Angie

Thesis watch · Week of November 8, 2026

Long durable scarcity. Short abundance.

Sorry, Angie is not public yet. The thesis is open to everyone while the rest of Angie gets ready. Sign in with Google to join the discussion.

Update · Thu, Nov 19

Anthropic's $81 billion IPO, and why "cheap" Nvidia is not the same as Nvidia

Anthropic began trading today on Nasdaq under the ticker ANT. It sold 980 million shares at $83, above its first range of $70 to $78, and raised about $81.3 billion. That is the largest IPO on record, at a value of roughly $1.4 trillion to $1.5 trillion depending on the share count used. Gray-market trading had pointed to a first trade as much as 27% above the offer price.

The first trade came at 11:48 a.m. ET at $111.50, 34% above the $83 offer and above what the gray market expected. It touched $114.99 in the first minutes and was about $111.90 just before noon. That is the boom part of my call. The hold and the bust are still ahead.

What the filing shows. Anthropic reported $17 billion of revenue in the third quarter and $9.9 billion of computing cost, so about 58 cents of every revenue dollar goes to compute. Government work is under 2% of revenue. Reuters, reporting on the prospectus, counted at least $518 billion of computing commitments over about ten years, most of it payable whether or not Anthropic uses it.

The tie to Nvidia. Alongside the IPO, Anthropic is selling about $26 billion of stock privately to its suppliers and partners: Nvidia and Google about $10 billion each, AMD $2.5 billion, and Broadcom, Micron and Microsoft about $1 billion each, as reported from the amended filing. So yes, the IPO helps Nvidia directly: its biggest customers get cash to keep buying. But read the circle. Suppliers put money in, and much of it comes back as chip orders. And Anthropic does not run only on Nvidia. It also runs on Amazon's Trainium chips and Google's TPUs. The IPO money is spread across several chip makers.

My read on the IPO, on the record before the first trade: boom, hold, bust, then hold at a fair price, the same path I think SpaceX took. People who could not get into Anthropic privately get their chance today, and that demand can carry the first days. The real test is whether revenue keeps outrunning computing costs. I also think the company that wins government work wins this race, the way Palantir built its business on governments and enterprises instead of consumers. Today that business is under 2% of Anthropic's revenue. I will score the call at 1, 5, 20 and 60 trading days.

Now Nvidia. People are calling it cheap at under 10 times earnings. That is true on analysts' estimates for next year. On what it earns right now, it is about 14 times: $152 a share against about $10.72 of yearly earnings at last quarter's pace. Either way, it is far below where it traded this summer.

The level I watch is $148 to $150. On November 12, the day Huang resigned, Nvidia touched $148.80 in the first minute of trading and closed at $166.19, up almost 12% from that low. This week it bottomed at $150.33 on Tuesday. Today it rose to $158.13 at the open and gave most of it back, to about $152 by noon, still above the zone. Buyers keep showing up in that zone. If it breaks below $148 and stays there, that tells me the cheap price is not cheap enough, and that the market is repricing Nvidia without its founder.

My read: cheap does not mean it is worth what it was. Jensen Huang was part of the price. If the market decides the founder premium is gone, the low multiple is the new normal, not a bargain. And if Nvidia, the leader, is priced like this, other chip names may come down to match it. The prices fit the idea so far. From the November 12 close to noon today, Nvidia fell 8.6%, Micron 9.6%, Oracle 9.7%, TSMC 10.7% and AMD 15.0%. Broadcom held up best, down 1.4%. That is consistent with my read, not proof of it.

The scorecard from Tuesday. I said a beat "won't mean anything." Wednesday, Nvidia closed down 2.9% at $150.96 while the chip index finished flat, up 0.1%. That call was a hit. The other side of the record, that proof the business is unchanged could spark a relief rally, missed on day one.

So far the IPO has not lifted Nvidia. It opened up 2.2% and was up only 0.7% at noon, after Anthropic began trading. AMD was down 6.2%, while the chip index was up 1.4%. One possible reason, my read and not proven: some money is moving out of chip stocks and into Anthropic itself. What I watch: ANT's first trade and close, whether chip stocks hold their gains into the close, and the Fed. Governor Waller said this week that a half-point hike in December should be on the table.

