The Great Repricing of Intelligence

“The next Great Split will not be between those who possess intelligence and those who do not — but between those who can make it productive.”

The AI Supercycle confronts its first systemic financial test. Global growth remains dependent on accelerated computing, while semiconductor valuations, credit markets and sovereign borrowing costs begin revealing the limits of capital-intensive expansion.

The contradiction is profound: the same technological investment cycle supporting global economic growth is creating new financial vulnerabilities. The IMF warned today that a reversal in AI investment, accompanied by falling asset prices and tighter financing, could reduce global output by 1.1% by 2028 — including a 1.8% reduction in U.S. output — relative to baseline. These are modeled downside effects, not forecasts.

The financial system is beginning to distinguish between the capital required to produce intelligence and the returns generated when intelligence becomes productive. That is the next phase of the Supercycle — and it may trigger a structural rotation across global equities, credit and industrial investment.

1. 01 The IMF Warning

   The October World Economic Outlook projects global growth of 3.1% in 2026 and 3.4% in 2027, with inflation of 4.7% then 4.1%. An adverse AI correction could cut global output 1.1% by 2028 (U.S. −1.8%); oil above $130 through Q1 2027 would reduce 2027 growth by roughly half a point. The Supercycle is now large enough to influence the global business cycle itself — a transition from sector theme to macroeconomic stability issue. signal: AI concentration has made global growth structurally dependent on continued investment.

2. 02 The Semiconductor Selloff

   South Korea's Kospi fell about 5.5%, Samsung 7.4% and SK Hynix 10%; Micron lost roughly 4.5%. The catalyst was reported progress by China's CXMT on denser memory architecture — commercial viability still uncertain. Valuations have capitalized sustained scarcity and pricing power, so a credible future competitor compresses margins before producing volume. A market can grow rapidly while its leading suppliers' profitability declines. signal: The next contest is controlling the cost curve of intelligence, not expanding capacity.

3. 03 Wall Street's Divergence

   Goldman Sachs reported $18.68B revenue and $5.29B net earnings, equities trading up 72% to $6.44B, IB fees up 21% to $3.22B, and an 18.7% ROE. JPMorgan posted EPS of $5.57 (below the $5.85 consensus), with NII ex-markets up 6% to $24.7B and transactions increasingly postponed. Citigroup earned $5.2B on ~$23.8B revenue. Volatility enriches intermediation even as long-term corporate financing grows more selective. signal: Wall Street can earn more from volatility while the real economy pays more for capital.

4. 04 The Energy–Bond Collision

   Brent traded near $106.10 and the 10-year Treasury eased to about 5.22% after last week's 5.36% peak — relief, not resolution. The Global Financial Stability Report flagged the market's capacity to absorb issuance: hyperscalers now account for ~10% of global investment-grade bond issuance, versus under 1% in 2024. Technology firms are now major participants in the market for long-duration capital, competing with governments and industrials. signal: The price of financing will increasingly determine the speed of economic transformation.

5. 05 Europe Repriced

   The STOXX 600 slipped about 0.1%; ASM International and Soitec fell over 6% while chemicals rose roughly 2%. Europe's ten leading luxury groups have lost about $350 billion in market value in 2026 on persistent Chinese demand weakness. The opportunity may come less from recovering yesterday's growth model than from financing tomorrow's industrial productivity. signal: Europe must convert industrial strength into advantage in the AI-native economy.

The evolutionary mechanism runs in five stages: AI investment accelerates activity; demand rises for computing, electricity and capital; scarcity raises costs and generates exceptional profits; competition and financial discipline erode scarcity premiums; and capital reallocates toward businesses generating durable returns from productive deployment.

That fifth stage marks the possible shift from the infrastructure-building phase toward the industrial-productivity phase. Hyperscaler leverage adds systemic sensitivity — a 100-basis-point rise on an illustrative unhedged $10 billion floating-rate infrastructure loan adds $100 million in annual interest. Capital will increasingly favor reliable electricity, contracted demand and credible financing: Scaleway's AMD-based European supercomputer in Sweden, Nvidia's minority stake in Doctolib, and the Lucid–Bolt plan for 25,000 autonomous vehicles all illustrate capital moving toward applications that can generate recurring value.

Key Takeaways

For investors, boards and policymakers

• Separate AI growth from AI profitability — demand can expand while margins decline.
• Reassess capital expenditure through risk-adjusted returns; financing cost and asset life are now critical.
• Treat electricity and connectivity as strategic financial infrastructure.
• Recognize Industrial AI as the next frontier — the strongest applications improve measurable performance.
• Make capital formation the central objective of European industrial strategy.

The first AI Supercycle was about producing intelligence. The next will be about turning intelligence into capital productivity — and the financial markets are beginning to price that distinction.