←所有簡報

DISCIPLES OF CLAUDE

AI × 市場 · 會議簡報
期號
出席Ryan · Charlotte · Ethan · Ashley · Daniel
60秒摘要
  • The room got bigger and the research got real: Ryan brought a 55-page robotics report, Charlotte brought a full CEG thesis with a named risk, and two new members introduced themselves.
  • One genuine disagreement was logged and left standing: Ethan’s counter-thesis on GEV — the binding constraint is the gas pipeline, not the turbine — against Charlotte’s grid value-chain read. It has since met its first arbiter.
  • One factual dispute went unresolved on the call — does China control ~70% or ~90% of rare earths? The desk has since run it to ground: both, at different stages. Answer inside.
  • Ten commitments entered the docket, from a bot feature with a two-week clock to four research assignments.
本場數據
5Present
71Minutes
20+Members
10Calls entered
房間

What the room argued about

01

Robotics: China’s edge is cost twice over

Raised by Ryan · 55-page research report
  • Four segments — industrial arms, warehouse/logistics, surgical, humanoids. Surgical carries the fattest margins, around 60%.
  • The US bottleneck is cost twice: rare-earth inputs (actuators are 40–60% of machine cost) and integration, which can rival the hardware bill. China is cheap on both.
  • Ryan’s closing claim, now on the ledger: whoever cuts actuator manufacturing cost wins the industry — flagged by the desk as needing a threshold and a date before it can be scored.
Room status · Converged — China structurally advantaged; MP named as the US counter-expression
02

CEG: the full thesis, with the risk named

Raised by Charlotte · the deepest single-name work of the night
  • Capacity factor, 24-hour baseload for datacenters, the 20-year Microsoft PPA, and the IRA nuclear PTC acting as a price floor even in a downturn.
  • The named risk — the house style at work — is FERC: regulators worried big tech siphons public supply and raises consumer power prices.
  • Ethan widened it to the fuel cycle: 10–15-year mine permitting means the uranium bid accrues to incumbents — CCJ.
Room status · Agreed on the name — Charlotte takes the regulatory-mandates research assignment
03

The grid disagreement: turbine vs pipeline

Charlotte’s value chain · Ethan’s counter-thesis · left standing on purpose
  • Charlotte’s chain: AI power deficit → baseload (nuclear or gas) → total grid upgrade. GEV and CEG undervalued on that read.
  • Ethan’s counter, sold his own GEV on it: the constraint is getting gas to the turbine, not building the turbine — and if hyperscalers outbid consumers for gas, turbine opex explodes while the fuel-indifferent hardware layer (ETN) just reprices its backlog.
  • Both agreed on where it converges: own the layer that doesn’t care which fuel wins.
Since this session · First arbiter fired — ETN Jul 31: Electrical backlog +43% y/y, EA orders +41%, margins above guide, organic guidance raised. The transformer side printed first.
04

The room doubled

Ashley · Daniel — first full introductions
  • Ashley: index-fund base (VOO), here to learn single-name work.
  • Daniel: tech-concentrated book, takes the SaaS research seat.
  • Standing division of labor proposed by Ryan and adopted in spirit: each member owns a domain — robotics (Ryan), energy/China (Charlotte), software (Daniel), options mechanics (Ethan, next session).
持倉

The watchlist

Supply side · long7
VSTIndependent power producer; nuclear and gas fleet sits directly under datacenter load growthEthan未說明時限
CEGNuclear baseload — the cleanest expression of firm, dispatchable powerEthan · Charlotte未說明時限
ETNElectrical equipment and grid buildout; picks and shovels for the wallEthan未說明時限
EPDMidstream gas; gas is the marginal fuel for new generationEthan未說明時限
EXEGas production, one step further upstream of the same constraintEthan未說明時限
NBISLow-cost neocloud; the efficiency side of the trade. Not adding hereEthan · Charlotte未說明時限
GEVTurbines and grid equipment; carried in from prior sessionsEthan未說明時限
Demand side · short4
STLALegacy auto caught between energy costs, the EV transition, and consumer creditEthan · puts未說明時限
RCLCruise economics compressed by record fuel costs and debt repricingEthan · puts未說明時限
MUCommodity memory cycle, capex drag, and energy costs landing at onceEthan未說明時限
SNDKSame thesis as MU; storage after a run, without the secular storyEthan未說明時限
Held · separate thesis2
NVDALegacy shares only; trimming remaining mega-cap into strengthEthan未說明時限
GOOGLHolding through late-July earnings; ignoring the near term deliberatelyCharlotte未說明時限
啟示

Takeaway

The teaching moment of the night hid inside an options question. Charlotte floated selling volatility — covered calls — on NXT and ENPH, where implied vol runs hot. Ethan’s rule: a covered call in a bull market converts your best outcome into someone else’s. The premium is real, but it’s payment for the exact right tail the thesis was built to own. The general form, which is the same lesson as the exit-gate rule from last session: know which outcome your position is for before you sell pieces of it. Income strategies belong on names whose upside you’d cap anyway.