Research and opinion, not investment advice.

ANT · NVDA · AMD · AVGO · GOOGL · AMZN

Earlier updates (9)

Update · Tue, Nov 17

Nvidia beat. Here is my scorecard, misses included.

Nvidia reported its first quarter since Jensen Huang resigned, and the numbers were not close. Revenue was $115.8 billion, up 103% from a year ago, against about $108.9 billion expected. Adjusted earnings were $2.68 a share against about $2.47. Data center revenue alone was $107.6 billion. For the next quarter, Nvidia guided to $130 billion, plus or minus 3%, against about $124.9 billion expected, and that guide assumes no data center computing revenue from China at all.

I put calls on the record before this report, so here is the scorecard, misses included.

Hit: "They'll beat." This morning, before the release, I said Nvidia would beat and that the sell-off was an overreaction. It beat on revenue by about 6% and on earnings by about 8%.

Hit: the bar. On November 9 I wrote that a beat inside the guidance range would not be enough this time. Nvidia had to clear the top of the range, $110.2 billion, hold a gross margin of at least 74%, and guide higher. It did all three: $115.8 billion, a 74.9% margin and a $130 billion guide.

Hit, with a caveat: "Huang leaving won't hurt the numbers." True, and the forward guide held too. To be fair to the other side, the quarter ended in late October, before he left. The real test of life after Huang is the next two quarters, not this one.

Open, leaning against me: on November 12 I said Nvidia might beat but "it won't mean anything," because the market would be pricing the leadership question. After the release the stock rose about 2.6% after hours, to about $159.50 from a $155.46 close. A beat this size did move the stock, though modestly. Wednesday's close decides it.

Miss: on November 12 I also said the board would stay quiet until the 18th. It did not. The same day, Nvidia announced Huang's resignation and named Colette Kress interim CEO.

Partial: I said Nvidia would fall hard into the 18th. From the November 12 close to Tuesday's close it fell 6.5%. But the chip index fell 8.2% over the same days. Nvidia fell with its sector, not because of its own news.

What the report means for the thesis. Two lines in it matter more than the beat. First, the guide excludes China data center computing completely, after the new rule on remote access to US chips. The demand outside China was enough to raise the guide anyway. Second, the gross margin guide drops to 71% to 72% from 74.9%. Nvidia is still growing fast, but each dollar of it is getting more expensive to earn, and that is exactly where the cost of money and the cost of the build-out show up first.

What comes next. The open questions are the permanent CEO pick, the new board committee on national security and export controls, and the margin path. I said on the record that the day Nvidia names Huang's permanent replacement will not be a good day for the stock. That one is still open, and I will score it when it happens.

Research and opinion, not investment advice.

NVDA

Update · Tue, Nov 17

Monday: the bounce is gone, and the filings show who borrowed too much

Last Monday the S&P 500 jumped 2.3% and carried the whole week. This Monday took it back. The S&P 500 fell 1.9% to 7,416.59, the Nasdaq 2.5% and the Dow 1.4%. The index now sits just 0.5% above where it closed on November 6.

The damage started in chips. The Philadelphia semiconductor index fell 5.3%. Micron fell 7.6%, AMD 5.6%, Nvidia 4.1% to $156.14 and Broadcom 3.9%. Tech as a sector lost 3.2%. Nvidia is now about 16% below its November 6 close, one day before it reports.

Washington added a reason. The Commerce Department's export office released a rule that reaches remote access to advanced US chips through data centers in third countries. Until now, a restricted buyer could rent the computing power without the chip ever crossing a border. The rule closes that route. How much revenue it touches depends on licenses that have not been decided yet.

Our defensive call held for a second session. Consumer staples rose 0.3% and utilities 0.2% while the S&P 500 fell 1.9%. Real estate slipped 0.3% and health care 0.5%. When the market sells the builders, it has been keeping the boring names.

The filings tell the same story from the inside. On Monday companies filed 809 current reports with the SEC. We read every one with a high-severity item. Three of them came from companies that borrowed heavily.

