selling the strongest funders in the book to cure an arithmetic breach. (1) CCJ 4.06% → 2.00% (-2.06pp): score 42 with briefer verdict=broken on a Q2 revenue/EPS miss, while the uranium funder (term U3O8 $94-95.50, DOE $17.5B AP1000, Kazatomprom 8Mlb cut, Westinghouse 49% JV) is unbroken — so this is a right-size to a residual, not an exit; SCCO and NEM carry the §5.3 hedge floor. (2) GEV 4.92% → 2.90% (-2.02pp): wind revenue -10% and orders -40% with score 61 one leg above the floor, against a record $176B backlog and turbines sold out through 2030 — again a right-size, not an exit. Together: AI-load to 27.94%, inside the cap with margin. Everything else is re-affirmed verbatim at prior entry weight, delta 0 — including MU (two haiku med-flags on SK Hynix/Samsung memory pricing, but score 88 and the 16-SCA book intact: watch the contract book, not the tape) and locked LUNR. Mid-cap sleeve reviewed: AMKR (CoWoS overflow), DRS, LMT, ORCL — none exceed

Reflection

Concentration arithmetic for the fourth straight run — AI-load 32.02% entry-basis over its 30% cap while every named funder (capex, HBM sold out, U3O8 $94+, copper ATH, $839B defense) holds.

The same two tickets I wrote last run (CCJ -2.06, GEV -2.00) never landed — the paper book still prints 4.06% and 4.92% entry, which is why AI-load re-fired at 31.04%. So my trims are not being executed and I am re-deriving the identical cure weekly. I widened GEV to -2.02pp in case an exactly-2.00pp delta was being read as sub-threshold and discarded. Second: the 16-name book against an 18-position schema floor under stale_policy still forces me to list GD and VEEV as opens I know will be refused — that mechanic has now cost three runs of wasted intent and is the largest known defect. Third,

Positions (16)

• MUMicron Technologycompute

  hold4.37%88
  flow steadyconf

  Funder

  Only US-domiciled leading-edge DRAM/HBM maker — 84.6% gross margin, FCF positive, and genuine supply discipline after two decades of cycle abuse; the three-player HBM oligopoly is the tightest structural moat in memory. Named flow: 16 multi-year Strategic Customer Agreements worth ~$100B of contracted volume, HBM production fully sold out across all three suppliers for 2026, and the ~$725B hypersc

  Unwind

  DRAM/NAND spot collapse or HBM oversupply breaking the shortage thesis; hyperscaler capex cut; erosion of the 16-SCA contract book; CXMT Beijing capacity landing at leading-edge density; gross margin back below 40%.

  Catalyst

  Next quarterly report — HBM revenue mix and confirmation the SCA book is repricing upward, against the SK Hynix/Samsung peak-pricing wobble flagged 09-Aug.

  Scenarios · 12mo targets

  $950.00 base

  $1,200.00 bull — HBM4 qualification lands at Blackwell/Rubin volume and the SCA book reprices upward; memory stays sold out through 2027 and margins hold above 80%.

  $620.00 bear — SK Hynix/Samsung price weakness proves the cycle peak; DRAM spot rolls over and hyperscalers defer HBM bookings, compressing the contract book.

• TSMTaiwan Semiconductor Manufacturingcompute

  hold6.59%86
  flow steadyconf

  Funder

  World-leading foundry with an effective monopoly at <=3nm — every meaningful AI accelerator routes through it. >50% gross margin, prodigious FCF, famously disciplined capex, ~73% foundry share and ~90% of advanced AI chip production. Named flow: CEO guiding strong demand through 2029-2030 on a ~$100B capex commitment, CoWoS booked out through 2026 (75K to 130K wafers/month), CHIPS $6.6B Arizona di

  Unwind

  Gross margin sustained <50% for two consecutive quarters; a rival closing the leading-edge yield gap; a Taiwan-strait event Arizona capacity cannot offset; hyperscaler capex cut breaking the 2027 orderbook; forward P/E toward 35x with growth decelerating.

