My commodity covered the full spectrum: HBM, NAND, SSD, DDR4, DDR5 (which had just hit the market), and even DDR6 (already in testing, years ahead of production). My suppliers? They’re the names you see trading today. At 19, I was getting a real Semiconductor 101 education, not from a textbook, but from the supply chain itself. (I was a Finance Major lol).

ChatGPT was going mainstream. NVIDIA was trading at $1,000 pre-split. I had a hunch, even a conviction, that semiconductors and memory would be foundational to the AI boom ahead. But I’d be lying if I said I fully grasped the magnitude of what was coming.

I could hear the music when no one else could. It was beautiful. Fast forward to Q3-Q4 2025: the rest of the world is finally hearing the melody I’d already caught. Today, I’m here to tell you that same song is getting louder. We’re very early to what I’m now calling The Memory Permanence.

Memory was cyclical for years. Now it's permanently structural.
I believe we are only in the 3-4 innings of The Memory Supercycle, this post will share my view on Memory. Enjoy!

The Number Nobody Is Talking About

This might be the only place you'll see this exact computation modeled this week. That's the edge you're paying for.

Let me give you one data point and let it sit for a second.

Read more

I’ve been skeptical of that framing for a while. And this week, Citadel Securities published a piece that puts the institutional weight of one of the most rigorous macro research shops on Wall Street squarely behind the contrarian view.

Subscribe now

The title: The Economics of Intelligence.

The conclusion: AI is a complement to labor, not a substitute — and the data are already proving it.

What Citadel Actually Found

Frank Flight and the Citadel Securities macro team did something most AI commentators haven’t bothered to do: they looked at the actual data.

Their analysis ran NLP-based keyword classification across S&P 500 earnings call transcripts, tracking how corporate executives have framed the relationship between AI and their workforces over time. The results are striking.

AI mentions on S&P 500 earnings calls grew from 20% to 64% of companies — a near-tripling that began accelerating sharply after ChatGPT launched in early 2023. If AI were actually gutting headcount, you’d expect that acceleration in AI adoption to correlate with aggressive labor cutting signals.

It doesn’t.

Labor cutting language — layoffs, severance, hiring freezes — peaked at 21% in Q1 2023 and subsided to 13% by Q4 2025. Hiring and talent language has remained steady. Restructuring language hasn’t budged. The AI adoption wave, in the actual corporate data, has not produced a labor market shock.

But here’s the number that stops me cold:

When executives mention AI alongside their workforce, they frame it as a complement to labor vs. a substitute by a ratio of roughly 8 to 1. 43% of firms call AI a complement. 51% are neutral. Only 5 → five percent, frame AI as a substitute for human labor.

These are not Silicon Valley futurists or ideologically motivated academics. These are CEOs and CFOs on earnings calls, speaking to institutional investors, where credibility is literally priced in. They are telling you, under that scrutiny, that they are using AI to make their existing people more productive — not to replace them.

The Physics Argument: Why Compute Intensity Changes Everything

Citadel’s framework builds on a structural argument that I find compelling from first principles, and one that maps directly onto my own research from earlier this year on AI infrastructure constraints.

Subscribe now

The core thesis has four linked components:

1. Compute demands scale quadratically with task complexity. This is not intuitive, but it’s crucial. The jump from automating a simple task — summarizing an email, generating a code snippet — to automating a complex, multi-step cognitive role is not linear. It requires orders of magnitude more compute. The physical constraints are real: silicon, power, cooling, data center buildout. These don’t scale overnight.

2. Physical scarcity limits generalized displacement. Because complex cognitive work is expensive to automate at scale, the economics favor augmentation over wholesale replacement. It is cheaper and faster to give a skilled worker an AI co-pilot than to build and run the compute stack that could fully replace them.

3. AI therefore integrates into existing workflows as a force multiplier. This is the radiologist story. This is the lawyer who drafts in half the time. This is the analyst who can synthesize ten earnings calls before the market opens. The tool makes the human more productive without eliminating the human’s indispensable judgment, relationships, and accountability.

4. History says this is the pattern — and demand is elastic. Technological revolutions don’t eliminate labor as an input. They alter the task composition of jobs while output expands to fill the new productive capacity. Microsoft Office didn’t replace office workers. It redefined what office workers could do — and the economy created more of them.

As Citadel puts it directly: “Was the advent of Microsoft Office a complement or substitute for office workers? Ex-ante the concern skewed towards substitution, ex-post it appears a clear complement.”

That’s the bet we’re making again, and the early data are on the right side of it.

Jevons Paradox: The Demand Expansion Nobody Is Modeling

There’s a concept embedded in Citadel’s analysis that I think is the most underappreciated macro dynamic in the AI debate: Jevons Paradox.

Named after 19th century economist William Stanley Jevons, who observed that more efficient coal-burning engines actually increased total coal consumption — because efficiency drove adoption and expanded use — the paradox describes how productivity improvements in a resource often increase total demand for that resource rather than reducing it.

Applied to AI and labor: as AI makes workers faster and more productive, the economy doesn’t simply need fewer workers. It expands the scope of what gets done. More tests get run. More clients get served. More software gets built. The demand for skilled human judgment scales with the expansion of output, not against it.

We’re already seeing this in the construction labor market. AI data center buildout is driving a measurable pickup in construction hiring. The infrastructure layer of the AI revolution is physically employing people. This is Jevons working in real time.

Jensen Has Been Saying This for Months — And He Was Right

I want to give credit where it’s due. Jensen Huang has been in front of this thesis longer than anyone in the financial media has taken seriously.

His radiologist example has become something of a signature argument, and for good reason — it is devastating to the displacement narrative.

Ten years ago, Geoffrey Hinton — the so-called “Godfather of AI” — predicted that AI would make radiologists obsolete. His technical forecast was correct: AI is now embedded in nearly every corner of radiology. But the labor outcome was exactly the opposite of what he predicted.

As Huang explained at the Milken Institute this week: “People who believe this misunderstand that the purpose of a job and the task of a job are related but not ultimately the same thing.”

The radiologist’s job is to diagnose disease. Reading scans is a task in service of that purpose. When AI made image analysis faster and more precise, radiologists could read more scans, serve more patients, and generate better economics for hospitals. The response? Hospitals hired more radiologists. The American College of Radiology projects the number of radiologists in America will grow by up to 40% between 2023 and 2055.

Huang has extended this logic broadly. In his recent appearance at the Milken Institute: “AI creates jobs. AI is the United States’ best opportunity to re-industrialize itself.”

And the data is starting to confirm it. Software engineer job postings on Indeed were up 11% year-over-year in early 2026 — in the middle of what was supposed to be an AI-driven tech layoff wave. Huang himself noted that AI has created more than half a million jobs in recent years.

“I’m fairly confident that AI will drive productivity, revenue growth, and therefore more hiring,” he said.

The inflection is happening. We’re watching it in real time.

The Macro Positioning Read

For those of us managing capital, what does this framework actually imply?

Citadel’s analysis makes an argument that I think has real alpha embedded in it: the market has been pricing AI as a labor destroyer, and that consensus is wrong. Sectors like Software have been brutally sold-off the past year; which if you believe in this color — then software presents an amazing opportunity. As evidence accumulates that AI adoption is not producing the employment shock the doomers forecast, several second-order effects follow:

• Wage growth remains stickier than feared → inflationary impulse doesn’t disappear the way bears assumed

• Consumer spending holds up better than recession models embedded in AI-displacement scenarios → consumer staples and discretionary don’t get the demand shock

• AI infrastructure investment remains front-loaded and massive → the real challenge is the payoff structure (decades of productivity gains vs. near-term capex) not the demand destruction narrative

• The big winner remains the infrastructure layer: compute, power, data centers — the physical real estate of intelligence

The Citadel piece explicitly flags what I think is the real risk to the AI trade: not that AI destroys jobs, but that the investment cycle is heavily front-loaded while the productivity payoff accrues over decades. That’s a very different bear case than the labor displacement narrative — and it’s a bull case for the physical infrastructure names I’ve been long. & .

The pie is getting bigger. The question for investors is simply: who has the biggest fork?

Disclaimer: This is for Educational Purposes Only. NFA.

The discount was rational. Korean corporate law genuinely did not protect minority shareholders. Directors owed fiduciary duty to “the company,” which in practice meant the controlling family. Treasury stock was used as a financial engineering tool to entrench control. Mergers happened at unfair ratios. Cross-shareholdings hid value. The market priced this correctly. The Korea Discount was not a mispricing, it was an accurate read on the legal architecture.

What changed in late 2025-2026 is that the legal architecture changed.

The regime change

On July 22, 2025, an amendment to Article 382-3 of the Korean Commercial Code took effect. It explicitly extends director duty of loyalty to shareholders, not solely to the company. The vote was 220 to 272 — not symbolic, statutory. On August 25, 2025, a second amendment expanded the aggregate 3% rule for audit committee elections, renamed outside directors as “independent directors,” and mandated hybrid AGMs from January 1, 2027. The FSCMA enforcement decree, effective December 31, 2024 — prohibited treasury stock allocation in mergers and spinoffs, closing the most abused governance loophole in modern Korean capital markets.

The Corporate Value Up program, launched February 2024, has now crossed 590 disclosure filings. KRX announced on April 15, 2026 that the simplified disclosure path will be eliminated in 2027 — full plans with status diagnosis, target setting, implementation timeline, and communication strategy will be required. KRX will publish a “low-PBR list” in H2 2026 designed to name and shame companies that have not engaged.

Korean activists now hold real war chests. Align Partners runs ₩1.235 trillion in AUM and has filed AGM proposals at six small-caps in the last 12 months. Dalton Investments has accumulated 5%+ stakes across multiple cosmetics, holdco, and mature-industrial names. Truston has campaigned at BYC since 2021. KCGI executed the first ownership change at Hanyang Securities in 68 years.

These are not promises. They are enacted laws, court precedents, and active capital deployments. The Korea Discount existed because the legal architecture justified it. The legal architecture is now changing faster than institutional pricing has adjusted.

That gap is the trade.

What I’m doing with personal capital

I’m currently building a 15-name basket of Korean equities across three theses: AI infrastructure suppliers selling into the NVIDIA, Micron, SK Hynix, Samsung and TSMC supply chains; deep-value small caps trading below tangible book with named institutional activists on the register; and founder-succession setups where inheritance tax of 50-60% on controlling shares structurally forces value-realizing events.

I’ll be deploying capital into these positions over the coming weeks, building the most liquid names first and accumulating the smaller positions over 4-8 weeks. This is personal capital with a 5-10 year horizon, not a short term trade. Position sizes and entry timing will be a function of liquidity constraints, not narrative urgency.

I’ll be transparent about which names I purchase and at what levels. Not all 15 will be sized equally. Some I’ll own forever. Some I’ll own for 24 months around a specific catalyst. The basket is calibrated to a regime change, not a single thesis.

Tier 1: AI infrastructure (the picks-and-shovels under NVIDIA and TSMC)

Read more

These seven charts, read together, paint a coherent and unsettling macro picture. This isn’t a soft patch — this is a regime transition with compounding signals across correlations, supply chains, capex, and FX. Here’s the breakdown:

Subscribe now

1. AI Dispersion (Chart 1) — The AI Trade Has Bifurcated

The 6-month pairwise correlation across /// has collapsed to ~0.20–0.35, the lowest since at least 2013. This is one of the most important structural signals in the chart deck. The “AI basket” trade is dead as a monolith. The market is now doing fundamental differentiation — pricing MSFT’s Azure growth differently from GOOGL’s search disruption risk, etc.

Implication: You can no longer be long “AI exposure” as a factor. You have to be right on the specific company. This also means dispersion vol is the trade — long individual names via options rather than basket hedges. The single-stock picker’s market is back.

2. Allied Defense Spending (Chart 2) — The Rearmament Supercycle Is Accelerating

Germany and Japan are the standout stories — near-vertical ramps into 2025–2026. Germany crossed ~$100B (nearly doubling in 3 years), Japan accelerating rapidly off a low base. This is a decade-long structural spend shift, not a cyclical bump.