Optimum Communications skipped about $102.5 million of interest due Monday on three bond series. It has a 30-day grace period, to about December 16, before that becomes a default. The stock rose 19% to $0.63, which tells you the market already expected trouble and is now trading the restructuring.

Cable One received a New York Stock Exchange notice: both its average market value and its stockholders' equity fell below $50 million. It has 45 days to file a plan. The stock is down 52% in a month, to $2.96. Two heavily indebted cable companies breaking at once is not a coincidence. Cable borrowed heavily against a shrinking base of customers, and with long-term rates near 5%, refinancing that debt keeps getting harder.

The Children's Place, already in bankruptcy, filed for new financing and reported a delisting notice.

Two companies said past financial statements can no longer be relied on. Lincoln National Life Insurance will restate its 2024 and 2025 results and its quarters through June 2026. It says the error is a GAAP-only accounting item that does not change its statutory capital or its parent company's consolidated results. Century Casinos will restate over the accounting for a stake in the real estate under its Reno casino.

Outside AI, the energy picture moved. Oil fell on Monday as Reuters reported Hormuz flows recovering, with WTI down 3.2% to $85.72. Crude is down about 17% in a month, but diesel remains tight, so cheaper crude does not yet mean cheaper shipping. In utilities, Black Hills closed its merger with NorthWestern Energy on November 13, one more sign that gas and power utilities are consolidating to fund the grid build.

The data was mixed. The New York Fed's Empire survey fell from 25.5 to 7.8, still positive, but its prices-paid index sits at 45.9. The 10-year Treasury yield ended near 5.12%. Bitcoin is around $65,000, down about 18% in a month.

Where this leaves the thesis. When money costs this much, the builders that borrow feel it first. Monday showed it twice: in the chip stocks that need the AI build to keep getting financed, and in the cable companies that cannot carry their debt. Our scope is widening. AI stays at the center, but energy and gas, politics and credit are now part of the same story, because they all run through the cost of money.

Nvidia reports on Tuesday, November 17, at about 4:20 p.m. ET. My read is unchanged: a beat may not be enough without a clear answer on leadership and compliance.

Research and opinion, not investment advice.

NVDA · OPTU · CABO · PLCE

Update · Sun, Nov 15

Week close: my Friday call missed. Nvidia reports Tuesday.

First, the scorecard. On Thursday night I called one more rally on Friday to end the week. It did not come. The S&P 500 fell 0.4% to 7,561.67 and the Nasdaq fell 0.9%. Only the Dow eked out a gain of 0.1%. Call that a miss.

The other half of the call held up better. I said I would hunker down rather than chase the "it's cheap now" talk. Friday punished exactly the stocks that had bounced. AMD fell 8.8%, Cisco 6.2% and Broadcom 4.8%, even after strong results from Cisco and Applied Materials the night before. Nvidia slid another 2% to $162.85. That is about 13% below where it closed on November 6.

Step back and the week was still positive for the index. The S&P 500 gained 2.5%, the Dow 2.1% and the Nasdaq 2.3%. Monday did almost all of it. The S&P 500 jumped 2.3% that day, about 94% of its weekly gain. For the Nasdaq and the Dow, Monday's gain was bigger than the whole week's, so Tuesday to Friday added up to a loss. The pain sat in the AI names and in bitcoin, which fell below $70,000 for the first time since August and logged its sixth losing week in a row.

The macro picture did not get easier. October producer prices rose 0.3%, a little cooler than the 0.4% expected, but they are up 6.5% from a year ago. The 10-year Treasury yield traded near 5.13% on Friday. Traders put the odds of a December rate hike near 87% for a quarter point and 13% for a half point. The cost of money is still the story.

The weekend added risk. Bahrain says it intercepted Iranian drones, and Iran's Revolutionary Guard claimed tankers hit mines near Hormuz. Reuters could not verify the mine claims. On Friday the IEA said crude is heading for a surplus, but diesel stays tight, near $180 a barrel in Europe. Watch oil when futures reopen.

