“The next advantage is not scarcity — it is turning abundant intelligence into returns.”
The AI supercycle is accelerating. Its financial foundations are being tested. The next market rotation will reward productivity over scarcity, cash generation over capital expenditure, and financial resilience over technological ambition.
Three developments define the landscape this morning. The AI trade is repricing sharply: South Korea's KOSPI has fallen more than 6%, with Samsung down roughly 8% and SK Hynix 10%, as reports of advances by China's ChangXin Memory Technologies (CXMT) challenge assumptions about the durability of memory-industry margins. The IMF has raised its 2026 global growth forecast to 3.1% partly because of AI investment — while warning that an AI correction, rising sovereign yields and growing leverage could threaten financial stability. And AI financing is now a competitive battleground: Tencent is marketing at least $4.5 billion in bonds, and hyperscalers now account for 10% of global investment-grade issuance, up from less than 1% in 2024.
The AI revolution is no longer simply creating new advantages. It is beginning to redistribute the financial returns of advantages markets had assumed were permanent. The fundamental question: are investors financing the next productivity revolution — or capitalizing temporary technological scarcity as if it were a permanent monopoly?
Five Signals Shaping the Next Economy
Where the returns migrate
-
Reports of CXMT advances triggered heavy selling across Asian memory stocks; Samsung, SK Hynix and TSMC fell, while SoftBank lost roughly 8%. The advances do not yet prove CXMT can match leaders at commercial scale, but this may be the first major competitive repricing within the mature phase of the AI supercycle — compressing memory margins while reducing computing costs for developers and enterprises. signal: Differentiate manufacturers exposed to pricing pressure from equipment suppliers and downstream businesses benefiting from lower computing costs.
-
Tencent is marketing bonds across four maturities, including 10- and 20-year dollar securities; its first-half capex more than doubled to 96.3 billion yuan. With hyperscalers' IG share rising from under 1% to 10%, AI is becoming a balance-sheet industry — introducing refinancing, duration and utilization risks, and favoring issuers with recurring revenues and liquidity. signal: Watch credit spreads, order books, cash-flow conversion and the link between infrastructure spend and contracted revenue.
-
The IMF expects 3.1% growth in 2026 and 3.4% in 2027. Worldwide trade rose 8.3% between 2023 and 2025, but AI-related trade grew 29%; without it, growth would have been about 4.6%. Growth is concentrating — 32 countries now account for 95% of global growth, down from 54. Productive deployers may capture the largest long-term gains. signal: The US and advanced Asia remain central, but opportunity broadens toward industrial automation, enterprise AI and intelligent logistics.
-
Brent remains above $105 despite added Russian diesel supply; OPEC trimmed its 2026 oil-demand growth forecast to 400,000 bpd. The Bank of Italy warns that projected deficits of 3.4% in 2027 and 3.2% in 2028 could complicate Italy's exit from the EU Excessive Deficit Procedure — just as Europe needs heavy investment in infrastructure and AI deployment. signal: Favor infrastructure with predictable revenues, productivity-enhancing technology and energy-resilience projects with demonstrable returns.
-
JPMorgan, Goldman Sachs, Citigroup and Wells Fargo report today. IMF stress tests show banks representing 22% of global assets would hold weak capital under the adverse scenario, up from 18% — a stress result, not a current finding of inadequacy. Banks must finance the AI cycle, deploy AI internally and manage the risks it creates; agentic systems enlarge cybersecurity and governance dependencies. signal: Capital strength, governance and operational resilience become decisive advantages.
The Hidden Pattern — The Great Split Moves Inside AI
Conditional, not inevitable
Technological diffusion can destroy scarcity premiums while expanding total demand — not bearish for AI; it changes where profits accrue. Capital formation becomes a mechanism of competitive selection: when financing stays elevated, firms with superior cash generation sustain investment while weaker rivals postpone it.
The first AI wave concentrated profits among infrastructure suppliers. The next could reward those applying AI to manufacturing, transport, financial services and healthcare. The sequence — computing scarcity → infrastructure expansion → competition → lower deployment costs → industrial productivity — is conditional. The decisive uncertainty is whether adoption and productivity gains develop fast enough to offset margin compression and financing costs.
Where Capital Could Move Next
Differentiation over exposure
Potential beneficiaries: industrial automation, robotics, enterprise AI, electrical infrastructure, efficient computing, advanced networking and financially resilient platforms. Potential underperformers: highly valued semiconductor businesses facing new competition, leveraged infrastructure developers, firms dependent on repeated refinancing, and energy-intensive operators without pricing power.
The distinction is between technological investment that adds capacity and investment that improves capital productivity. Before the close, the decisive question is whether investors distinguish competitive repricing within AI from a broader deterioration in expected returns — tested today by US bank earnings, Tencent's bond pricing, sovereign yields, and tomorrow's US inflation print alongside ASML and TSMC results.
The next great advantage is not technological scarcity — it is the capacity to turn increasingly abundant intelligence into sustainable financial returns.