Implication: Defense primes (, , , BAES.L, Rheinmetall) and European defense ETFs (NATO proxy plays) remain multi-year longs. The market has partially priced this but the duration of the spend is underappreciated. This is 10-year capex commitments, not 2-year. My personal favorite way to play the European Rearmament is (Euro Defense ETF).

3. Europe Jet Fuel Supply Shift (Chart 3) — Energy Supply Chain Rewiring in Real Time

European jet fuel/kerosene imports show Middle East supply cratering in April (barely ~100K bbl/day vs. 300K+ earlier) with North America filling the gap at ~200K bbl/day. This is almost certainly Iran sanctions/Hormuz disruption driven. Another win for the US and the Trump Admin.

Implication: The Hormuz choke-point risk is real and already showing in physical flows. This is not priced in airline operating costs, and it’s bullish for US refined product exporters , and LNG shippers. European energy security premium is rising.

4. Core Capital Goods Orders (Chart 4) — Capex Surge, But Read It Right

March 2026 core capital goods orders (ex-aircraft, ex-military) surged ~3.5% MoM — the biggest print since COVID. This looks bullish on the surface, but context matters: this likely reflects front-loading ahead of tariffs, not genuine organic demand expansion.

Implication: Don’t read this as a capex cycle inflection. It’s a pull-forward. The second derivative will likely turn negative in Q2–Q3 as the front-loading exhausts and tariff uncertainty paralyzes forward commitments. I’d watch the next 2–3 prints closely. Industrial machinery and equipment names could give back recent gains.

5. Stock-Bond Correlation (Chart 5) — The Most Dangerous Chart in the Deck

The S&P 500 / Treasury return correlation has flipped positive and is now running near +0.15 to +0.20 — a regime that has historically only persisted during inflationary supply shocks (1970s, 2022). The traditional 60/40 hedge is broken. Bonds don’t protect when stocks sell off in an inflation regime.

Implication: This validates my personal “Broken 60/40” thesis and it means the Fed is pinned — cutting into an inflation scare kills the bond rally and doesn’t help equities. Gold, commodities, real assets, and short duration are the only clean hedges. This chart alone justifies a significant portfolio construction shift.

6. Housing Starts (Chart 6) — Positive Surprise, But Tariff Shadow Looms

Total housing starts hit 1.5M SAAR — the best in over a year, with single-family at ~1.0M. This is a genuine beat and signals that builders are getting ahead of expected tariff-driven lumber/materials cost increases.

Implication: Same dynamic as capex orders — likely a pull-forward, not a trend. Lumber futures and homebuilder stocks , , may have a near-term ceiling as this front-loading exhausts. The real question is mortgage rate trajectory — if the stock/bond positive correlation persists, rate relief is not coming, which chokes off sustained housing demand.

7. USD/JPY and BoJ Intervention (Chart 7) — Japan at 160, No Intervention Yet

Dollar-yen is back near 160 in 2026 — the same level that triggered massive BoJ intervention in 2024 (¥5.62T and ¥5.92T interventions). Critically, no intervention has occurred yet at this level despite the yen sitting right at the prior trigger zone.

Implication: Either (a) the BoJ has decided to tolerate more weakness given US tariff/trade dynamics, or (b) intervention is imminent and the risk/reward on USD/JPY is capped at current levels. For a US macro book, this matters because yen weakness = dollar strength, which is a headwind for EM and commodities priced in USD. We should all watch for a coordinated G7 FX statement — that would be the catalyst for a sharp yen snap-back.

Synthesis: The Big Picture

1. AI Dispersion Collapse — Late-cycle differentiation, monolith trade over

2. Defense Ramp — Multi-year structural, not cycle-dependent

3. Jet Fuel Rerouting — Hormuz risk real, energy supply chain fracturing

4. Capex Surge — Front-loading, not organic — second derivative turns

5. Stock/Bond Correlation Positive — Inflationary regime, 60/40 broken

6. Housing Beat — Pull-forward, rate-dependent sustainability

7. USD/JPY Near 160, No Intervention — Intervention risk building, dollar cap forming

Bottom line: It appears we are in a late-cycle inflationary supply shock regime with geopolitical overlays compressing traditional hedges. The plays are: long real assets/energy/defense, short the 60/40 assumption, be selective (not basket) in AI, and watch USD/JPY as the canary for FX intervention risk.

Disclaimer: This is for educational purposes only. NFA.

Two charts came across my feed this morning from Apollo Chief Economist Torsten Slok — and together they tell one of the cleaner second-derivative macro stories I’ve seen in a while. The market hasn’t caught up yet.

Subscribe now

Chart one: median wage growth is bending back up. After cooling from the post-COVID peak, the job-switcher series — historically the leading indicator for broader wage moves — is re-accelerating into early 2026. Chart two: net immigration has collapsed from 2.8 million in 2024 to a CBO-projected 0.4 million in 2025, recovering only modestly to 0.6–0.7 million through 2027.

Put them together and the story is simple: a labor supply shock is tightening the market in exactly the sectors that relied most on immigrant workers. Native-born workers in construction, agriculture, meatpacking, and hospitality are starting to see real wage gains. And those workers vote.

“The market is pricing the generic ballot. The pocketbook voter is pricing their paycheck.”

The Political Transmission Mechanism

Pocketbook voting is one of the most robust findings in political science. When real wages rise visibly, the incumbent party benefits. Reagan ‘84, Clinton ‘96, the 2019 low-income wage surge all confirm it. The question isn’t whether the mechanism exists — it’s whether the wage acceleration is large enough, fast enough, and felt by the right voters to shift the outcome in swing districts.

The Slok data suggests the answer is yes on all three. The job-switcher series is your early signal — it moves first, stayers follow with a 3–6 month lag. We are currently in that lag window. By September and October, when early voting starts, the wage improvement will be visible in paychecks across the states that matter most.

The critical variable the prediction markets are missing: this isn’t a national story. It’s a state-specific labor market story. And some states are dramatically more exposed than others.

The State-by-State Mispricing Map

If you're interested in trading any of these markets, consider using my referral link to sign up for Polymarket → it helps support independent research like this. 🔗

https://polymarket.com/?r=Kairos7

Texas Senate

This is the most egregious single misprice on the board. Texas has one of the highest concentrations of immigrant labor in construction, oil field services, and agriculture in the country. An 86% collapse in net immigration hits Texas working-class voters hardest and most visibly — and the wage uplift in those sectors will be felt first.

Democrats are currently priced at 43 cents to win the Texas Senate race. Democrats have not won a Texas statewide race since 1994. Thirty years. I have this as 25–28 cents fair value at most. That’s a 15–18 cent mispricing on a $183K volume market — liquid enough to size a real position.

Iowa Senate

Iowa is ground zero for this thesis. Meatpacking and agricultural processing dominate the Iowa interior — Tyson, JBS, and similar facilities employed enormous immigrant workforces. With immigration collapsed, native-born wage pressure in those plants is real, immediate, and politically legible to the voters who work there.

GOP is currently priced at 60 cents. I have fair value at 70–72 cents. The market is underpricing rural Iowa’s sensitivity to exactly the labor dynamic the Slok charts describe. Buy the Republican Iowa Senate position.

Alaska Senate — Mary Peltola vs. Dan Sullivan

This is the cleanest individual race mispricing on the board. Peltola is priced at 64 cents — Sullivan at 36 cents — in a state Trump won by 13+ points in 2024. The market is giving significant RCV incumbency premium to Peltola, but Alaska’s fishing and resource extraction industries are not insulated from the labor dynamics in play.

More importantly: Alaska at 64% Democrat in the current political environment is simply wrong on priors alone. I have Sullivan at 45–50 cents fair value. That’s a 9–14 cent edge on the highest-volume individual Senate race on the page ($306K).

Ohio Senate

Ohio is the interesting swing case. Manufacturing and construction workforce with meaningful immigrant labor overlap, in a state Trump won by 11 points in 2024. Democrats are priced at 58 cents here — a slight favorite. If the wage acceleration story plays out in the Ohio Rust Belt, this flips. Medium conviction, 8–10 cent edge toward GOP.

The Full Alpha Stack

What Kills This Trade

Two things break the thesis and you should watch both closely.

First: tariff-driven goods inflation. The April 2026 tariff regime on steel, autos, and broad consumer goods could re-ignite inflation and compress real wage gains back toward zero. Nominal wages up 4% plus goods prices up 3% equals a real gain of roughly 1% — which voters don’t feel. Watch the April CPI print (May release) as a pivot point. If goods prices reaccelerate, the Slok thesis weakens as a political argument.

Second: sector-specific unemployment. The immigration collapse has two sides. Tighter labor supply pushes wages up, but it also slows hiring in construction and agriculture. If regional unemployment spikes in immigrant-dependent sectors — Texas homebuilding stalls, Florida agriculture contracts — the job loss narrative will dominate the wage gain narrative in local media and voter sentiment.

The Data to Watch:

Atlanta Fed Wage Growth Tracker (monthly), BLS JOLTS job openings in construction + leisure/hospitality, generic ballot by state (not national), and April CPI goods ex-food/energy. When the generic ballot in Texas or Iowa tightens toward Dem +2 or better, that's the signal the market is repricing — and your position is moving.

The prediction markets are running a national model on what is fundamentally a disaggregated labor story. The crowd is pricing the generic ballot from three months ago. The edge is in being early on the second derivative — the rate of change in real wages for the workers in the states where the immigration-to-wage transmission is fastest.

The market will be the last to know. That’s where the alpha is.

Intel delivered its sixth consecutive quarter beating expectations, with Q1 2026 revenue of $13.6 billion, up 7% YoY.

The beat was broad-based — revenue came in$1.4 billion above the midpoint of guidance, driven by supply improvements, better mix, and pricing actions.

Subscribe now

The GAAP loss was distorted by a $3.9 billion non-cash goodwill impairment charge on Mobileye, driven by increased discount rates and macro uncertainty — not an operational deterioration. AI-driven businesses now represent 60% of revenue and grew 40% YoY!

Q2 2026 guidance: Revenue of $13.8–$14.8 billion, non-GAAP gross margin of 39.0%, non-GAAP EPS of $0.20. DCAI guided to grow double digits sequentially.

LRMI: Analysis: Data Center & AI Infrastructure Demand

I believe this is where the real signal is. Management was unusually direct about a structural shift in the CPU-to-accelerator compute ratio, and the data center numbers confirm it.

DCAI Segment — The Core Thesis

• DCAI revenue: $5.1 billion, up 22% YoY and 7% sequentially, described as “well above expectations.”

• ASIC revenue nearly doubled YoY and was up >30% sequentially — an underappreciated growth vector.

• Operating profit: $1.5 billion at 31% margin, up ~$292M QoQ on better yields and cycle times on Intel 3.

The CPU-to-GPU Ratio Inflection — The Most Important Data Point on the Call

Management quantified what the market has debated: the ratio of CPUs to GPUs in AI workloads is structurally shifting toward CPUs as AI matures:

• Training: ~7–8 GPUs per 1 CPU

• Inference: now closer to 3–4:1

• Agentic / multi-agent: approaching parity or better for CPUs

CEO Lip-Bu Tan reinforced this: “The backbone of AI computing in production remains a CPU-anchored architecture... CPU is the orchestration layer and critical control plane for the entire AI stack.”

Inference requires CPU-heavy orchestration for managing multi-agent data flows where power-per-performance matters.

Demand Signals Are Getting Stronger, Not Weaker

• Server CPU unit TAM outlook improved over the last 90 days — Intel now expects double-digit unit growth for the industry and for itself, with momentum extending into 2027.

• Supply, not demand, is the binding constraint. Management explicitly stated demand continues to outpace their growing supply.

• To address this, tool spending is increasing ~25% YoY even as total CapEx stays flat — a signal of deliberate capacity acceleration on the manufacturing side.

Customer Wins — Structural, Not Transactional

• Google signed a multi-year long-term supply agreement for Xeon IPU in Q1, with Lip-Bu describing it as a “long-term trusted partnership.” More contracts to be announced.

• NVIDIA’s DGX Rubin NVL8 selected Xeon 6 as the host CPU — Intel is embedded in the most advanced AI training systems being shipped today.