Nvidia, the week that decides it. The company reports on Tuesday, November 17, at about 4:20 p.m. ET. It guided to $108 billion of revenue, plus or minus 2%, and a 74% gross margin. It is the first report since Jensen Huang resigned, with Colette Kress as interim CEO. HSBC cut its rating to Hold and its target to $170 from $365. Treasury Secretary Bessent said on Sunday that the board should hire an outside CEO and answer questions about its export-compliance culture within 10 days. That is his opinion, not a finding.

My read for Tuesday is the same as before. A beat may not be enough. The market wants to hear who leads, how the China and compliance questions get answered, and whether the next quarter's guide holds. Strong numbers with a shaky answer on leadership could still sell off.

The rest of the week. Anthropic's IPO is expected to price around November 18 and start trading November 19, per its filing. Dates can change. President Xi and President Trump are expected to meet Thursday after the first trade council talks. SpaceX caught its Starship upper stage on Saturday, which happened after Friday's close.

I am still playing it safe. One call missed, and I will say so when I miss. The thesis does not change: when money costs this much, the builders that borrow feel it first.

Research and opinion, not investment advice.

NVDA · AMD · AVGO · CSCO

Update · Thu, Nov 12

Jensen Huang resigns. Colette Kress is interim CEO.

Jensen Huang resigned this morning as Nvidia's president, CEO and director, effective immediately. He founded the company in 1993 and ran it for 33 years.

Colette Kress, Nvidia's CFO since 2013, is interim CEO and stays CFO. Before Nvidia she spent about 24 years in finance at Texas Instruments, Microsoft and Cisco. Harvey Jones, a former CEO and executive chairman of Synopsys, becomes Nvidia's first-ever board chair. The board is searching for a permanent CEO and will look inside and outside the company.

Nvidia gave no reason, and the release carries no statement from Huang. It came four days after his letter to staff admitting an affair, and one day after the New York Times reported a second relationship with an executive who reports to him. The board's review of the foreign intelligence reports was not mentioned in the release.

Earnings are unchanged: Tuesday, November 17, at about 4:20 p.m. ET, with the call at 5 p.m. ET. The board chair said it expects "operations, customer commitments and product roadmap to continue without interruption."

The stock fell as much as 14% after the announcement, to a low near $149, and dropped below $4 trillion in value. CNBC says it is on pace for its worst day since January 2025. It closed at $173.73 on Wednesday.

Here is how my call held up. I said the pressure would last until the board spoke. That part held: the stock fell before the open, the board spoke, and it is still down. I also said the board would wait for the November 18 briefing. That part was wrong. They moved in four days.

What I think matters now. The things that make Nvidia hard to copy do not leave with one person. The factories, the memory contracts and the software stay. What leaves is the founder who set the strategy, sold to the biggest customers and handled Washington. The next test is November 17: can Kress deliver the quarter and the guide with steady hands?

One more thing. Elon Musk replied on X late Wednesday to a post urging him to hire Huang: "That would be an honor." He named no company or job. His companies are among Nvidia's biggest buyers. SpaceX said in August it would build exclusively on Nvidia's next chips, and Nvidia disclosed a stake in SpaceX. It is a reply on social media, not an offer.

Research and opinion, not investment advice.

NVDA

Update · Thu, Nov 12

Nvidia: a second report, a board meeting, and the wait

Nvidia's board met last night to decide whether to remove Jensen Huang. As of 3:30 a.m. ET, it has not said a word. Nvidia's newsroom shows no new statement since November 10.

Here is what changed on Wednesday. The New York Times reported that Huang also had a relationship with Kimberly Powell, Nvidia's vice president of healthcare, who reports directly to him. She joined the company in 2008. The Times cited five former employees, who describe a video from Nvidia's March conference. The Times says it has not seen the video. Reuters says it could not verify the report. Nvidia and Powell have not commented. Treat it as a report, not a fact.

Then the Wall Street Journal reported that the board was preparing to remove him. Reuters says the board met late Wednesday. Nvidia has 11 directors, and 10 of them are independent. A simple majority can remove the CEO, with or without cause. No successor has been named. Bloomberg reports that an outside firm is helping the board look.

