Chapter Summaries

Chapter 1: What Made This Go Viral James van Geelen published a Substack piece on February 27, 2026 that reached ~30 million views. He’s the founder of Citrine Research, which does thematic equity and macro research — looking for cohesive narratives that connect disparate market moves. He noticed software stocks, fintech, private equity, and bonds all moving in similar directions and constructed a scenario to explain the underlying mechanism. He explicitly assigns this a 10-15% probability — not a forecast, but a tail risk scenario worth thinking through. He now regrets naming individual stocks in the piece and focuses on sectors only.

Chapter 2: The Core Thesis — AI Capability Acceleration Is the Key Variable The 10-15% probability scenario hinges on one question: what if AI’s exponential capability curve continues rather than levels off? AI agents went from ~2 minutes of autonomy on intellectually complex tasks two years ago to 8-16 hours today — in just two years. At what point do they reach multi-day autonomy? Multi-week? The problem historically: agricultural mechanization, the industrial revolution, and the internet all took 50-100+ years to displace workforces, giving economies time to adapt. AI is compressing that timeline dramatically.

Chapter 3: The Deflationary Crisis Loop The mechanism: (1) AI drives extreme supply-side productivity — massive cost deflation; (2) simultaneously, AI displaces white-collar workers rapidly; (3) income collapse leads to consumer spending collapse; (4) deflationary spiral accelerates. The 2028 central scenario: unemployment above 10%, stock market down 40%. Van Geelen emphasizes this is a scenario — not a prediction — that he uses to stress-test investment frameworks.

Chapter 4: Private Credit Is the Hidden Vulnerability White-collar workers typically have ~780 FICO scores (excellent credit) and are not historically modeled as default risk. If displaced by AI, they default on private credit loans, mortgages, and fintech lending. Private credit has exploded in size over the past decade. A cascade of high-quality borrower defaults — from people who have never been default risk before — would create systemic stress in private credit markets that existing models don’t anticipate. Apollo Capital’s early 2025 reduction of software lending is cited as an early institutional signal.

Chapter 5: Markets Already Pricing Elements of This Software stocks, fintech stocks, private equity, and bonds have all been selling off in coordinated fashion. Bond markets rallied (anticipating rate cuts). Van Geelen connects these as the same mechanism playing out — markets are implicitly beginning to price the white-collar displacement scenario even if no one is naming it explicitly. The job posting paradox: software job postings are up 11% YoY, but the composition is shifting entirely toward AI/ML engineers — a signal of replacement, not growth.

Chapter 6: Moat Compression — The Oligopoly Problem AI agents doing comprehensive, frictionless price comparison don’t experience tedium. This erodes the pricing power of any business that has benefited from consumer inertia or comparison friction: payment networks, delivery platforms (using a burrito example — AI finds cheapest delivery option, drivers find highest-paying gig, platform margins collapse), intermediaries in any market. Van Geelen: “The idea of taking half the delivery fee as the company goes away because your margin is my opportunity.” Businesses with network effects sustained artificially high margins when comparison was difficult; AI makes comparison effortless at scale.

Chapter 7: Rebuttals and Counterarguments Four main rebuttals addressed: (1) Government can intervene — van Geelen agrees but argues proactive frameworks don’t exist yet; AI lab executives are reportedly “pleading” with government to prepare, and no formal plan exists. (2) Job posting data shows net growth — true but misleading; the composition shift to AI/ML engineering represents replacement, not new job creation. (3) Historical disruptions worked out fine — true over 100 years; real-time suffering was significant (Luddites didn’t suffer in hindsight). (4) White-collar work requires personality and nuance — this is legitimate and is why the probability is 10-15%, not 50%+. The meta-insight: “Everyone seems very, very comfortable this is all going to be okay” — hindsight bias distorts our view of historical disruptions.

Chapter 8: Investment Framework and Actionable Takeaways Van Geelen’s philosophy: feel comfortable when you can envision the bull case, bear case, base case, and worst case. Find cohesive narratives connecting disparate market moves. Follow early institutional movers (Apollo’s software lending reduction). When markets latch onto any scenario with this intensity, there’s real nervousness underneath. Scenario planning matters even at 10-15% probability.

---

Summary

James van Geelen’s Substack piece went viral (30 million views) not because it predicts doom, but because it names a coherent mechanism explaining market moves that were previously unexplained. He assigns the scenario a 10-15% probability — a tail risk, not a base case — but argues it’s important enough to think through rigorously, especially given the absence of formal government response frameworks.

The core mechanism in plain terms: AI capability is compressing what used to take 50-100 years of workforce displacement into potentially a decade or less. The specific danger is that white-collar workers — who historically have excellent credit and are not modeled as default risk — could be displaced fast enough to cause a cascade of private credit defaults that existing financial models have never priced. This would hit private credit (which has ballooned in the past decade), mortgages, and fintech lending simultaneously, at the same moment consumer spending collapses from income loss, creating a deflationary spiral.

Investment implications and actionable insights:

Sectors with structural vulnerability (10-15% tail risk considerations):

• Software stocks — Already selling off; job postings up but composition shifting entirely to AI/ML. Revenue moats thinning as AI can replicate specialized functions.
• Fintech and payment platforms — AI agents doing frictionless price matching erode oligopoly pricing power. Any business extracting rent through consumer inertia or comparison friction is at risk.
• Delivery platforms — The burrito example: agents optimize both sides of the marketplace simultaneously, collapsing platform take rates.
• Private credit funds and life insurers — Exposed to a class of borrowers (780 FICO, white-collar) that has never historically been modeled as default risk.

Sectors better positioned:

• AI infrastructure plays — Demand for compute continues regardless of which scenario plays out.
• Caterpillar (CAT) — Mentioned positively as an old-economy beneficiary of Stargate data center construction.
• Businesses with unique value beyond price optimization — Any company where value is relationship-based, expertise-based, or not reducible to a price comparison.

Apple (AAPL) — Van Geelen was bullish in 2024 on the thesis that Apple would control the personal AI agent (it knows everything about you via your phone). The call didn’t work out — Apple let competitors move first. He still considers the platform thesis interesting but acknowledges the execution risk.

Tactical and portfolio considerations:

• A deflationary scenario (unlike inflation) would be bullish for bonds. If this tail risk materializes, long-duration bonds would likely outperform.
• Equity drawdowns of 40%+ in this scenario suggest protective hedges (puts, short-vol strategies on the most exposed sectors) are worth pricing.
• Following institutional early movers: Apollo Capital’s 2025 de-risking of software lending is the benchmark for how sophisticated investors are acting.

Scenario planning over prediction: Van Geelen’s explicit advice is not to forecast but to construct multiple scenarios and ensure your portfolio is not catastrophically exposed to any single outcome. At 10-15% probability, this scenario should inform hedging decisions without dominating positioning.

Watch for: Government policy response as the key swing variable. In a deflationary shock, fiscal capacity is actually higher than in inflation — the question is whether governments have a framework ready to deploy quickly. AI lab executives privately advocating for government preparation is itself a signal worth tracking.

Career insight: Van Geelen’s research approach — finding cohesive narratives that connect disparate market moves rather than analyzing individual stocks in isolation — is presented implicitly as the alpha-generating framework for thematic macro equity research.