• SambaNova signed a multiyear collaboration for next-gen heterogeneous AI inference architectures combining SambaNova RDUs with Xeon 6.

Advanced Packaging — A Scarce and Strategic Asset

The semiconductor TAM is approaching $1 trillion driven by AI demand. Intel has identified three strategic moats: the x86 CPU franchise, advanced packaging, and the manufacturing network.

Advanced wafer and packaging capacity remains inshort supply industry-wide, which benefits Intel’s vertically integrated supply chain. The company is expanding its back-end facilities in Penang, Malaysia on a multi-year horizon.

Near-Term Gross Margin Watch

CFO Zinsner flagged that Q2 gross margin will step down from 41% to 39% due to Panther Lake (18A) volume ramping ~6–7x QoQ — still below corporate average margins. Input cost inflation (substrates, T-glass, memory) is also a headwind in H2. This is a ramp tax, not a demand problem.

Stanley Druckenmiller Variant Perception

Using Druckenmiller’s framework: find the consensus that is wrong, identify the leading indicator the market is missing, and size the bet on the rate of change, not the level.

Subscribe now

What the Market Has Wrong

The market’s dominant narrative has been “GPU wins AI, CPU is commoditized legacy silicon.” Intel has traded as a restructuring story, not a growth story. The variant perception emerging from this call is that this framing is 3 years stale.

The Asymmetric Setup

1. The CPU TAM is being re-rated in real time. The market modeled GPU/accelerator-centric AI infrastructure. The shift to inference and agentic AI is a second derivative change that dramatically increases CPU intensity per unit of compute output. Intel’s own numbers — DCAI +22% YoY, server CPU TAM now expected up double-digits vs. flat expectations 90 days ago — are the leading indicator. The rate of change in estimates is what Druckenmiller would focus on.

2. Supply is the constraint, not demand. When demand exceeds supply and the company is accelerating tool spend 25% YoY, you have a pricing power and revenue upside setup into H2 and 2027. The market is discounting Intel on margin concerns (18A ramp costs) without properly crediting the demand backlog building underneath.

3. 18A is a free call option the market is not pricing. Yields are running ahead of internal projections — described as a “meaningful inflection.” If 18A yields continue to improve and external foundry customers commit (Intel has flagged that 14A continuation depends on external design wins), Intel goes from a one-asset story (x86 CPUs) to a two-asset story (x86 + foundry), re-rating the stock.

   

4. The ASIC business is hiding in plain sight. ASIC revenue nearly doubled YoY and grew 30%+ sequentially. This is structurally attached to the custom silicon wave (hyperscalers building their own chips). Intel’s packaging and manufacturing capabilities make it a natural beneficiary. The street is not modeling this segment discretely.

5. Consensus is modeling PC weakness as a drag. Full-year PC TAM is expected down low double digits and H2 client is weak. This is the consensus anchor weighing on Intel’s multiple. But CCG is now >60% AI PC mix. As AI PC penetration grows and the platform refreshes, the downside case on CCG may be more muted than feared.

The Risk That Kills the Variant View

The bear case is straightforward: 18A fails to attract sufficient external design wins, making the foundry investment a value destroyer with no revenue to justify it. Management explicitly flagged that 14A development could be paused if external customer demand isn’t secured.

Additionally, input cost inflation (substrates, memory, helium shortages) compressing margins in H2 could cause estimate cuts that overshadow the demand recovery narrative. The bet requiresprocess execution— which Intel has historically struggled to deliver on time.

Bottom line: The earnings call revealed a company where the demand picture is materially better than expected and improving, the CPU structural narrative is gaining real-world confirmation, and the primary risk is self-inflicted execution (18A ramp, costs) rather than market demand. The variant perception is that Intel is being priced as a restructuring story when the data now supports framing it as an AI infrastructure beneficiary in an early innings re-rating.

With CPU plays like , , and surging +40% in just two weeks, it’s clear the market is aggressively correcting a previous mispricing in the space. I anticipate a rotation of profits back into GPU leaders like , whose May 20 earnings (T-24 days) will serve as a definitive litmus test for both GPU demand and the broader AI sector.

Disclaimer: This is for Educational purposes only. Not a solicitation to buy or sell securities. NFA.

• Forward Observations Desk is a field-deployed intelligence function within an investment organization that gathers real-time, on-the-ground insights → through site visits, and supply-chain checks— to inform trading decisions and validate or challenge market narratives ahead of broader consensus.

Today’s piece is brought to you by LRMI’s Forward Observations Desk, Enjoy!

A SemiAnalysis tweet hit X yesterday with two GB300 dev racks CoreWeave scrambled to deploy after their cluster went down. Standard AI infrastructure stuff — until you actually look at the photo.

Read more

Revenue: ₩52.6T +198% YoY · beat est +7%

Op. Margin: 72% record · +30pp YoY

EBITDA: ₩41.3T 79% margin · beat est

Net Cash: ₩35T flipped from net debt

This quarter was not supposed to be possible

To put 72% operating margins in context: Apple, the most profitable consumer hardware company in history, runs at roughly 30%. SK Hynix — a Korean memory manufacturer that analysts have modeled as a commodity cyclical for two decades — just more than doubled that, in Q1, which is historically their softest seasonal quarter.

Revenue hit ₩52.58T, up 60% from the prior quarter and 198% year-over-year. DRAM average selling price surged mid-60% quarter-on-quarter with flat bit growth — meaning the entire revenue jump was pure pricing power, not volume. NAND ASPs were up mid-70% QoQ. This is not a demand-pull inventory cycle. This is structural scarcity pricing.

Four consecutive record operating profit quarters — Q2 '25 through Q1 '26 — with COGS rising only 6% QoQ on 60% revenue growth. That is what operating leverage looks like when a commodity business stops being a commodity.

“Customers’ demand for the next 3 years far exceeds our current supply capacity. Within the limited supply capacity, we are doing our best to supply as much HBM as possible to our customers.”

— Kim Gi-Tae, Head of HBM Sales · SK Hynix Q1 2026 Earnings Call

A three-year demand visibility statement. In memory semiconductors. That has never happened before in this industry's history.

LRMI Derivative Analysis: Where we are in the cycle: momentum, acceleration, jerk

Most earnings recaps will tell you revenue went up. Here's the more interesting question, I’ve developed: is the rate of improvement getting faster or slower? Running the three mathematical derivatives on quarterly revenue gives a precise cycle positioning signal most analysts never compute.

Read more

$5.1 billion in revenue. $66.8 billion in backlog. The fastest cloud in history to $5B. And the smartest money in the world — Jensen Huang, Leopold Aschenbrenner, Jane Street — is still buying.

Rome had aqueducts. The AI era has CoreWeave.

— LRMI, November 2025

Time for my thesis for CoreWeave

Subscribe now

The Smartest Money in the World Just Showed Its Hand

This week, Jane Street — a global quantitative trading firm that handles billions in daily volume and operates with arguably the strictest security and performance requirements of any financial institution — committed approximately $6 billion to use CoreWeave’s AI cloud platform. On top of that, Jane Street made a $1 billion equity investment in CoreWeave Class A common stock at a purchase price of $109.00 per share. Article Link

Let that sink in. The most mathematically rigorous firm in finance didn’t just rent GPUs. They bought a billion dollars of the stock. At $109.

Under the expanded agreement, CoreWeave will provide Jane Street with next-generation compute across multiple facilities, including access to NVIDIA’s Vera Rubin architecture and the full software and services stack required to deploy and scale AI at the frontier of quantitative research. CoreWeave was selected for its ability to combine high-performance compute with an integrated software layer that operates efficiently and consistently under real-world conditions — with dedicated connectivity, custom storage configurations, and responsive technical support tailored to Jane Street’s research operations.

This isn’t a press release you skim. This is a signal you study. Jane Street joins NVIDIA, Meta, Anthropic, and OpenAI in placing multi-billion-dollar, multi-year bets on CoreWeave as the essential infrastructure layer for AI. The contract backlog now includes $21B from Meta (through 2032), $6.3B from NVIDIA (capacity backstop through 2032), a multi-year agreement with Anthropic (announced April 10), and now $6B from Jane Street — over $33 billion in committed or contracted demand from four of the most sophisticated capital allocators on the planet.

The question every investor should be asking: what do they know that the market doesn’t?

The Rating That Started It All

To understand why this caliber of customer keeps choosing CoreWeave, you have to understand the SemiAnalysis ClusterMAX™ Rating System — the closest thing the neocloud industry has to a Moody’s-style credit rating, but for GPU cloud performance.

SemiAnalysis

ClusterMAX™ 2.0: The Industry Standard GPU Cloud Rating System

Introduction…

Read more

6 months ago · 141 likes · 3 comments · Jordan Nanos, Daniel Nishball, Michelle Shen, Cheang Kang Wen, Wei Zhou, Jeremie Eliahou Ontiveros, and Dylan Patel

SemiAnalysis (founded and run by ) is the most respected independent semiconductor and AI infrastructure research shop in the space. Their ClusterMAX system independently tested dozens of GPU clouds covering approximately 90% of the entire GPU rental market by volume, evaluating providers from the perspective of a typical enterprise customer across five tiers: Platinum, Gold, Silver, Bronze, and UnderPerform. ClusterMAX 2.0, released in November 2025, reviewed 84 providers across a market map of 209 total companies, incorporating interviews with over 140 end users.

CoreWeave is the only AI cloud provider to achieve the Platinum rating. Twice. No other provider came close — to reach Platinum, a provider needs to score 90+ out of 100 in nearly every category with almost no weak points. The full tier breakdown tells the competitive story clearly:

• Platinum: CoreWeave (sole provider)

• Gold: Nebius, Oracle, Azure, Crusoe, Fluidstack

• Silver: Google Cloud, AWS, Together.ai, Lambda

• Bronze: DataCrunch, TensorWave, and others

AWS is Silver. Google Cloud is Silver. Azure is Gold. CoreWeave stands alone at the top.

What SemiAnalysis Actually Tested

The benchmarking wasn’t a survey or a questionnaire. SemiAnalysis ran a year-long series of tests spanning single-node setups to massive 1,000+ GPU clusters, evaluating ten criteria including security, storage, orchestration, reliability, networking performance, technical expertise, and pricing.

Some concrete results from ClusterMAX 2.0: on the PyTorch install test (a proxy for WAN connection quality and local disk I/O), CoreWeave completed the install in roughly 3.2 seconds versus a median of 8.5 seconds across all providers — and a staggering 41.2 seconds for one Bronze-tier provider. On the 22.1GB NGC container pull test, CoreWeave completed it in under 10 seconds (they maintain local NGC mirrors), while Gold-tier providers took 45-60 seconds.

But the benchmarks that matter most for the investment thesis are the ones that translate directly into dollars.

The Six Competitive Advantages That Create the Flywheel

1. Goodput & Reliability — The Cash Generation Engine

CoreWeave’s Mission Control platform delivers 50% fewer job interruptions when running GPU clusters of over 1,000 nodes, achieving a goodput rate as high as 96% versus the industry average of 90%.

This is the single most important number in the entire neocloud economy. On a 10,000-GPU training run at $2.35/GPU/hr, that 6-percentage-point delta means roughly $12.3 million in saved compute per year per cluster. CoreWeave operates 250,000+ GPUs — the aggregate efficiency advantage is worth hundreds of millions annually to its customers. This is why they keep coming back, and why they sign multi-year deals worth billions.

2. Pricing Power — The Platinum Premium

SemiAnalysis validated that Platinum-tier providers command a pricing premium over competitors because the total cost of ownership is better, even when the raw $/GPU-hr is higher on paper. CoreWeave is the only cloud to consistently command premium pricing across SemiAnalysis’s interviews with 140+ end users.

This isn’t a vanity award. It’s a validated pricing moat.

3. Storage Innovation — CAIOS & LOTA

CoreWeave’s proprietary AI Object Storage (CAIOS) and Local Object Transfer Accelerator (LOTA) were rated industry-leading by SemiAnalysis. LOTA transparently caches data on compute node local drives, eliminating storage bottlenecks that can represent over 20% of cluster TCO for data-heavy workloads like video generation, weather prediction, drug discovery, and robotics.