The stock had closed up on Wednesday. It fell more than 8% after hours once the reports landed. Reuters puts the loss near $600 billion of value if it holds. Bernstein cut its rating and halved its target from $400 to $200. Baird cut to Neutral. Morgan Stanley and Bank of America took Nvidia off their top-pick lists.

My call, on the record before it happens. Investors who are late to the news will keep pressure on Nvidia all morning, and that pressure lasts until the board speaks. I also think Nvidia could beat earnings on November 17 and it would not matter much, because the market will be pricing the leadership question instead.

The honest challenge to my own call. I assumed the board would wait for the House briefing on November 18. That is the wrong anchor. November 18 is a date Congress asked for, not a board deadline, and the board has already met. A fast, clear decision with a credible successor could end the uncertainty and lift the stock, even if the news is that Huang leaves. Silence is what hurts most.

When could they speak? Boards usually announce big changes before the open or after the close, not in the middle of a trading day. That is my read, not a rule. Watch 4 p.m. today and before the open on Friday.

What this does to the market. Nvidia was about 8% of the S&P 500 and the Nasdaq-100 in October, a little less now after its fall. Simple math: an 8% drop in Nvidia alone takes roughly 0.6% off either index. Overnight, S&P 500 futures were down about 0.5% and Nasdaq-100 futures about 0.9%, so Nvidia explains most of the gap by itself. Bitcoin was down about 0.7%. Bitcoin has no Nvidia exposure. It moves with the risk mood, so a bad day for big tech tends to show up there too.

One more thing on Congress. Rep. Brian Mast said he is prepared to issue a subpoena if Huang did not answer the briefing request by Wednesday. No answer from Huang has been reported.

What to watch: a board statement, most likely outside market hours. Nvidia earnings on Tuesday, November 17, at about 4:20 p.m. ET. The requested House briefing on November 18.

Research and opinion, not investment advice.

NVDA · CRWV

Update · Tue, Nov 10

Midday update: hot prices, Nvidia's second leg down, Anthropic sets a price

Yesterday the market had its best day since April. This morning one inflation report took a piece of it back, and Nvidia fell again.

Prices ran hot. October consumer prices rose 0.4% against 0.3% expected. Core prices, without food and energy, rose 0.3% against 0.2%. The 10-year Treasury yield climbed about 8 basis points to near 4.95%. Traders now put the odds of a December rate hike above 80%, and the odds of a half-point hike went from zero to about 15%.

The Fed had warned us. Governor Waller said this morning, before the report, that a 0.3% core reading means "a 50 basis point hike in the policy rate should be considered." In his words, "Hiking only 25 basis points would put our credibility at risk."

This is the part of the AI story that people skip. The Fed does not have to touch a single chip to slow the build. Higher rates raise the cost of the debt that pays for data centers. Companies that pay with their own cash feel it least. Companies that borrow to build feel it most.

Nvidia, day two. The stock was down about 5% late this morning, near $169, after a 4% drop on Monday. The new piece is a report from The Information: at least $7.5 billion of Nvidia hardware reached Chinese companies through a network linked to Alice Huang since 2023, per two people familiar with the matter. This is a report, not a finding. Nvidia has said the Megaspeed diversion claims were investigated and debunked last year. A shareholder class action was also filed in San Jose.

Here is my honest read. On Sunday I said a drop on the headline alone looks like a sale price, and that the view dies on real evidence of a leak or an export failure. This report is aimed at exactly that second question. It is not proven, so my view stands for now. But the stock is no longer trading on gossip. It is trading on an export question, and that one can cost real money.

Anthropic set its price. It plans to sell 825 million shares at $70 to $78, raising up to $64.35 billion. At the top of the range, the fully diluted value is about $1.4 trillion, up from $965 billion in May. Seven companies, including Nvidia, AMD, Broadcom, Google, Microsoft, Micron and TSMC, plan to buy $26 billion more at the IPO price. Pricing is expected November 18, with trading November 19 under ANT on the Nasdaq. Our read: this is funded demand. The cash pays compute bills, and much of it flows back to the suppliers who are buying in.