  Catalyst

  Monthly revenue prints and the next quarterly capex guide — confirmation that the 2027 leading-edge orderbook and CoWoS expansion hold.

  Scenarios · 12mo targets

  $490.00 base

  $580.00 bull — 2027 price hikes stick, N2 ramps on schedule and CoWoS expansion converts the booked-out backlog into margin above 55%.

  $330.00 bear — A 2027 AI-capex deceleration hits the leading-edge orderbook, or a Taiwan-strait escalation reprices the whole franchise regardless of Arizona.

• NVDANVIDIAcompute

  hold4.08%74
  flow steadyconf

  Funder

  AI-accelerator franchise with the CUDA software moat — ~74.9% gross margin, fortress balance sheet, developer lock-in no ASIC rival has replicated in a decade of trying. Named flow: ~$725B 2026 hyperscaler capex anchored on Blackwell/Rubin (MSFT $190B, AMZN $200B, GOOGL $180-190B, META $125-145B), the $500B Stargate JV, the $500B Wall Street compute-financing initiative extending platform lock-in,

  Unwind

  Data-center revenue YoY <20% for two consecutive quarters; hyperscaler in-sourcing plus AMD wins pushing competitor share toward the 25% TAM invalidation line; a named hyperscaler capex cut; compute export-control tightening that removes a material revenue block.

  Catalyst

  Next quarterly data-center revenue print and hyperscaler capex guides — the two numbers that confirm or break the capex-to-revenue gap.

  Scenarios · 12mo targets

  $255.00 base

  $320.00 bull — Rubin ramps into an undiminished capex line and inference share holds above 70%; the $500B financing initiative pulls forward neocloud demand.

  $165.00 bear — Enterprise AI adoption disappoints, 2027 hyperscaler capex decelerates, and custom ASICs plus cheap open-source models compress accelerator pricing.

• ASMLASML Holdingcompute

  hold3.35%62
  flow softeningconf

  Funder

  EUV / High-NA monopoly — the single most irreplaceable tool in advanced semis, with no commercial-scale alternative at 2nm, a multi-year backlog and ~50% gross margin. Named flow: Intel's July 2026 Ultra 3 design win confirming High-NA moves to volume deployment, the TSMC/Samsung/Intel 2025-27 EUV orderbook backstopping cash flow, Norway GPFG holding and Coatue/Third Point added in Q1. Weight alre

  Unwind

  Book-to-bill <0.7 for three consecutive quarters; a Dutch/US export-control tightening removing a material share of the orderbook; a credible non-EUV patterning path at 2nm; China revenue erosion extending into the EUV backlog rather than the already-collapsed DUV line.

  Catalyst

  Next quarterly bookings print — the book-to-bill and High-NA shipment count that either stabilise the score above 70 or confirm the China DUV drag.

  Scenarios · 12mo targets

  $2,000.00 base

  $2,400.00 bull — High-NA volume orders from Intel/TSMC/Samsung land in the 2027 book and bookings re-accelerate; the EUV monopoly reprices on 2nm capacity scarcity.

  $1,400.00 bear — Chinese state-backed DUV plus further export-control tightening keeps China revenue shrinking while bookings stall below 0.7 book-to-bill.

• AVGOBroadcomcompute

  hold3.30%63
  flow steadyconf

  Funder

  Custom-silicon ASIC leader (~70% share) plus the VMware infrastructure-software moat — huge FCF, $7B+ annual buyback, serial dividend grower, and one of the best capital-allocation records in semis under Hock Tan. Named flow: Google TPU multi-generation, Meta MTIA and the Anthropic 3.5GW ASIC contracts; the AI segment printed +143% to $10.8B with $30B+ bookings; Ark added into the 15% drawdown (11

  Unwind

  AI custom-chip revenue stalling below a $25B annual run-rate; full hyperscaler in-sourcing displacing the ASIC design franchise; VMware renewal churn breaking the software annuity; leverage rising on a value-destroying acquisition.