4. Orchestration — SUNK & CKS

SemiAnalysis recognized CoreWeave’s Slurm on Kubernetes (SUNK) and CoreWeave Kubernetes Service (CKS) as best-in-class for managing distributed AI workloads. The validation that matters: even Meta and Jane Street — organizations with world-class internal engineering — choose CoreWeave’s managed orchestration rather than building their own.

5. Security — The Enterprise Moat

SemiAnalysis specifically highlighted CoreWeave’s GPU/InfiniBand-specific pentesting, VPC isolation, and real-time threat detection as industry-leading. This matters because enterprises rarely rent from emerging neoclouds — they gravitate toward CoreWeave because it has crossed the security hurdle that blocks every other neocloud from enterprise adoption. Jane Street, with arguably the strictest security requirements of any firm in the world, chose CoreWeave. That’s the ultimate security certification.

6. Customer Entrenchment

SemiAnalysis concluded that CoreWeave is entrusted to manage the large-scale GPU infrastructure for AGI labs like OpenAI, high-frequency trading firms like Jane Street, and even NVIDIA’s internal clusters. Read that again — NVIDIA trusts CoreWeave to run its own compute.

The Internal Pitch: What CoreWeave Showed Jensen

I think the surface-level narrative — “of course he loves them, they buy his chips” is a lazy analysis. NVIDIA’s 13F shows CoreWeave as its largest equity investment by far: over 24 million shares worth $3.2+ billion, representing more than 90% of NVIDIA’s entire reported equity portfolio. NVIDIA increased its stake from 17.9 million shares at IPO to 24.28 million shares during Q2 2025.

Jensen isn’t diversifying. He’s building a vertically integrated AI ecosystem. Here’s what was really on the pitch deck:

The Capacity Backstop — A Zero-Downside Floor

CoreWeave entered a $6.3 billion agreement where NVIDIA is obligated to purchase any unused cloud capacity through April 2032. The mechanic is elegant:

• NVIDIA sells GPUs to CoreWeave → revenue for NVIDIA

• CoreWeave rents them at premium pricing → revenue for CoreWeave

• If demand softens → NVIDIA buys the residual capacity → CoreWeave still generates revenue

• NVIDIA uses that capacity for internal R&D or partner allocations → NVIDIA gets utility

CoreWeave effectively holds a put option on its own inventory, written by the strongest balance sheet in the semiconductor industry. This crushes CoreWeave’s cost of capital — the single biggest variable in neocloud economics — because lenders know NVIDIA is backstopping billions in demand.

Jensen didn’t invest in a customer. He built a demand-guaranteed distribution channel for his own silicon.

The AWS Playbook on GPUs

The pitch wasn’t “we rack and stack GPUs.” It was: “We’ve built a proprietary software stack that makes GPU compute actually work at scale.” Jensen recognized this as the AWS playbook replayed on accelerated compute — the infrastructure layer becomes the moat, not the hardware.

Scale Economics

CEO Mike Intrator said the quiet part out loud: “If we stop growing at the rate we’re growing, this company would become enormously profitable within three months. We’re making the strategic decision to get to scale.”

The cost structure is almost entirely fixed — capex-funded GPUs, datacenter leases, networking. Once you cross the utilization threshold, every marginal dollar of revenue falls to the bottom line at 60-80% incremental margins. Revenue guidance: $5B to $12B+ by end of 2026. The GPU pricing environment validates the model — H100 1-year rental pricing surged almost 40% from $1.70/hr in October 2025 to $2.35/hr by March 2026, and on-demand capacity is completely sold out.

The Variant Perception: What Leopold, Jensen, and Jane Street Know

Leopold Aschenbrenner’s Situational Awareness fund — the $1.5B+ AGI-focused hedge fund backed by the Collison brothers, Daniel Gross, and Nat Friedman — has made CoreWeave one of its highest-conviction positions. In Q4 2025, the fund added 9.4 million shares of CoreWeave call options, a 672% increase in position size. The fund’s total US equity and options portfolio expanded from $254 million in Q4 2024 to over $5.5 billion by Q4 2025, with CoreWeave as a top holding.

Leopold exited NVDA and AVGO in Q4 2025 and rotated aggressively into the physical infrastructure layer — CoreWeave, Core Scientific, Applied Digital, Bloom Energy, Solaris Energy, and Bitcoin miners pivoting to AI hosting. His thesis: the bottleneck in the race to AGI has shifted from chips to power, cooling, and physical data center capacity. CoreWeave is the operator that converts that physical infrastructure into revenue-generating compute.

The variant perceptions that Jensen, Leopold, and now Jane Street are all pricing in:

1. CoreWeave is a compute utility, not a GPU rental company. GPU rental implies commodity, low-margin, replaceable. Compute utility implies recurring revenue, switching costs, pricing power, and operating leverage. The Platinum rating proves it. You don’t call AWS a “server rental company.”

2. The debt is a feature, not a bug. The market sees $13B+ in debt and panics. The smart money sees an asset-backed lending structure where every dollar of debt is secured by physical GPUs with contracted cash flows — backstopped by NVIDIA through 2032. The cost of capital compresses as CoreWeave scales, which is exactly what Intrator described.

3. The GPU shortage is structural, not cyclical. SemiAnalysis identified a massive disconnect between tightening supply dynamics on the ground (40% pricing surge, sold-out on-demand capacity) and bearish public market sentiment still anchored to narratives of eventual oversupply. The reality on the ground points to sustained scarcity and pricing power.

4. The inference transition creates a second S-curve. Most of CoreWeave’s revenue today is training workloads. But inference — running models in production — is a much larger TAM, more recurring, and more latency-sensitive, which means the quality premium that justifies Platinum pricing gets even wider. CoreWeave holds a leading 20% share of the neocloud market and is positioned as a primary beneficiary of this transition.

5. Jane Street’s $1B equity buy at $109 is the ultimate validation. This isn’t a strategic partnership disguised as investment. Jane Street runs quantitative models on everything. They modeled CoreWeave’s unit economics, stress-tested the balance sheet, analyzed competitive positioning, and put a billion dollars of their own capital on the line at $109. When the most mathematically sophisticated firm in finance — a firm that literally makes markets for a living — buys equity alongside Jensen Huang and Leopold Aschenbrenner, the signal-to-noise ratio is deafening.

The Bear Case (and Why It’s Getting Weaker)

Every honest thesis addresses the risks. The bears have three arguments:

“The hyperscalers will catch up.” Maybe. But SemiAnalysis has evaluated them twice, and AWS is Silver, Google is Silver, Azure is Gold. CoreWeave has been Platinum both times. The gap isn’t closing — it’s widening on dimensions like storage (CAIOS/LOTA), orchestration (SUNK/CKS), and security.

“The debt is unsustainable.” It would be, without contracted demand. But $33B+ in committed revenue from Meta, NVIDIA, Jane Street, and Anthropic — plus 96% utilization and rising GPU pricing — transforms the debt from existential risk into leveraged upside. Every dollar of debt is matched by a GPU generating contracted cash flow.

“GPU prices will collapse.” SemiAnalysis data says the opposite is happening right now. H100 rental pricing surged 40% in six months. On-demand is sold out. The shortage is structural, driven by the inference transition and hyperscaler capex acceleration (Goldman projects $432B in hyperscaler spending in 2026).

The Pitch Deck Jensen Saw (Reconstructed)

If I had to guess the five slides that helped CoreWeave close the deal with Jensen Huang:

1. “We run your chips better than anyone in the world.” 96% goodput, Platinum rating, trusted by OpenAI + Jane Street + NVIDIA itself.

2. “The flywheel: your backstop de-risks our capex, which de-risks your channel.” The circular economics of the $6.3B capacity agreement.

3. “We’re building the AWS of AI before anyone else gets to scale.” First-mover on GB200/GB300 deployment, SUNK orchestration, Mission Control.

4. “The unit economics at scale are 60%+ incremental margins.” Operating leverage from fixed-cost infrastructure with rising utilization and pricing.

5. “The TAM is every AI model that will ever run in production.” The inference transition multiplies the addressable market by 10x.

Jensen didn’t invest in a GPU buyer. He invested in the distribution layer that makes his entire ecosystem work. CoreWeave is to NVIDIA what AWS was to Intel in 2006 — except this time, the chip company owns equity in the cloud.

Bottom Line

The market sees a levered, unprofitable GPU rental shop. Myself, Jensen Huang, Leopold Aschenbrenner, and Jane Street see the essential infrastructure utility for the most important technology buildout in human history — with contractually guaranteed demand, pricing power validated by independent testing, a $33B+ contract backlog, and a path to enormous profitability that management says is three months away if they just stopped growing.

They’re not stopping.

Disclaimer: This is not financial advice. Do your own research before purchasing securities

“Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultraintelligent machine could design even better machines; there would then unquestionably be an ‘intelligence explosion,’ and the intelligence of man would be left far behind.”

— Irving John Good, Speculations Concerning the First Ultraintelligent Machine (1965)

I.J. Good wrote that in 1965. He was a British mathematician who worked with Alan Turing at Bletchley Park cracking Nazi codes during World War II. He wasn’t writing science fiction. He was writing a technical paper for Advances in Computers, Volume 6, published by Academic Press.

Sixty years later, Leopold Aschenbrenner — former OpenAI safety researcher, now at one of the most well-capitalized AGI labs in existence — opens Part II of Situational Awareness with that exact quote. Not as a curiosity. As a thesis statement.

His message is clear: we are no longer in the era of speculation.

Read more

The surface narrative: Trump wants peace. Vance flies to close the deal. Markets price de-escalation.

The actual signal: Trump is sending his family office to negotiate — not the State Department, not the NSC, not career diplomats. Rubio is nowhere near this meeting. That is not an accident.

Napoleon at Tilsit, 1807.

He showed up having already won. Alexander thought he was negotiating. Prussia had no seat at the table.

The deal was structured before the meeting started.

Sound familiar?

Subscribe now

Layer 1 — This Is a Property Acquisition, Not a Peace Deal

Trump doesn’t think in geopolitical frameworks. He thinks in assets. What does Iran have that Trump wants?

• The Strait of Hormuz — the most valuable maritime chokepoint on earth

• Iranian oil — currently sanctioned, sitting idle, representing massive latent supply

• Regional stability as a product he can sell to Gulf states who will pay for it

Kushner’s entire Middle East playbook in Term 1 was the Abraham Accords — which were explicitly transactional. Normalize relations, unlock economic flows, take a fee in the form of geopolitical credit. The Gaza deal followed the same structure. This is deal number three in the same franchise.

The variant perception: Trump isn’t ending a war. He’s closing a real estate transaction where the asset is Middle East stability and the buyers are Saudi Arabia, UAE, and the Gulf states who will fund the reconstruction. Kushner has the relationships with every Gulf sovereign wealth fund. That’s not coincidence — that’s the monetization layer.

Layer 2 — Vance Is There Because He’s Running in 2028

This is the piece nobody is saying out loud. Vance doesn’t fly to Islamabad to sit in a room for optics. He goes because he wants the photograph and the headline — VP brokers Iran peace deal — on his résumé before the 2028 primary begins.

Trump knows this. Trump uses this. Vance’s personal political incentive to close is the enforcement mechanism that ensures the delegation doesn’t come home empty-handed. You don’t send someone with a legacy motive unless you want maximum closing pressure in the room.

The variant perception: Trump has aligned every principal’s personal incentive with the outcome he wants. Vance needs a win. Kushner needs a legacy. Witkoff needs the next deal. Iran gets to survive. Trump gets the Nobel narrative. This isn’t diplomacy — it’s incentive architecture.

Layer 3 — Iran Already Knows They Lost

Here’s the read the market is completely missing. Iran published a 10-point proposal. Think about what that actually means strategically.

When you publish your terms in a negotiation, you’ve already surrendered the leverage of ambiguity. You’ve told the other side exactly what you need, which means they know exactly what to give you and what to withhold. Professional negotiators never do this. Iran did it because they needed to signal to their own domestic audience that they extracted terms — not because they had leverage.

