The Great Split Reaches Corporate Credit

“The Great Split is moving from technology valuations into the balance sheet.”

The AI Supercycle is entering a decisive financial transition. Schneider Electric has raised its earnings outlook as data-centre demand accelerates, while CoreWeave’s bonds trade at yields near 15%, reflecting concern about debt-financed AI expansion. China is preparing new controls on AI technology exports, and the IMF warns that Europe must reconcile weak growth with increasingly restrictive sovereign finances.

The evidence reveals a new regime: computing demand keeps accelerating, but the cost of financing it is becoming decisive. The first phase of AI rewarded technological scarcity; the second rewarded infrastructure accumulation. The third will distinguish businesses that convert accelerated computing into recurring cash flow from those whose expansion depends on continuous access to expensive financing.

1. 01 The AI Credit Shock — Capital Is Repricing the Supercycle

   Reuters reports that CoreWeave bonds issued in June at a 9.675% yield have traded recently near 15%; the shares fell roughly 15% between Monday’s open and Thursday’s close before rebounding on renewed Nvidia support. The market is beginning to differentiate three AI models: established platforms with substantial cash flow, industrial suppliers paid as infrastructure is built, and developers that must finance expensive assets before achieving utilization. As borrowing becomes costlier, weaker operators may delay projects, accept dilution or sell to stronger competitors — the Supercycle can expand even as the number of companies able to finance it contracts. signal: Project backlog is not cash flow, and contracted demand does not eliminate refinancing risk.

2. 02 Industrial AI — The Physical Economy Is Capturing the Returns

   Schneider Electric reported third-quarter revenue of €11.46 billion, up 16.9% year-on-year, raising its 2026 organic revenue outlook to 12–14% and adjusted EBITDA growth to 18–21% on exceptional data-centre demand — though shares still slipped in morning trade. The contrast with CoreWeave is revealing: Schneider earns as projects advance, while developers must finance assets and then monetize capacity. Electrical equipment, transformers, advanced cooling, power electronics, storage, grid software and automation are becoming essential to accelerated computing. signal: The metric is not AI exposure, but the share of spend converted into cash-generating orders and durable margins.

3. 03 China — The AI Race Moves From Chips to Technology Control

   China proposed expanded restrictions covering 40 technologies, including 23 AI-related categories spanning models, intelligent agents and computing clusters; consultation runs to November 15, and access via APIs, apps and SaaS is excluded. Washington restricts China’s access to advanced semiconductors; Beijing now seeks to control outbound transfer of its own capabilities. The likely outcome is not full separation but managed interdependence combined with strategic control over intellectual property and infrastructure. signal: The line between access to AI services and ownership of the underlying architecture will grow decisive.

4. 04 Energy — The Geopolitical Premium Is Now a Capital Cost

   Brent traded near $105 on Friday as the IEA agreed to accelerate release of remaining March-pledged emergency stocks, including over 20 million barrels of diesel; roughly 330 million of the 400-million-plus barrel commitment is already released. Higher energy prices raise operating costs and can sustain higher rates, lifting financing costs for capital-intensive investment. signal: Energy resilience is no longer defensive — it is an instrument of capital productivity.

5. 05 Europe — Sovereign Debt Meets the Industrial AI Imperative

   The IMF warned European economies must raise growth or cut borrowing, forecasting eurozone GDP growth of 1.1% in 2026 and 1.2% in 2027, with roughly three-quarters of countries needing adjustments averaging 3.5% of GDP against current plans near 2%. Europe must invest more in industrial AI, defence, energy and networks even as public finances tighten — yet consolidation without productivity growth risks cutting the very investment capacity needed for long-term sustainability. signal: Europe’s advantage depends on whether its financial system can turn savings into productive investment at continental scale.

Afternoon releases include U.S. import/export prices (14:30 CEST) and Fed industrial production (15:15 CEST). On Wall Street (15:30 CEST), watch whether markets keep separating resilient technology companies from leveraged infrastructure providers. In Europe, monitor the German Bund (near 3.42%) and French spreads after the IMF warning, with the STOXX 600 up around 0.6%, plus IEA reserve releases and tanker security near the Strait of Hormuz.

The next inflection may emerge from the interaction between AI-related credit spreads and government-bond yields. If corporate credit risk rises while sovereign yields decline, the market is signaling a flight to financial quality — not a general easing of conditions.

The AI Supercycle is not ending its technological expansion; it is entering a far more demanding phase of financial selection. The next trillion dollars will not be won by those who spend it fastest, but by those who can convert it into productive cash flow before the cost of capital catches up.