未決問題

尚未定論的問題

Does China control ~70% or ~90% of rare earths? — SETTLED SINCE
Ryan’s report said ~90; Ethan said ~70. The desk ran it down: both are right about different stages — roughly 69% of mining, 90–91% of refining and separation, 90–94% of magnet manufacturing (USGS 2026, IEA 2026). The dispute was a units problem, and it’s exactly why claims now enter the ledger with the quantity named.
GEV: undervalued grid winner, or turbine business with a fuel problem?
Charlotte holds the value-chain read; Ethan holds the pipeline-bottleneck counter. ETN’s print supports the fuel-indifferent layer, but that grades ETN, not GEV. GEV’s own next print is the scored test.
Is short-term SaaS investable, or first out the door in a downturn?
Ethan: money rotates out of software first; long-term constructive on NOW only. Daniel takes the research seat. Not converged.
Where does the wheel belong in this book?
Deferred whole to next session’s walkthrough — with the covered-call caution from THE LESSON as its opening constraint.
Emergency rate cut before the election?
Ethan speculates yes eventually, but no falling-rate environment before mid-2027 — now a ledger claim. Three FOMC dissents for a hike since the session lean his way.
日程

值得關注的日期

JUL 30 ✓June PCEHeadline 4.1→3.7, core 3.4→3.3 — the desk’s 70% May-peak call hit; the relief came from a truce that has since collapsed.
JUL 31 ✓ETN · CCJ, pre-marketETN confirmed the transformer thesis; CCJ mixed — realized prices up, volumes and Westinghouse down.
AUG 6CEG earningsCharlotte’s thesis meets its arbiter.
AUG 26July PCEThe sterner test of May-was-the-peak, with energy re-accelerating.
TONIGHT AUG 2The Wheel — session lessonCovered calls + cash-secured puts, mechanics and when not to use them.
跟進事項

誰負責什麼

Ethan · —Bot user-tagging in leaderboard + summariesCommitted with a two-week clock — due around Aug 9.
Ethan · —Wheel-strategy walkthroughCovered calls + cash-secured puts, mechanics and when not to — next session’s segment.
Ryan · —Leaderboard scaffold · robotics slide deck · best-robotics-company name to SlackDesign structure for the redesigned leaderboard; deck consolidating the 55-pager.
Charlotte · —Nuclear regulatory mandates · space thesis · ICE analysis · China robotics articlesThe regulatory work feeds directly into the CEG arbiter on Aug 6.
Daniel · —SaaS / software sector researchNOW, CRM, ORCL, SNOW — the rotation-risk question from UNSETTLED.
Ethan · 未定截止日期VST · State a horizonThe room named this position without a timeframe.
Ethan · Charlotte · 未定截止日期CEG · State a horizonThe room named this position without a timeframe.
Ethan · 未定截止日期ETN · State a horizonThe room named this position without a timeframe.
Ethan · 未定截止日期EPD · State a horizonThe room named this position without a timeframe.
Ethan · 未定截止日期EXE · State a horizonThe room named this position without a timeframe.
Ethan · Charlotte · 未定截止日期NBIS · State a horizonThe room named this position without a timeframe.
Ethan · 未定截止日期GEV · State a horizonThe room named this position without a timeframe.
Ethan · puts · 未定截止日期STLA · State a horizonThe room named this position without a timeframe.
Ethan · puts · 未定截止日期RCL · State a horizonThe room named this position without a timeframe.
Ethan · 未定截止日期MU · State a horizonThe room named this position without a timeframe.
Ethan · 未定截止日期SNDK · State a horizonThe room named this position without a timeframe.
Ethan · 未定截止日期NVDA · State a horizonThe room named this position without a timeframe.
Charlotte · 未定截止日期GOOGL · State a horizonThe room named this position without a timeframe.
本週市場史
JULY 27, 2015

A rally built on state buying lost 8.5% the day the buyers went quiet

On July 27, 2015, the Shanghai Composite fell 8.5% to 3,725.56, its worst day since February 2007, with 75 stocks falling for each one that rose. The index had gained about 150% in a year, then lost 32% in 18 sessions from a June peak above 5,100. Beijing suspended new listings and organized share purchases with central bank cash while at one point about half of all listed stocks sat frozen in trading halts, and the index bounced 16% from July 8. That Monday the state buying was hard to see, and PetroChina fell a record 9.6%.

A price held up by a buyer of last resort is only as firm as that buyer's appetite on the next bad day.
— C. LEBRUN