The Hormuz closure was their nuclear option. They played it. The US didn’t blink — they escalated harder, hit Kharg Island, and sent the VP. Iran’s only remaining play is to get the best possible terms on the way out.

The variant perception: Iran walked into this negotiation having already lost. The 10-point proposal is a surrender document dressed up as a negotiating position. The question isn’t whether a deal gets done — it’s how much face-saving architecture Witkoff builds around the Iranian capitulation.

Layer 4 — The Pakistan Choice Is a Message to China

This is the deepest layer and the one zero people are talking about.

Why Pakistan? Pakistan is militarily and economically closest to China of any US-adjacent mediator. Pakistan’s powerful military has deep PLA relationships. Choosing Pakistan as the venue is a direct signal to Beijing: we are resolving the Middle East on our terms, through your allies, without you.

Trump is showing China he can stabilize the region that China depends on for energy — without Chinese involvement — right before his planned Beijing visit. That’s not a coincidence. That’s positioning.

The variant perception: The Islamabad meeting is partially a message to Xi. Trump is demonstrating that he controls the energy security architecture of the Middle East and can turn it on or off. That’s leverage for the Beijing summit. Iran is almost a sideshow — the real audience for Saturday’s deal is China.

Layer 5 — The “Fraudulent” Comment Was Deliberate

Remember Trump said the 10-point proposal was “workable” — then hours later called it “fraudulent.” Markets whipsawed. Everyone called it chaos.

It wasn’t chaos. It was a negotiating tactic so old it has a name: good cop / bad cop compressed into one person. By calling the proposal fraudulent after agreeing to the ceasefire, Trump:

1. Signaled to his domestic base that he’s not being soft on Iran

2. Reset Iran’s expectations downward so any final deal looks like Iran “won something back”

3. Created urgency — Iran now needs to prove their proposal isn’t fraudulent by showing up Saturday

The variant perception: Trump manufactured the perception of a collapsing deal specifically to make Iran more motivated to close. Every apparent miscommunication in the last 48 hours was pressure, not incompetence.

The Bottom Line Read

The market thinks this is a VP flying to Pakistan to negotiate peace.

What it actually is: a fully engineered transaction where every principal’s personal incentive is aligned to close, the counterpart has already signaled capitulation through their own proposal, the venue choice sends a parallel message to China, and the apparent chaos of the last 48 hours was deliberate pressure tactics.

Trump is two steps ahead. The meeting in Islamabad is the closing table. The deal was structured before the ceasefire was announced.

That’s the variant perception.

Most traders are watching the headlines. You should be watching the playbook.

We’ve seen this exact team operate before — and the methodology is fully documented.

Subscribe now

The Gaza playbook is the analytical anchor here. From primary sources, the deal team’s exact methodology is on record. Witkoff and Kushner used the intensely personal techniques of real estate deal makers — dangling presidential promises, protections, or punishments to get Israel and Hamas to agree. Witkoff and Kushner outlined that it was crucial to include tangible benefits for both sides, and that Trump’s personal backing made those guarantees credible.

• The critical intelligence: the willingness of Witkoff and Kushner to meet Hamas leaders directly, despite political risks, demonstrated to the group that the US was serious about getting and enforcing a deal.

• And the exact closing move: Egypt’s intelligence chief returned to tell Witkoff and Kushner: “Based on the meeting we just had, we have a deal.”

Now here’s the full strategic model — built exactly as I think Professor Jiang would construct it:

THE THREE-STEP PLAYBOOK. LIVE AGAIN THIS SATURDAY.

Step 1 — Isolate the counterpart. In Gaza, Arab mediators (Qatar, Egypt, Turkey) told Hamas: deal now or lose your global cover. Hamas caved. This weekend, Pakistan plays the same role with Iran. When your only mediator tells you the window is closing, you deal.

Step 2 — Personal credibility deposit. Witkoff opened the Gaza room by offering condolences to the Hamas lead negotiator whose son had just been killed. He turned a terrorist negotiation into a human exchange. That’s not diplomacy school. That’s real estate psychology — read the room, close the gap.

Step 3 — The Trump guarantee as the closing instrument. “You will be treated fairly. He stands behind every point.” Delivered in person. That’s what broke the deadlock. Iran knows this team carries Trump’s word directly — no translation layer, no bureaucratic chain.

THE GAME THEORY IS CLEAN.

1. Iran’s dominant strategy is to deal. They published a 10-point proposal. Parties that intend to walk don’t publish proposals. Holding the Hormuz closure costs them global legitimacy every single day.

2. Kushner’s motive is different from any diplomat you’ve ever seen. He has said publicly he won’t be remembered for business deals — only peace deals. That’s a man with a dominant strategy to close regardless of the optics.

3. Vance is there for one reason: to make it unambiguous that Trump is personally behind whatever gets signed. The VP flying to Islamabad is the credibility signal markets haven’t fully digested yet.

THE PROBABILITY READ:

🟢 Deal reached Saturday — 42%

🟡 Framework agreed, details deferred — 35%

🟠 Talks collapse, resume within 72hrs — 16%

🔴 Hard walkout — 7%

Deal + Framework = 77% chance of a positive resolution by Sunday morning.

The 16% collapse scenario isn’t terminal either — the Gaza precedent shows this team comes back within 72 hours when talks break down. They don’t walk away from the table permanently.

THE MARKET EDGE NOBODY IS TALKING ABOUT.

The announcement, if it comes, lands Saturday evening Islamabad time — early Saturday morning US time. Before markets open Monday. The gap between diplomatic reality and market pricing is the trade.

Oil has already priced some of this. Equities haven’t fully priced the second-order effects — financials reopening, M&A restarting, credit spreads tightening, AI rotation resuming.

The window to own the position is Friday close to Saturday evening.

Not after the announcement. Before it.

Disclaimer: For Educational purposes only. This is not financial advice. Do your own research.

He’s not wrong that confirmation bias is real. But he’s also wrong that the data has nothing to say.

So I built a model to find out for certain.

Subscribe now

The Setup

The 2Y10Y yield curve — the spread between 10-year and 2-year Treasury yields — is the most-watched recession indicator in finance. Every major U.S. recession since 1976 has been preceded by an inversion. The conventional wisdom is that a recession follows 12–18 months after inversion ends.9.7 month median.

But “conventional wisdom” is just vibes dressed up in a suit. I wanted the actual numbers.

I pulled every daily 2Y10Y observation from FRED going back to 1976 and identified 44 distinct inversion episodes. Then I built a three-phase ML system to answer two questions:

1. Does inversion duration actually predict recessions, and if so, what’s the threshold?

2. What additional features push predictive accuracy from “better than random” to “actually useful”?

Phase 1: Duration Alone

The first surprising finding: the optimal duration threshold is 2 days.

That sounds like noise, but it’s actually meaningful. The model isn’t saying that 2-day inversions are dangerous — it’s saying that any sustained inversion carries signal. The threshold is effectively zero. My professor is right that there’s no magic number. The wrong conclusion is that there’s therefore no signal.

The logistic regression trained only on duration and depth features returned an AUC of 0.590 — barely better than a coin flip. My professor would say “I told you so.” But that’s Phase 1.

Phase 2: Adding Macro Context

When I added 14 features across five domains — credit spreads, labor market momentum, Fed policy regime, ISM/industrial data, and curve shape — the model jumped to AUC 0.819.

The LASSO automatically selected 8 of 21 features as actually predictive. The top three by SHAP importance:

• Unemployment distance from cycle low — how far unemployment has risen from its trough when the inversion begins

• Industrial production trend — the 6-month slope of IP at inversion start

• 10Y yield level at inversion start — the absolute rate environment

The single strongest binary signal: whether the 3M10Y spread also inverted (the “double inversion”). When both the 2Y10Y and 3M10Y are simultaneously negative, the model treats it as a high-confidence confirmation. In the 2022–2024 cycle, both inverted — the 3M10Y stayed negative from October 2022 all the way through December 2024.

Phase 3: Live Tracker

The system now runs daily, pulls fresh FRED data, and scores the current macro environment against the trained ensemble. It outputs a 0–100 recession risk score, updated every evening at 6pm ET.

Here’s what it said this morning:

• Recession Risk Score: 37/100 — MODERATE LASSO: 38.2% | RF: 36.7% | Ensemble: 37.4%

• Days since uninversion: 590 (19.4 months)

• Status: BEYOND P90 WINDOW — RISK RECEDING

The historical median lead time from inversion end to recession start is 9.7 months — that window passed in June 2025. The P90 upper bound was January 2026. Both have passed without a recession.

Why 4 Red Flags Are Still Lit

A 37/100 score with a “BEYOND P90” status sounds like an all-clear. It isn’t — and this is the part worth paying attention to.

The model flags three features still in the RED zone from the 2022–2024 episode:

1. Log-duration = 6.7 (783 days). This inversion was the longest in recorded history. Every historical lead-time distribution in my training data was built from episodes averaging far shorter. When the model extrapolates a P90 window, it’s doing so from a sample that has no true analog to a 783-day inversion. The confidence interval on that P90 is wide.

2. HY credit spread at inversion start: 580 bps. High-yield spreads were already elevated when the curve first inverted in July 2022. That’s the credit market pricing in stress before the curve even broke. Historically, elevated HY spreads at inversion onset are associated with faster and deeper downstream damage.

3. Fed was hiking during inversion. The Fed raised rates 525 basis points across this cycle. Hiking while the curve is inverted accelerates the credit transmission mechanism — banks tighten lending standards, refinancing costs spike, and the damage compounds. The full effect of that tightening cycle may not have fully cleared.

The honest read: the base case is no recession. The +52 bps spread, no double inversion, and a BEYOND P90 status all point toward a soft landing holding. But the structural damage from the longest inversion in history — running through an aggressive Fed hiking cycle against a backdrop of already-stressed credit — means tail risk is non-trivially elevated. The tariff shock arriving in April 2026 is exactly the kind of exogenous catalyst that could activate latent damage the model is still flagging.

The Methodology Note

This model was trained on 44 inversion episodes with approximately 8 recession outcomes — a small-N problem I’m not going to hide. AUC 0.819 with LOO cross-validation is the right methodology for that sample size, and the improvement from 0.59 → 0.82 by adding macro features is robust. But wide confidence intervals come with the territory. I’m not publishing a 95% confidence interval because it would span most of the probability space — that’s the honest truth of building a recession model on 50 years of data with infrequent outcomes.

What I can say: the feature selection is economically intuitive, the validation is structurally sound, and the live tracker updates daily from official FRED data. That’s more rigor than most of what you’ll read on financial Twitter.

Disclaimer: Not investment advice. Model trained on historical data; past relationships between yield curve inversions and recessions may not persist.

I was at the dealership with my parents last week. My parents both have 800+ credit scores, as clean as it gets. The salesman rattled off the banks they use for financing: Chase, BofA, and Ally. Chase won the rate, but the name "Ally" stuck with me. Because I know something most car buyers don't.

Ally Financial isn’t a diversified bank. It’s a car loan company that happens to have a banking charter. Roughly 59% of its total assets — about $115 billion — are tied directly to auto lending. The vast majority of its revenue and profit come from one thing: Americans making monthly payments on their cars.

And that got me thinking. In 2008, the biggest liability most Americans held was a mortgage. The second biggest? A car loan. If there’s another recession → a real one, with unemployment spiking and consumers stretched thin — the company most exposed to that second-biggest liability is Ally .

So I went digging. I used Quartr AI to pull every earnings transcript, every slide deck, every management admission from the past three years. What I found was — frankly — remarkable.

Here’s the full forensic breakdown.

The Macro Setup: America's Auto Debt Problem

Before we even touch Ally’s filings, look at what’s happening to American car buyers right now.

30.5% of car buyers trading in a vehicle are underwater — they owe more than the car is worth. That’s up 4.2 percentage points from a year ago (JD Power, March 2026).

The average amount of negative equity hit an all-time record of $7,214. The average monthly payment for these underwater buyers is $916. And 40.7% of them are on 84-month (7-year) loans just to afford the payment.