Physical bottlenecks keep shipping. AAON's data center cooling sales rose 218% to $397.2 million last quarter, and the stock was up more than 20% this morning. One caution: AAON also restated its first-half results.

At 11:40 a.m. ET the S&P 500 was down about 0.75%, the Nasdaq about 1% and the Russell 2000 about 1.3%.

Tonight: CoreWeave reports after the close. Its own guide is $3.45 billion to $3.60 billion of revenue, with interest costs of $860 million to $940 million for the quarter. In a week like this, the interest line matters more than the sales line.

Research and opinion, not investment advice.

NVDA · CRWV · AAON

Update · Fri, Nov 6

Stocks bounced. The bill for the AI build went up anyway.

Stocks bounced after Thursday's crash, but the bounce is fading. By about noon ET the Nasdaq was up about 1.5% at 24,679, the S&P 500 up about 0.8% at 7,413 and the Dow flat at 51,539. Chips led the early move. The semiconductor index rose about 5% in the first hour after its 12% fall on Thursday, and Nvidia was up about 6.5% before 10 a.m. By noon Nvidia was up about 3.9%. The VIX, the market's fear gauge, fell about 16% to about 23.6. Trading was light, so this is not yet proof that big buyers are back. The Nasdaq is still about 10.6% below its October 6 high, so it is still in a correction.

The bigger news came at 8:30 a.m. The U.S. added 165,000 jobs in October. Forecasts were near 65,000. August and September were also revised up by a combined 84,000. Unemployment was 4.1%, and average pay rose 0.4% in the month and 3.1% over the year. That is good news for workers. For markets it raises the chance that the Federal Reserve raises interest rates again in December. Traders put those odds at about 70% to 78% this morning, depending on the source and the time. The 10-year Treasury yield sat near 4.86% to 4.88%.

Here is why that matters for the thesis. The AI build runs on borrowed money. Data centers, power deals and chip orders are paid for years before they earn anything. When rates rise, every one of those projects costs more, and a late project costs more again. On Thursday the market priced the risk of slower permits. Today it priced the risk of dearer money. A Barclays credit strategist put the new shape of the race well: scarce power "favors the biggest balance sheets." Local pushback makes the build more expensive, not impossible. The companies that can pay cash win. The ones that need loans pay more.

IREN is the test case, and it rose about 19% today. It said on Thursday evening that its AI cloud revenue reached $200 million in the quarter, up from $71 million the quarter before, with a $20.3 billion backlog of signed contracts. Its target of more than $4 billion in yearly AI revenue is contracted and funded, but the start moved to January and possibly February. The open question is the next layer. Of about $15 billion it still plans to spend this fiscal year, about $9 billion is in hand or committed. About $6 billion is still being arranged. One new five-year contract has the customer prepaying about 90% of the chip cost, which is the best defence against dear money. Real revenue earned the bounce. The $6 billion is the next test, and a rate hike makes that test harder.

The AI IPO got cheaper too. Anthropic aims to list as soon as the week of November 16. Investors now talk about a value between $1 trillion and $1.3 trillion, according to Bloomberg and The Wall Street Journal as reported by Yahoo Finance. Last month buyers talked about $2 trillion. Anthropic's filing shows $33.2 billion of revenue for the nine months to September 30. Nothing in that filing changed this week. The price people will pay for it did. That is the same lesson AMD taught on Thursday: a price is a guess about the future, and the guess just got more careful.

Firefly signed a rocket-engine deal. Firefly says AST SpaceMobile will pay more than $1.3 billion, in cash and AST shares, for 36 engines, a licence to build them and three years of engineering help. Firefly rose about 30%. It is a sign that launch engines are scarce outside SpaceX. Shares are part of the price, so the final value depends on AST's stock. SpaceX itself was up about 7%.

Cloudflare beat. Revenue rose 39% to $782.5 million, and the stock rose about 11%. It still reports a loss under standard accounting rules.

DraftKings fell about 27%. Revenue grew 16%, but it will spend $650 million to $700 million this year building a prediction-markets business, and it expects almost no revenue from it this year.