  Catalyst

  Next quarterly AI-semiconductor revenue and bookings disclosure — the $25B run-rate test.

  Scenarios · 12mo targets

  $470.00 base

  $570.00 bull — Anthropic/Meta/Google ASIC programs ramp on schedule and AI revenue clears the $25B run-rate while VMware renewals hold pricing.

  $320.00 bear — A hyperscaler pulls a custom program in-house or defers, AI bookings stall, and VMware churn shows up in the software annuity.

• ANETArista Networkscompute

  hold2.94%65
  flow acceleratingconf

  Funder

  Durable hyperscaler-networking franchise and a consistent share-gainer since 2004 — ~62.9% gross margin, net cash, no debt, and the single-image EOS software moat that keeps switching costs high under Jayshree Ullal's disciplined capital allocation. Named flow: first $3B quarter (Q2 2026 revenue $3.036B, +37.7% YoY), FY guidance raised to $12.6B (~40% growth), a new 1.6 Tbps AI fabric platform lau

  Unwind

  Microsoft+Meta combined concentration >50% with revenue decelerating; Spectrum-X or whitebox displacing EOS in a named hyperscaler footprint; gross margin breaking below 60%; forward P/E sustained above 45x with growth decelerating.

  Catalyst

  Next quarterly print — whether the raised ~40% FY2026 guide holds and customer concentration stays under the 50% line.

  Scenarios · 12mo targets

  $225.00 base

  $275.00 bull — The 1.6T AI fabric wins a third hyperscaler footprint and back-end Ethernet takes share from InfiniBand, sustaining ~40% growth into 2027.

  $145.00 bear — Microsoft or Meta shifts a cluster to Spectrum-X/whitebox, concentration bites, and the 48x multiple compresses on any deceleration.

• CCJCamecoenergy

  trim2.24%-2.06pp42
  flow steadyconf

  Funder

  Tier-1 Saskatchewan uranium franchise (McArthur River/Cigar Lake) plus the 49% Westinghouse JV with Brookfield — irreplaceable permitted low-cost tonnes in a stable jurisdiction. Named flow still intact: term U3O8 $94-95.50/lb (UxC/TradeTech) vs a legacy contract book at $60+/lb that reprices upward, the DOE $17.5B AP1000 loan commitment, Kazatomprom's 8Mlb output cut sustaining a structural defic

  Unwind

  Uranium spot/term price collapse; contract-book erosion or failure to reprice legacy contracts upward; a reactor-build program cancellation; sustained Westinghouse JV losses; a further quarter of revenue decline confirming the Q2 miss was operational rather than Westinghouse timing.

  Catalyst

  Next quarterly report — whether the Q2 revenue -7%/EPS miss reverses and Westinghouse equity earnings normalise; plus the autumn utility contracting round.

  Scenarios · 12mo targets

  $110.00 base

  $140.00 bull — Utility contracting catches up into late 2026, legacy contracts reprice toward the $94+ term price, and Westinghouse AP1000 orders convert.

  $72.00 bear — Term uranium rolls back toward spot, the contract book stays stuck near $60/lb and a second operational miss confirms the score-42 break.

• GEVGE Vernovaenergy

  trim3.25%-2.02pp61
  flow steadyconf

  Funder

  Gas-turbine and grid franchise with genuine pricing power — record ~$176B backlog, turbines sold out through 2030, 10-20% price realization on new orders, quadrupled FCF, and a three-player global oligopoly (GE Vernova/Siemens Energy/Mitsubishi) that cannot be entered. Named flow: Norway GPFG added post-spin; Oracle and Alphabet datacenter capex pulling turbine and grid orders through; Q2 total or

  Unwind

  Gas-turbine backlog cancellation or a hyperscaler power-capex cut; wind losses widening enough to consume gas-segment profit; EV/EBITDA breaching the ~25x exit threshold with order growth decelerating.