The average new car now costs $49,353 — up 30% from 2020. Americans owe $1.7 trillion in total auto debt. And many of the cars being traded in underwater were purchased during 2022–2023, when prices were at their absolute peak.

This is the consumer environment Ally Financial operates in. Now let’s look inside.

Next is my research from Quartr:

Red Flag #1: The Asset-Liability Time Bomb

This is the structural flaw at the center of everything. Michael Burry’s key insight in 2005 was that mortgage lenders had locked in long-duration fixed-rate income but funded themselves with short-duration, rate-sensitive liabilities. Ally has the exact same architecture — just in auto! Literally holy smokes!

Ally’s consumer auto portfolio is fixed-rate — loans originated at whatever rate was set, locked in for 5–6 years. But Ally funds itself primarily with floating-rate deposits (high-yield savings, money markets) that reprice immediately when rates move.

Read more

“And Fannie and Freddie are stupidly cheap. Asymmetry at its best. They could be a 10X and it could happen soon.”

2.6 million views in under 24 hours. Then Michael Burry — the “Big Short” guy, @michaeljburry — replied with: “Cannot emphasize enough how rare this is in this market.”

Two of the most credentialed contrarian investors alive, publicly aligning on the same trade, at the same moment. I’ve been building this thesis for months. This is the writeup I’ve been working toward.

Subscribe now

What Fannie Mae and Freddie Mac Actually Are

If you haven’t thought about these companies since 2008, here’s the 60-second version.

Fannie Mae (FNMA) and Freddie Mac (FMCC) are the two government-sponsored enterprises — GSEs — that underpin approximately 70% of the U.S. mortgage market. They don’t originate mortgages. They buy them, package them into mortgage-backed securities, and guarantee them. They are the plumbing of American homeownership.

In 2008, they nearly collapsed under the weight of subprime exposure. The government took them into conservatorship under the FHFA, injected $193 billion, and has controlled them ever since. The original deal: all profits get swept to Treasury. That “net worth sweep” continued for years — by the time it was amended in 2021, Fannie and Freddie had already returned over $300 billion to the U.S. government. They repaid the bailout in full, with a substantial premium on top.

Since 2021, they’ve been allowed to retain earnings. And the capital accumulation that followed is the most important number in this thesis.

The Second Derivative Nobody Is Watching

Fannie Mae’s total equity: $47 million in FY2021. $109 billion by FY2025. That is $95.5 billion of capital accumulation in five years — not from government transfers, but from retained earnings on a $14+ billion annual earnings machine.

Freddie Mac went from $28 billion in equity to $70.4 billion over the same period.

The rate of accumulation matters more than the level. These entities are compounding equity at scale, every quarter, like a clock. Meanwhile, operating costs are falling — Fannie cut noninterest expense by $141 million year-over-year in 2025, including a $40 million reduction in administrative expenses from headcount cuts and contract renegotiations. Freddie’s noninterest expenses fell to $8.6 billion, down $38 million year-over-year.

Revenue up. Costs down. Equity compounding. In any normal corporate context, this trajectory commands a premium multiple. These stocks trade at roughly 0.1x book value on Fannie and 0.04x on Freddie.

Why? Because they’re still in conservatorship — and most institutional investors literally cannot own them.

The Perception Gap: Why Nobody Owns This Yet

This is the core of the asymmetry. Run through who can’t own FNMA/FMCC right now:

1. Dividend and income funds — no dividends under conservatorship. Hard screen out.

2. ESG mandates — government-controlled, governance red flags. Hard screen out.

3. Active growth managers — no CEO, no strategic autonomy, FHFA controls everything. Pass.

4. Style-box investors — not classified as a bank, insurer, or fintech. Doesn’t fit the box.

5. Index funds and quants — OTC-traded, not listed on NYSE or Nasdaq. Not in the index.

6. Risk-managed macro funds — political binary risk. Avoid.

Every single one of these constraints evaporates upon conservatorship exit. A re-listing, a PSPA restructuring, or a permanent CEO appointment would force a mass reclassification of these securities across hundreds of institutional mandates. That forced reclassification is the re-rating mechanism. The buyer base that doesn’t exist today would materialize overnight.

The Capital Shortfall: What It Is and Why It Matters

Freddie Mac stated explicitly on their Q4 2025 earnings call: “Capital shortfall to regulatory capital, excluding buffers, was $106 billion because the $73 billion of senior preferred stock does not qualify as regulatory capital.”

That sentence is doing a lot of work. The shortfall is not an operating insolvency. It’s a structural accounting artifact. The senior preferred stock owed to Treasury doesn’t qualify as regulatory capital under the ERCF framework — so on paper, the capital gap looks enormous. In reality, any PSPA restructuring that converts, cancels, or reclassifies that Treasury preferred could eliminate the shortfall essentially overnight.

Management didn’t bury this disclosure. They surfaced it, explained the mechanism, and named the instrument. That’s not an accident — that’s an invitation for sophisticated investors to model what the re-rating looks like if the preferred gets restructured.

Ackman has done exactly that. His base-case valuation, using 16x and 13x estimated 2026 earnings respectively, gets you to approximately $34/share for both entities at IPO. From FNMA’s current price around $4.86 and FMCC around $4.38, that’s roughly 600–700% upside in his base case. If they revisit pre-2008 levels in the mid-$40 range, you’re looking at roughly 900% upside from current prices. Stocktwits

What Burry Actually Said — And Why It’s More Important Than Ackman’s Tweet

Burry’s amplification was notable — but his full framing is more nuanced than the retweet suggests. According to Seeking Alpha, Burry now sees FNMA and FMCC IPOs as a 2027 proposition at best. Seeking Alpha He’s not calling for a near-term catalyst. He’s calling for the trade to be real, the thesis to be valid, and the timeline to extend — which means the entry window is still open.

This is actually the most constructive possible framing for a new position. The “stupidly cheap” language with a 2027 timeline is Burry saying: the market has overcorrected on timeline pessimism. Both stocks have lost more than half their value over the past six months, reaching their lowest levels in a year GuruFocus — not because the earnings thesis broke, not because privatization became less likely structurally, but because a Bloomberg article reported that a planned 5% secondary offering stalled.

The reaction to that stall is the mispricing. A failed secondary offering isn’t a failed privatization thesis — it’s a scheduling delay.

The Management Signal You’re Not Supposed to Notice

Subscribe now

Management of both GSEs cannot speak openly about conservatorship exit. They are regulated entities under FHFA control. But they can change who they’re speaking to and how.

On Freddie Mac’s Q1 2025 earnings call, CFO Jim Whitlinger named FHFA Director Bill Pulte by name — explicitly, on a public earnings call — and connected regulatory reform directly to revenue growth and net worth accretion. He called it a roadmap for a better U.S. housing finance system. Compare that to Q4 2024, where the same CFO spoke exclusively about financial metrics with zero mention of the transformation agenda.

Three calls in, the messaging cadence is different. That’s not boilerplate drift. That’s positioning language for an investor base that doesn’t exist yet.

At Fannie Mae, a new Acting CEO — Peter Akwaboah — opened his first full-year earnings call by describing his tenure as happening “at such a pivotal time.” New leadership, mid-conservatorship, front-loading a phrase like “pivotal time” in his opening remarks. That’s not coincidental language.

The conspicuous absence of the words “privatization,” “IPO,” or “conservatorship exit” in either company’s transcripts is the tell. They are prohibited from front-running policy. The playbook is silence + capital build + efficiency narrative. They’re executing it precisely.

The Bear Case (I’m Not Ignoring It)

The single biggest structural risk is Treasury warrant dilution. The federal government holds warrants to acquire 79.9% of common equity in both entities. Even in a favorable privatization scenario, existing common shareholders could face severe dilution depending on how those warrants are exercised. If the warrant structure isn’t restructured in a shareholder-friendly way — and there’s no guarantee it will be — the common equity upside gets capped or eliminated entirely. This is the variable that determines whether this is a double or a washout.

The second risk is political binary. This trade lives and dies on one administration’s political calendar. A policy reversal, a housing market shock, Congressional pushback from Elizabeth Warren’s coalition — any of these can stall the thesis indefinitely. Sen. Warren has stated publicly she is “very worried that the Trump administration is very focused on how the billionaires are gonna do in any Fannie/Freddie deal.” NPR That’s a real political headwind.

Third: investor confidence in the Trump administration’s plan to sell additional shares of Fannie Mae and Freddie Mac is waning, as the initiative has stalled with no significant updates. GuruFocus Timeline risk is real. The 2026 IPO window that Ackman projected is closing.

The fourth risk is earnings deceleration — Fannie’s net income has fallen from $22.2 billion in FY2021 to $14.4 billion in FY2025. Freddie has been range-bound at $10-12 billion for three years. Credit normalization and rising multifamily delinquencies are real headwinds that could slow the capital build.

How to Size This

This is not a core position. This is a 3–5% starter to core, time-gated thesis with an embedded binary option.

The asymmetry math is straightforward: at Ackman’s base case, this is a 6–7x from current prices for the common in a favorable scenario. At his earnings multiple range with a NYSE re-listing, you get meaningful institutional forced-buying on top of the fundamental re-rating. The upside is real and math-able.

The downside is also real — common equity in a conservatorship has no fundamental floor. If the warrant structure doesn’t resolve favorably, existing common shareholders could see near-zero outcomes even in a technically “successful” privatization.

Watch the VIX and credit spread environment as your regime filter. This thesis works better in a risk-on backdrop with tightening credit spreads. In a risk-off credit event while both entities are still in conservatorship, the political will to push privatization could evaporate.

The Bottom Line

Ackman has described this as potentially generating more than $300 billion in additional profit for the U.S. government Stocktwits — which means this is a trade where the government’s interest and shareholders’ interests are aligned. That’s rare. Treasury Secretary Bessent has described a potential partial sale as “one of the biggest deals, maybe the biggest deal in history.” FHFA Director Bill Pulte’s fingerprints are on every Freddie earnings call.

The market is treating FNMA and FMCC as stranded assets in permanent bureaucratic limbo. The capital math, the management language, the political alignment, and now the loudest voices in contrarian investing are all pointing the same direction.

Ackman said it best: asymmetry at its best. The option value of privatization is currently priced at zero. The probability is not zero.

Disclaimer: Positions and price levels current as of March 30, 2026. This is not investment advice. Do your own due diligence.

The short version: he and his team have been using Claude to autonomously find real, undiscovered security vulnerabilities in critical software. No custom tooling. No security expertise required. Just a prompt that says “find the most serious vulnerability you can,” and the model goes to work.

What it found was not trivial. A popular content management system with 50,000 GitHub stars — never had a critical bug in its history. Claude found the first one, then wrote a working exploit from scratch that extracted admin credentials with zero authentication. Carlini wrote none of that code.

More significant: the Linux kernel. One of the most battle-hardened pieces of software on earth. The model found a remotely exploitable heap buffer overflow in the NFS network file system daemon — a bug requiring understanding of a multi-step interaction across a network protocol handshake that no fuzzing tool would ever surface. It has been sitting in the kernel since 2003. It predates git.

The capability cliff is recent. Models from six months ago couldn’t reliably do this. The models released in the last three to four months can. The doubling time on AI task autonomy is running at roughly four months.

Subscribe now

Why the US government will move fast on this

The same capability that finds vulnerabilities for defenders finds them for attackers. If you can deploy an AI that autonomously scans any software in the world for exploitable flaws — without needing elite human researchers — you have just handed one actor a structural offensive cyber advantage.

The US currently sits at the frontier. China is an estimated six to nine months behind, partly constrained by export controls on the compute clusters required to train frontier models. That gap is real and closing.

The DoD doesn’t need convincing. What’s happening now — quietly — is accelerated classified procurement across AI-native security vendors and government integrators who can deploy this in cleared environments. That spending will show up in contract awards and revenue over the next twelve to twenty-four months. We are inside that window.

The market’s mistake

Here’s the irony. The narrative driving cybersecurity stocks lower has been: AI will disrupt cybersecurity companies. The fear is that AI automates security work, compresses headcount, and reduces the need for enterprise software. Qualys, Rapid7, and Tenable all dropped sharply this week on that thesis. SentinelOne hit a multi-year low in February. CrowdStrike is down roughly 25% from its November peak.