Shoppers feel a little better. The University of Michigan's early November sentiment reading rose to 51.0 from 48.0. Still, 57% of people said high prices are hurting them. The survey ended before the election result.

What to watch: the close, the Fed's next words on rates, Anthropic's final price, Firefly's call on November 10, and whether IREN closes its $6 billion.

Research and opinion, not investment advice.

NVDA · AMD · MU · SPCX · TSLA · IREN · CRWV · FLY · ASTS · NET · DKNG

Update · Thu, Nov 5

Chips did not break. The build plan did. Permits are the new chips, and late fees are the new risk.

Yesterday's election rally is gone. By late morning the Nasdaq was down about 5%, the S&P 500 about 3.4% and the Dow about 1,270 points, or 2.4%. The Philadelphia semiconductor index, the main chip-stock benchmark, fell about 11%. The VIX, the market's fear gauge, rose about 55%. On Wednesday the Dow had closed up 1,569 points, its best day since April 2025.

The selling hit the AI build hardest. Nvidia fell about 8%, AMD about 13% and Micron about 16%. Sandisk fell about 17%, so the memory trade lost most of its recent gains. The companies that rent out AI computing power fell further: in the morning CoreWeave was down about 9%, Nebius about 13% and IREN about 11%, and IREN was later down about 20%. The selloff reached Asia too, where SK hynix fell about 9% and Samsung Electronics about 7% in Seoul.

Here is what we know and what we do not. The fact is a large, fast repricing. No data-center contract was cancelled today, no chip order was withdrawn and no new law passed. The cause is the market's reading of the election. With Democrats now running both chambers, and Texas's new governor backing a pause on new data centers, investors now doubt how fast the AI build can move. One analyst, Dan Ives of Wedbush, estimated that 800 to 1,000 U.S. data centers could stall over the next two years. That is one analyst's estimate, not a count of real projects. Part of the chip drop has a separate cause: Qualcomm fell about 14% after a weaker smartphone forecast, and that has nothing to do with AI.

Tesla and SpaceX belong in this story, but only in part. SpaceX was down about 10% and Tesla about 9% to 10% (SpaceX about 7.6% at 9:56 a.m. ET). The main reason is political. Democrats have said they want to look hard at Elon Musk's role as a federal contractor, and SpaceX depends on government launch and satellite contracts. A planned Tesla and SpaceX merger also now looks harder to approve. The second reason connects to our thesis: SpaceX's AI ambitions need data centers, and those now face the same question about permits and power as everyone else. Musk also spent heavily in the Texas races, and his side lost. So the Musk drop is politics first and the data-center question second.

That data-center question is the thesis. We look for things the world needs and cannot quickly make more of. First the scarce input was chips. Then it was power. Today the market priced the next one: permission to build. A chip that cannot be plugged in earns nothing. A site with power connected and permits in hand can be leased for a premium, as Riot showed on Wednesday with a 15-year lease of its powered Texas campus worth about $28 billion over the initial term.

When a building is late, someone pays. There are three parties: the tenant (an AI lab or a giant like Microsoft), the builder or landlord (a neocloud or developer), and the lender. The tenants have the power. According to The Information, as reported by several outlets in October, tenants now ask for more than $5 million a day in late-delivery penalties on new AI data-center leases, up from about $1 million in the spring. Insurers have started to sell delay policies to cover those fees. A neocloud is a company like CoreWeave or Nebius that rents out AI computing power and borrows to build it.

The big tenants also protect themselves. In September, Oracle sent a force majeure notice on its New Mexico campus. A force majeure clause excuses a party for delays caused by events it does not control. Oracle's notice lets it pay lower, pre-opening rent for longer if power arrives late. Oracle says the project is on schedule, and its landlord says the financial commitments are unchanged. Google now backs some of Anthropic's lease payments at TeraWulf's Kentucky site, but the support starts only when each data hall is finished. So even with a strong partner, the builder still carries the cost of being late.