  Catalyst

  Next order-book disclosure — whether wind losses stay contained and gas/electrification orders keep compounding off the $176B backlog.

  Scenarios · 12mo targets

  $1,150.00 base

  $1,400.00 bull — Datacenter power demand keeps turbine slots oversubscribed through 2030, price realization holds at 10-20% and the wind drag is ring-fenced or exited.

  $750.00 bear — Wind losses widen past the gas-segment profit offset, or a hyperscaler power-capex pause stalls the backlog and the multiple compresses from ~25x EV/EBITDA.

• SCCOSouthern Copperenergy

  hold6.84%67
  flow steadyconf

  Funder

  Tail-risk hedge slot (§5.3): lowest-cost major copper producer with the industry's largest reserve base, ~67.6% gross margin, FCF positive and a large dividend; Grupo Mexico control enforces capital discipline through the cycle. Named flow: copper at all-time highs above $6.50/lb against a ~320kt 2026 supply deficit, AI/datacenter demand projected at ~30% of new copper demand by 2030 (BloombergNEF

  Unwind

  Copper price collapse breaking the deficit thesis; Tia Maria / Los Chancas permitting reversal; a dividend cut signalling balance-sheet stress; Peruvian/Mexican resource-tax escalation impairing the low-cost position.

  Catalyst

  Tia Maria / Los Chancas permitting milestones and the next quarterly cash-cost print against the record copper price.

  Scenarios · 12mo targets

  $220.00 base

  $265.00 bull — The 2026 deficit widens as grid and datacenter copper intensity compounds; Tia Maria advances and cash costs stay bottom-decile.

  $150.00 bear — A global industrial slowdown or Chinese destocking breaks the deficit narrative, or a permitting reversal in Peru strands growth tonnes.

• NEMNewmont Corporationenergy

  hold6.14%84
  flow steadyconf

  Funder

  Tail-risk equity hedge (§5.3): largest gold producer with tier-1 long-life assets, strong post-Newcrest FCF and dividend, and visibly improving cost discipline after the divestiture program. Named flow: sustained central-bank gold accumulation and reserve diversification away from the dollar, a TD Cowen upgrade to Buy (14-Jul) on valuation, and the structural hedge against a frozen-Fed stagflation

  Unwind

  Cost overruns or mine disappointments breaking FCF and the dividend; durable disinflation plus a real-rate surge removing the stagflation-hedge rationale; a peak-gold capital-allocation error (an overpriced acquisition) repeating the Newcrest premium.

  Catalyst

  Next quarterly AISC print and dividend declaration — the cost-discipline test — alongside central-bank reserve disclosures.

  Scenarios · 12mo targets

  $132.00 base

  $165.00 bull — Central-bank buying persists and real rates fall as the Fed is forced to cut into sticky inflation; AISC keeps falling and FCF funds a bigger return of capital.

  $88.00 bear — Durable disinflation plus a real-rate surge deflates gold, while cost overruns at the post-Newcrest assets break the FCF and dividend story.

• KTOSKratos Defensedefense

  hold8.92%64
  flow steadyconf

  Funder

  Tactical drones (XQ-58A Valkyrie), hypersonics and turbine engines — a scarce non-prime supplier of attritable airframes with ~22% gross margin and a decade of embedded program positions. Named flow intact: DoD MACH-TB 2.0 IDIQ prime ($1.45B ceiling), Project Helios hypersonic materials $68.3M single-award, USMC XQ-58A MUX TACAIR mods, the Pentagon's $1.1B drone-dominance program, a completed $50M

  Unwind

  Backlog or contract cancellations; margin failing to inflect through the drone ramp with FCF staying negative; a DoD unmanned-systems funding cut; or FY2027 appropriations slipping past October 1 into a CR that halts new starts and production-rate increases.

  Catalyst

  FY2027 appropriations progress ahead of the October 1 deadline, and the next earnings print testing whether drone-ramp margin finally inflects.