The market has the causality exactly backwards.

AI is not compressing demand for enterprise security software. It is dramatically expanding the attack surface those companies must defend. When AI can generate hundreds of kernel-level vulnerabilities faster than any human team can validate and patch them — which is literally what Carlini described, noting he has “several hundred crashes I haven’t had time to validate yet” — organizations need better security tooling, not less of it.

We ran earnings call transcripts across every major cybersecurity name for the last two quarters, filtering specifically for executives who spoke about this problem in concrete terms — not generic “AI enhances our platform” marketing language. The findings were clear and they point to specific names.

The winners — and why

CrowdStrike

CrowdStrike’s CEO George Kurtz disclosed something on an earnings call that barely made headlines: China state-sponsored actors were already using public LLMs to build what he called “AI-type malware.” Not traditional executables — prompts that instruct a system to look around, identify what’s interesting, and generate scripts on the fly. Unique every time they hit a new system. No static signature to catch.

This is Carlini’s thesis playing out in real enterprise environments, right now, confirmed by the CEO of the largest pure-play cybersecurity company in the world.

CrowdStrike’s structural advantage here is their data moat. Their Falcon platform sits on sensors across hundreds of thousands of endpoints globally, generating labeled telemetry that no LLM provider can replicate with a generic model. Charlotte AI — their autonomous SOC analyst product — is growing over 85% annually and now runs end-to-end security workflows without human intervention. That’s the right architecture for a world where attacks arrive faster than analysts can triage them. Down ~25% from peak on growth deceleration fears, this is the clearest entry point in the large-cap names.

Palo Alto Networks

PANW has the most comprehensive AI threat taxonomy of any vendor in the space. Their Unit 42 threat research team published data showing end-to-end cyberattacks are now 4x faster than a year ago, with nearly a quarter of breaches resulting in data exfiltration in under an hour. That’s the real-world version of Carlini’s capability cliff hitting enterprise environments.

What sets PANW apart is forward-looking product specificity. CEO Nikesh Arora named MCP server hijacking — a vector most security researchers hadn’t publicly identified yet — as an emerging attack surface. Their Prisma AIRS 2.0 platform includes autonomous AI red teaming, real-time agent defense against prompt injection, and deep model inspection. They are not reacting to the AI threat; they mapped it before it became consensus. At roughly $147 versus a consensus analyst target around $215, the discount is real despite the premium multiple.

SentinelOne — highest asymmetry

SentinelOne is the most interesting risk/reward in the basket. The stock hit a multi-year low in February, trades at roughly 3.5x forward sales — a fraction of CrowdStrike and PANW — and carries an average analyst price target implying nearly 50% upside from current levels.

The bear case is that AI disrupts endpoint security and larger platforms consolidate the market. The bull case has a very specific piece of evidence: a top frontier AI lab — almost certainly one of the organizations building the exact models Carlini was demonstrating — selected SentinelOne’s Singularity platform to protect its own mission-critical infrastructure and model development environment. The architects of the offensive threat chose SentinelOne to defend against it. That is third-party validation you cannot manufacture with marketing spend. If the labs building the attack surface are buying S1 to protect themselves, the market’s AI disruption narrative is probably wrong about this name specifically.

Zscaler

Zscaler’s CEO didn’t speak in hypotheticals. He disclosed a real incident on an earnings call: a large AI company had its coding assistant hijacked by bad actors, who used it to autonomously execute a large-scale cyberattack across multiple organizations. This is the scenario Carlini described — AI-assisted lateral movement at scale — already happening.

What makes Zscaler uniquely positioned is their inline network vantage point. They process over 500 billion transactions daily — more than 20 times Google’s search volume — which means they see AI-assisted attack traffic in real time, at the network layer, before it reaches endpoints. Their acquisition of Red Canary layered agentic SecOps AI on top of that data fabric, enabling threat detection that the CISO of a Fortune 100 company said was “finding things every month we aren’t able to find ourselves.” That combination of inline position plus agentic response is the right architecture for catching AI-generated exploits before they land.

Qualys

Qualys is the most operationally specific name in the cohort when it comes to solving the exact problem Carlini described — the patch pipeline overwhelm. Their platform doesn’t just find vulnerabilities. It runs an autonomous loop: detect the vulnerability, validate whether it’s actually exploitable in your specific environment, apply the remediation, then re-verify the fix worked. All without a human in the loop. Their CEO called this out directly: “You cannot show up for the AI fight today with your Jira tickets. You have to be able to do automation and autonomous decision-making.”

That’s the answer to a world where AI generates vulnerabilities faster than security teams can process them. Qualys caught in the broad sector selloff creates an entry point in a name with arguably the most complete autonomous remediation product.

Booz Allen Hamilton

Booz Allen is the government infrastructure play in this thesis. They are already delivering AI capabilities for classified missions at the Defense Intelligence Agency, the Army National Guard, and across DoW/DoD, and they have the cleared workforce and mission relationships to be the primary integrator of AI-assisted cyber capability in classified environments.

What distinguishes them from other defense contractors is a named product: Vellox Reverser, their AI-native malware reverse engineering platform that performs fully automated analysis in minutes versus days of traditional work. Their CEO explicitly called AI-enhanced cyberattacks “one of the primary threats of 2026” — unusually direct language for a defense contractor. As classified DoW/DoD spending on AI cyber capability accelerates, Booz Allen is the most direct conduit for that spending in the public markets.

The chart — dislocated vs. how specifically management understands the threat

Below is where these names sit on valuation versus the conviction of their AI-specific evidence — the stocks with the strongest earnings call language and the most dislocated prices are in the top-right.

The bottom line

Carlini closed his talk with a practical admission: he has hundreds of Linux kernel vulnerabilities he can’t responsibly disclose because his team can’t process them fast enough. AI found them faster than humans can validate them.

That gap — between AI-generated vulnerability discovery and human capacity to patch — is the central problem these companies exist to solve. It is about to get dramatically harder. The market is selling the companies that solve it because it thinks AI will make them irrelevant. The companies themselves, and the researcher at Anthropic who gave this talk, believe the opposite.

One side is right.

Disclaimer: This is for educational purposes only. NFA.

Subscribe now

The company is Aritzia (TSX: ATZ) (OTCMKTS: ATZAF). The trade is a 2–4x over 36 months. And the setup is almost structurally identical to what Lululemon - looked like in 2016 — right before it became one of the best long trades of the decade.

Let me show you exactly what I mean.

Subscribe now

First, a History Lesson

In 2016, Lululemon was a punchline.

The company had just gone through a multi-year nightmare: a supply chain reset, an infamous product recall (the see-through yoga pants debacle), a CEO departure, and a gross margin that had collapsed from ~51% to ~46%. The stock had been cut in half from its 2013 highs. Institutional investors who had been burned wanted nothing to do with it. The sell-side had largely moved on.

The consensus view: Lululemon is just a yoga pants brand that got lucky, and the margin expansion story is over.

What actually happened: LULU’s gross margin went from 46% in FY2016 to 57% in FY2021. The stock went from ~$55 to over $400. A 7x run driven entirely by a thesis that was hiding in plain sight → if you understood what was driving the margin recovery and what came next.

That exact setup is playing out again. The company is different. The product is different. But the financial fingerprints are identical.

What Aritzia Actually Is (vs. What the Market Thinks It Is)

The market thinks Aritzia is a premium Canadian fashion retailer — cyclical, geographically constrained, and still recovering from a painful margin compression episode in FY2024 that burned investors.

I think that perception is wrong.

What is actually happening: Aritzia is executing a Lululemon-2014-style US market penetration, with only 72 US boutiques open today against a long-term target of 180–200+. The US already represents 62%+ of total revenue and is growing at +45–54% YoY. The company’s CEO, Jennifer Wong — who has been with Aritzia for 39 years — said on the most recent earnings call:

“I’ve been with the company now for a very long time. I’m coming up on 39 years, and I’ve never been more excited about the business as I am right now.”

That is not a boilerplate statement. A founder-operator with nearly four decades of institutional knowledge deploying “never been more excited” is a very specific, personal conviction signal. I pay attention to those.

The Numbers That Matter

Before I get into the LULU comparison, here’s where Aritzia’s business actually stands today:

• Revenue growth: ~33% YoY LTM — running faster than Lululemon at a comparable revenue scale

• EBITDA margin: ~21.4% — virtually identical to LULU FY2017–18 (21.3–21.7%)

• Boutique payback periods: Under 12 months vs. a 12–18 month target. Improving as they scale, not deteriorating. This is the single most important unit economics data point.

• Free cash flow: Swung from -CAD $68M in Q1 FY25 to +CAD $286M in Q3 FY26 — in five quarters.

• Balance sheet: CAD $620M cash, zero debt. On track to exceed $1 billion cash by FY2027.

• US active client growth: +40% YoY for two consecutive quarters. Organic awareness, not promotional pull.

The business is not recovering. It is accelerating!

The LULU Comparison — Mapped Precisely

I ran Aritzia’s current financial profile against Lululemon’s historical financials year by year to find the closest analog. The answer is uncomfortable for the bears.

Aritzia today most closely resembles Lululemon in late FY2017/early FY2019 — past the operational trough, in the middle of its US buildout, with EBITDA margins converging toward prime-LULU levels. But with one critical structural difference that makes the ATZ setup potentially more compelling than the LULU setup was:

Aritzia is generating comparable revenue with 3x fewer doors.

When LULU was generating $2.3–2.6B in revenue, it had ~411 stores. Aritzia is generating a comparable CAD revenue base with just 139 boutiques. That means Aritzia’s revenue per boutique is roughly 3x higher than LULU’s was at the same stage. Higher-format flagships. E-commerce carrying more load. Both are structurally positive — the remaining 60+ US boutiques needed to hit management’s target are starting from a higher productivity base than LULU ever had.

Here’s the comparative scorecard:

1. Revenue growth — LULU FY2019: +24% / ATZ LTM: +33% → ATZ wins outright. Growing faster on a comparable revenue base.

2. Gross margin — LULU FY2019: 55.2% / ATZ LTM: ~44.8% → LULU ahead by ~1,000 bps. This is the single unproven chapter of the ATZ thesis and the number to watch every quarter.

3. EBITDA margin — LULU FY2019: 25.2% / ATZ LTM: ~21.4% → LULU ahead, but the gap is closing. ATZ is already running above LULU’s FY2017–18 EBITDA profile on a tariff-adjusted basis.

4. FCF trajectory — Both accelerating sharply off trough. ATZ swung from -CAD $68M to +CAD $286M in five quarters. Comparable inflection shape.

5. Store count vs. revenue — LULU FY2019: 440 stores generating $3.3B / ATZ LTM: 139 boutiques generating ~$3.4B CAD → ATZ wins decisively on productivity. Revenue per door is approximately 3x higher, meaning remaining US openings start from a structurally better base.

6. Buildout runway — LULU was mid-cycle in FY2019. ATZ is early-cycle with 72 of 200 target US boutiques open. More compounding runway remaining.

The One Number That Decides Everything

There is a single metric that will determine whether this trade re-rates like LULU did or plateaus as a well-run retailer with a ceiling.

1. Gross margin.

LULU’s entire re-rating story — the reason the stock went 7x from trough to peak — was built on gross margin expansion from ~46% to ~57% over five years. Every leg of the stock’s move corresponded to a specific gross margin trigger:

• Leg 1 (+55%): Gross margin confirmed a floor. Trough is behind us. Multiple expanded from ~25x to ~35x.

• Leg 2 (+65%): Gross margin broke 53% as DTC mix shifted and menswear ramped. New investor base entered.

• Leg 3 (+70%): Gross margin hit 55%+ and the narrative shifted from “premium yoga brand” to “luxury lifestyle compounder.” Multiple broke 50x.

Aritzia’s gross margin trajectory is following the same shape with an 8-quarter lag. ATZ’s trough was Q2 FY24 at 34.99% — the company’s equivalent of LULU’s 2016 capitulation. It has since recovered to 47.15% in Q1 FY26. That is 1,216 basis points of recovery in eight quarters.