This is the split we flagged last week. A hyperscaler like Amazon or Google pays for data centers from its own cash flow. If one site is late, it waits or builds elsewhere. A neocloud borrows against each site opening on time. Its interest is due whether the building is ready or not. A delay is an inconvenience for the first and a financing problem for the second.

What would prove us wrong: projects keep their permits, sites open on time, and the neoclouds keep raising money at today's cost. If that happens, today was a panic, not a repricing. We will watch Texas permit decisions, utility connection approvals, the next neocloud debt raise and whether any signed lease is delayed or cut.

Research and opinion, not investment advicw.

NVDA · AMD · MU · SNDK · QCOM · INTC · TSLA · SPCX · CRWV · NBIS · IREN · WULF

Update · Wed, Nov 4

Congress flipped and stocks rallied. The bottleneck for AI just moved to the permit office.

Democrats won both the House and the Senate, according to AP projections. The official count is still under way. President Trump said he would accept the result. By late morning the Dow was up about 1,480 points, or 2.9%. The S&P 500 was up about 1.8% and the Nasdaq about 2%. The VIX, the market's fear gauge, fell about 17%.

The simple reason is gridlock. When one party no longer controls Washington, fewer big laws pass, and markets like fewer surprises. Removing the risk of a disputed result helped too. One strategist put it plainly: earnings matter more than elections. That is the right frame for the rest of the year.

The more important story is underneath the index. The rally was not even. The biggest cash-generating tech companies jumped. Apple and Microsoft each rose about 6% to records, Amazon about 7% and Alphabet about 5%. The companies that rent out AI computing power on borrowed money fell. CoreWeave was down about 8%, and Nebius and IREN also fell. Nvidia slipped.

The reason is Texas. Democrat Gina Hinojosa won the governor's race and has pledged a pause on new data-center construction. Texas holds roughly a quarter of America's planned data-center capacity. A pledge is not a law, and nothing has been enacted. But investors now have to price a new risk: that building a data center depends on a permit, not just on money and chips.

This is the split we flagged last week. A hyperscaler is a company like Amazon or Google that runs its own giant data centers. It pays for them from its own cash flow and has sites around the world. If Texas slows down, it builds somewhere else. A neocloud like CoreWeave rents out computing power, borrows to build, and needs each site to come online on schedule to pay its lenders. A delay is an inconvenience for the first and a financing problem for the second. Credit markets already see it: the cost of insuring AI borrowers' debt has risen as much as 80% since August.

For the thesis, this adds a new kind of scarcity. We look for things the world needs and cannot quickly make more of. Chips were scarce, then power. Now a data-center site with power connected and permits in hand may be the scarcest input of all. Owners of ready sites gain pricing power. Builders who still need approvals carry more risk.

The rest of the market reads the same way. Solar stocks jumped, with First Solar up about 10%, on hopes of friendlier energy policy. Hospital operators and health insurers rose. Defense stocks fell. The 10-year Treasury yield eased to about 4.90%.

What would prove us wrong: Texas projects keep their permits, construction stays on schedule, and the neoclouds keep raising money at today's cost. If that happens, today's selloff was an overreaction, not a repricing. We will watch Texas permit rulings, utility connection approvals and the next neocloud debt raise.

Research and opinion, not investment advice.

AMZN · GOOGL · MSFT · AAPL · CRWV · NVDA · FSLR · AMD · CVS · FIS · XYZ

The house thesis

We look for durable scarcity with demand that someone can keep paying for, and with cash that actually reaches shareholders, at a price that does not already assume perfection.

This week the market told us which part of that sentence it cares about most right now: demand that someone can keep paying for. AI demand did not weaken. The price of the money that funds it went up. Our directional call for the next few weeks follows from that. Funded capacity beats borrowed capacity, owners of physical bottlenecks keep their pricing power, and companies whose plans depend on cheap new debt carry the most risk while rates and credit spreads move against them.

Isiofia

Keep reading. It is free.

Sign in with Google to read the full letter every Sunday, join the discussion, and get each new letter by email if you want it.

Research by Isiofia for education. Not investment advice. Scores rank where research attention goes. They do not predict returns.

IsiofiaIsiofia