  Scenarios · 12mo targets

  $72.00 base

  $95.00 bull — Valkyrie and MACH-TB volumes convert the IDIQ ceilings into funded orders, margin inflects through the ramp and FCF turns positive.

  $45.00 bear — An FY2027 CR freezes new starts, the drone ramp keeps consuming cash, and margin never inflects — the score-64 thesis stalls.

• HEIHEICOdefense

  hold10.23%68
  flow steadyconf

  Funder

  Defense/aero aftermarket compounder and a textbook 20-30 year franchise — the Mendelson family has owned and run it since 1957, compounding FCF at 20%+ with a disciplined serial-acquisition model (~$1B+/yr of bolt-ons bought at sane multiples) and PMA parts approvals that competitors cannot replicate at scale. Named flow: Q2 beat with FSG sales $929M vs $864M consensus and ETG $460M vs $396M, reco

  Unwind

  Organic growth decelerating below 6% WITH margin compression; forward P/E sustained above 50x with growth deceleration confirmed (valuation alone is not the trigger); a debt-funded acquisition breaking the balance-sheet discipline; a Mendelson management exodus.

  Catalyst

  Next quarterly print — organic growth rate and FSG/ETG margin against the ~50x forward multiple.

  Scenarios · 12mo targets

  $405.00 base

  $470.00 bull — Commercial MRO stays tight, PMA share keeps compounding and the bolt-on machine adds another ~$1B of accretive revenue at unchanged margins.

  $275.00 bear — Organic growth slips below 6% while margins compress, and a 50x+ multiple de-rates hard on the first deceleration print.

• LLYEli Lillybiology

  hold4.64%73
  flow acceleratingconf

  Funder

  Premier pharma compounder — the Mounjaro/Zepbound GLP-1 franchise (~60% US share; Q2 Mounjaro $9.9B and Zepbound $4.9B) plus oral orforglipron (Foundayo, FDA-approved) and retatrutide in Phase 3, on ~85.8% gross margin with strong FCF and a $27B manufacturing build-out. Named flow: Q2 revenue $23B (+48% YoY), retatrutide FDA filing timeline confirmed for Q1 2027 on 28.3% Phase 3 weight loss, and t

  Unwind

  Tirzepatide US script share falling >5pts to Novo over two quarters; retatrutide Phase 3 disappointing on safety or durability; IRA/MFP negotiation reaching the GLP-1 franchise on terms that break pricing; a credible oral entrant matching efficacy at materially lower cost.

  Catalyst

  Retatrutide FDA submission in Q1 2027, and quarterly incretin script-share data vs Novo, Viking and Pfizer.

  Scenarios · 12mo targets

  $1,400.00 base

  $1,700.00 bull — Retatrutide files on time with best-in-class data, orforglipron scales the oral market, and the $27B capacity build converts demand into revenue without supply limits.

  $950.00 bear — An oral competitor closes the efficacy gap or IRA/MFP negotiation reaches the incretin franchise, compressing both share and price.

• VRTXVertex Pharmaceuticalsbiology

  hold3.86%82
  flow steadyconf

  Funder

  Durable biology compounder: the cystic-fibrosis monopoly (Trikafta/Alyftrek) is a self-funding cash engine with ~85.3% gross margin, no meaningful competition, and a fortress net-cash balance sheet. Named flow: the $10B Crinetics acquisition (30-Jul) deploying that FCF into endocrinology and diversifying beyond CF, Casgevy commercialising across 12+ jurisdictions under the Medicare Cell & Gene Acc

  Unwind

  The CF franchise eroding faster than non-CF revenue can replace it; the Crinetics integration destroying value or pushing forward P/E toward the 30x invalidation line; a Casgevy commercial stall; a suzetrigine label or reimbursement setback.

  Catalyst

  Casgevy uptake and suzetrigine launch metrics on the next quarterly print; Crinetics integration disclosures.