The language from Aritzia’s CFO on the Q3 FY26 call:

“The increase in gross profit margin was driven by IMU improvements, lower markdown rates, savings from our smart spending initiative and lower warehousing costs.”

That is word-for-word the same language Lululemon’s management used during Phase 2 of their gross margin recovery. Same playbook. Same execution. Same results so far.

The next chapter — Phase 3 — requires e-commerce mix shift to start accreting to gross margin the way LULU’s DTC ramp did between FY2019 and FY2020. LULU’s e-commerce went from 12% of revenue in FY2015 to 26% by FY2019, adding 300+ basis points of structural gross margin expansion in the process. ATZ’s e-commerce is growing 30%+ but currently adding fulfillment cost, not removing it. That reversal — confirmed by Q2 gross margin printing above 45% — is the single most important catalyst to watch.

The target number: 47% gross margin. If ATZ crosses that threshold on a sustained basis, the Leg 2 re-rating trigger fires.

Why The Stock Is Suppressed (And Why That’s Your Edge)

Here is the honest answer to why a business growing 33% with $620M cash, zero debt, accelerating FCF, and boutique paybacks under 12 months is trading at a ~35x forward P/E instead of 45–50x:

1. It’s a TSX-listed stock. US growth funds that would own a LULU analog cannot easily buy a CAD-denominated Toronto-listed security. Only 85 US institutions have filed positions in the OTC-traded ATZAF. Compare that to LULU at a similar stage — overwhelmingly institutionally owned by US growth funds.

2. FY2024 scar tissue. The stock dropped 65% in 2023 on the margin compression episode. Institutional memory is viscous. Funds that got burned are on the sidelines waiting for proof points that — when they arrive — will already be priced in.

3. Style box exile. US growth funds screen it out because it’s “Canadian.” Canadian funds look at 35x P/E on a “fashion retailer” and pass. It falls through the cracks of two separate institutional mandates simultaneously.

4. The CAD/USD translation friction. ATZ reports in CAD but 62%+ of revenue is USD. Analysts doing a quick screen see CAD-denominated numbers and underestimate the actual USD economic reality.

Every single one of these factors is a temporary structural condition, not a fundamental problem. And the catalyst that resolves all of them simultaneously is a US exchange listing — which becomes economically inevitable as the US boutique count crosses 100+ and revenue approaches $4B+ CAD.

This is exactly what Lululemon did in reverse. LULU listed primarily on Nasdaq with a secondary TSX listing, then delisted from the TSX entirely in 2013, stating that “Nasdaq accounts for nearly all of the company’s current trading volume.” ATZ is running the same movie backwards. The end state is the same: one dominant exchange, and it will be American.

The Shopify precedent makes this more likely, not less. Shopify moved its primary US listing to Nasdaq in March 2025, triggering immediate S&P 500 eligibility speculation. The template is established. The Canadian technology and growth company playbook now explicitly includes US exchange migration as a standard scaling move.

The Price Target Framework

Based on the LULU re-rating playbook mapped to ATZ’s current financial profile:

• Bear case (12 months): CAD $70–96 — US consumer cracks, comps go negative, gross margin stalls at 43–45%, multiple de-rates to 25–30x. Downside: -38%.

• Base case (24 months): CAD $190–253 — Phase 2 holds, FY27 guidance raises EBITDA target above 19%, Q2 FY27 gross margin prints above 45%. Multiple expands to 38–44x. Upside: +70–126%.

• Bull case (36 months): CAD $338–425 — Full LULU replay. Gross margin crosses 50%, US boutique count hits 120+, US listing announced, institutional mandate unlocks. Multiple re-rates to 45–50x. Upside: +200–280%.

The asymmetry: approximately 6:1 reward-to-risk at current levels.

The Gate Map — What to Watch and When

I’m tracking six sequential catalysts:

1. Q4 FY26 earnings (April/May 2026) — Can ATZ lap +26% comps? If yes, add.

2. Q2 FY27 gross margin (Oct 2026) — Above 45% = Phase 3 activated. This is the most important single print.

3. 100 US boutiques crossed (late 2026/early 2027) — The psychological milestone that breaks the “Canadian retailer” narrative.

4. Revenue crosses CAD $4B (FY2027) — Index eligibility threshold.

5. US dual listing announcement (late 2027/early 2028) — Re-rates 15–25% on announcement day alone as US institutional mandates unlock overnight.

6. S&P/Russell index inclusion (2028–2029) — Passive money forced to buy. This is the exit.

We’re currently at Gate 1. You have approximately 24–30 months before the mechanical institutional forces take over and make this position expensive to build.

The Positioning Framework

This is not a trade. The FY2027 boutique opening cycle, international e-commerce scaling, and mobile app monetization are all mid-cycle. The re-rating happens when US boutique count crosses 100+ and the consensus realizes the TAM math.

Conviction: Medium-High.

• Now (late March 2026): Half position. Seasonal entry in the Q1 trough window. Thesis intact.

• Q4 FY26 earnings (May 2026): If comps hold HSD+ and margin guidance raised — add to full position.

• Q2 FY27 gross margin (Oct 2026): If 45%+ print confirmed — maximum size.

• November 2026 onward: Hold through the re-rating. Trim into any US listing announcement.

The kill switch: If US active client growth — which ran +40% YoY for two consecutive quarters — decelerates sharply below 20%, or if management stops leading with that metric, exit immediately. Everything in the thesis flows from US brand awareness compounding. If that stalls, the re-rating timeline extends indefinitely.

The One-Sentence Version

The market is pricing Aritzia as a recovered Canadian fashion retailer. It is actually a US luxury brand in its earliest stages of national buildout — with only 72 of a potential 200 US locations open, boutique paybacks under 12 months, a mobile app driving 58% e-commerce growth, and structural margins already exceeding their own 19% long-term target when stripped of tariff noise.

The gross margin trajectory is following Lululemon’s post-trough playbook with an 8-quarter lag. The institutional money that will eventually re-rate this stock cannot easily buy it yet. And the catalyst that forces them to — a US exchange listing — is a matter of when, not if.

You are positioned 24–30 months ahead of that event.

Disclaimer: This is not investment advice. Do your own research. All price targets are estimates based on scenario analysis and comparable company frameworks. Past performance of comparable securities does not guarantee future results.

Share

Subscribe now

It was the ultimate toll booth. And it was extraordinarily profitable → 98% gross margins, growing 20% annually, with almost no capital intensity. Nvidia , the most dominant chip company in the world right now, has margins in the mid-70s. Arm’s are near 100%.

That business is not going away. But today in San Francisco, CEO Rene Haas announced something that changes the entire investment thesis.

Arm is now selling its own chip.

What Actually Happened

At its Arm Everywhere event, Arm unveiled the AGI CPU — a 136-core data center processor manufactured by TSMC at 3nm, designed to work alongside AI accelerators like Nvidia’s GPUs. Meta is the launch customer. OpenAI and SK Telecom are committed. Production ramps second half of 2026.

Read more

Today I want to walk you through a chart from my Bloomberg Terminal that captures something institutional traders obsess over but most investors have never heard of: the VIX-to-VIX1D spread. It just hit its widest level since October 2024. Here’s why that matters.

First, What Is the VIX — Really?

The VIX measures the market’s expectation of volatility over the next 30 days. It’s derived from S&P 500 options prices. When traders buy more puts (downside protection), those options get more expensive, and the VIX rises. When traders are complacent, puts are cheap, and the VIX falls.

But here’s the nuance: the VIX is a 30-day average expectation. It doesn’t tell you whether the market is scared about tomorrow or scared about three weeks from now. That distinction is everything.

Enter the VIX1D — a newer index that measures expected volatility over just the next single trading day. It was introduced by the CBOE in 2023 and gives us a real-time read on how anxious the market is about the immediate future versus the medium-term.

The Spread: Where the Signal Lives

When you subtract VIX1D from VIX, you get the VIX-to-VIX1D spread. This tells you something simple but powerful:

Positive spread (VIX > VIX1D): The market expects MORE volatility in the coming weeks than it does tomorrow. Traders see event risk ahead but aren’t panicking yet.

Negative spread (VIX < VIX1D): The market is more afraid of TOMORROW than the next month. This only happens during acute panic events.

Think of it like a yield curve, but for fear. “Contango” means forward fear exceeds spot fear. “Backwardation” means today’s fear exceeds forward fear. Just like an inverted yield curve signals recession, an extreme reading on the VIX spread signals a regime shift in volatility.

Reading the Chart

The Bloomberg chart shows this spread from early 2024 through today. Three features stand out.

The baseline band. For most of the chart’s history, the spread oscillates in a tight range between roughly -5 and +5. This is “normal” — near-term and medium-term fear are aligned. The dashed line near +7 marks the upper boundary. When the spread pushes above it, something unusual is happening.

The April 2025 crash. The most dramatic feature is the plunge to approximately -40 in April 2025. That was the tariff shock — an acute, immediate panic event. VIX1D (single-day fear) massively exceeded VIX (30-day fear). The market was saying: “Today might be catastrophic, but we’ll figure it out over the next month.” That kind of extreme negative reading is rare and typically marks a capitulation point. Notice the spread snapped back to normal within days. Acute fear burns out fast.

Right now. Today’s reading is the opposite. The spread has pushed to its widest positive level since October 2024. The market is telling you the next 30 days contain significantly more risk than tomorrow. This is anticipatory fear — not panic, but preparation.

Why This Matters

The timing is important. Tomorrow (March 20) is a major options expiration, and April has historically been a re-risking window after quarterly OpEx. The vol term structure is saying the March OpEx might act as a temporary pin — holding prices steady through expiration — but the market expects turbulence immediately after.

The last time the spread was this wide was October 2024, right before the November election. Another period where traders could see a binary event on the horizon and were positioning weeks in advance. The spread normalized after the event passed.

The message is consistent: wide positive spreads precede event-driven volatility. They don’t tell you which direction the market will move, but they tell you it’s likely to move a lot.

Connecting the Dots

This signal doesn’t exist in isolation. Other data from the Terminal this week tells the same story:

• High-yield bond funds have seen six straight weeks of outflows, the largest in 11 months. Credit markets are de-risking.

• The 2s10s yield curve has flattened to 43.5 basis points, its narrowest since April. The bond market is pricing economic deceleration.

• Defense spending is concentrating into a handful of prime contractors — the government is prioritizing security even as the broader economy shows stress.

All of these point the same direction: the market is transitioning from complacency to anticipation. The VIX-to-VIX1D spread is just the clearest single chart that captures that shift.

Key Takeaways

• The VIX alone is incomplete. The VIX1D and the spread between them reveal HOW fear is distributed across time.

• Today’s wide positive spread means the market sees event risk in the coming weeks but isn’t panicking about tomorrow.

• Historically, wide positive spreads precede periods of elevated market movement. The direction isn’t predetermined, but the magnitude is.

• This is confirmed by credit outflows, yield curve flattening, and other macro indicators pointing toward a regime shift.

Volatility isn’t something to fear — it’s information. Once you learn to read the VIX term structure, you stop seeing the market as a random walk and start seeing it as a conversation between millions of participants expressing their views on risk across different time horizons.

The chart above is one of the most information-dense single images in finance. Now you know how to read it.

Capitulation is the act of surrendering or yielding to an opponent under agreed-upon conditions. In a militaristic context, it is the formal document or set of terms that outlines the cessation of resistance by an armed force or a sovereign government, typically when the cost of continued conflict (or blockade) becomes unsustainable.

Unlike a total “collapse,” a capitulation involves a structured transfer of power or territory, often negotiated to prevent the complete destruction of the state’s infrastructure or civilian population.

BLUF (Bottom Line Up Front)

The Trump administration is rerunning the Venezuela playbook in Cuba — pressure via oil blockade, negotiations for leadership change, economic liberalization concessions — The market is starting to figure this out: Polymarket has Díaz-Canel out by 2027 at 69%.

This matters for Operation Epic Fury positioning and broader LATAM geopolitical risk pricing.

Read more