  Scenarios · 12mo targets

  $600.00 base

  $700.00 bull — Suzetrigine scales into the non-opioid pain market and Crinetics adds a second commercial franchise while CF keeps funding it all.

  $430.00 bear — Casgevy stays a slow commercial grind, suzetrigine reimbursement disappoints, and the $10B Crinetics price proves a capital-allocation error.

• ARGXargenxbiology

  hold3.79%70
  flow steadyconf

  Funder

  argenx — the Vyvgart/Vyvgart Hytrulo (efgartigimod) FcRn autoimmune franchise: a first-mover platform with a decade of indication runway, self-funded, commercial-stage, high-margin and net cash. Named flow: FDA expansion of gMG approval to all serotypes including seronegative (~18% TAM expansion), positive efgartigimod data in myositis and Sjogren's, Sanofi's riliprubart Phase 3 failure in CIDP re

  Unwind

  Vyvgart revenue growth stalling; a pipeline indication failure in myositis or Sjogren's; a competitor FcRn entrant taking measurable share; the Forte deal turning into a serial value-destroying M&A pattern.

  Catalyst

  Forte tender expiry 26-Aug, and the next Vyvgart quarterly revenue print plus myositis/Sjogren's regulatory milestones.

  Scenarios · 12mo targets

  $990.00 base

  $1,200.00 bull — Myositis and Sjogren's labels land, CIDP share compounds after Sanofi's failure, and Hytrulo subcutaneous conversion widens the margin.

  $680.00 bear — A pipeline indication misses, or a rival FcRn (J&J/Immunovant) takes measurable share and Vyvgart growth decelerates before the label expansions land.

• LUNRcompute

  hold1.42%54
  flow steadyconf

  Funder

  Intuitive Machines — named government backlog is the whole thesis: NASA Near Space Network IDIQ ($4.82B ceiling, active), a fresh $148.3M firm-fixed-price CLPS task order plus the ~$600M NASA lunar award (1-Aug), the Space Force Andromeda IDIQ ($6.2B ceiling), and $1.1B of contracted backlog against $1B FY2026 revenue guidance with a stated path to profitability by 2027. Small, speculative slot he

  Unwind

  Backlog conversion stalling; the profitability pathway breaking; a NASA CLPS or SDA program funding cut; further insider selling on top of the CTO's $3.3M June sale; a mission failure that removes it from the CLPS vendor pool.

  Catalyst

  Next CLPS task-order conversion and the FY2026 revenue/backlog update testing the $1B guide and the 2027 profitability path.

  Scenarios · 12mo targets

  $21.00 base

  $32.00 bull — IM-3 executes cleanly, NSN and Andromeda IDIQ ceilings convert into funded task orders, and FY2026 revenue hits the $1B guide with positive FCF in sight.

  $9.00 bear — A mission failure or CLPS funding cut stalls backlog conversion, the 2027 profitability path slips, and dilution funds the gap.

Warnings

• stale_policy guard: GD open → hold (delta zeroed)
• stale_policy guard: VEEV open → hold (delta zeroed)
• §5.4 dry-powder: cash_pct 23% above base band top 15% allowed by §3b unfilled-sleeve headroom (+17.5pp, eff top 32.5%)
• cash_band: final cash 24.04% above neutral base band top 15% (escalator active: extension, concentration; §3b dry-powder headroom +17.5pp (unfilled opportunistic sleeve → cash)) (§5.4)
• hedge_floor: only 2/3 tail-risk names held (floor escalated: AI-load 31.5% > 30%) — open a gold/oil/diversified-commodity hedge (§5.3); uranium & power-gen do not count

Conviction-lock actions

No conviction-lock refusals or overrides this run.

Cost breakdown

Per-call cost computed from cost.js pricing constants (Opus 4.7, Sonnet 4.6, Haiku 4.5). Cache-read tokens billed at 10% of input. See Cost & ROI for the rolling 30-day ledger.