The Velocity Edge
Morning Brief · 07:52

The Great Split: When Growth Stops Creating Value

The AI Supercycle can keep accelerating even as the financial returns on the capital financing it begin to deteriorate.

Francesco De Leo Kaufmann October 29, 2026 5 min read Capital
+$18.6bnMicrosoft quarterly free cash flow
−$9.07bnAlphabet quarterly free cash flow
80%SK Hynix Q3 operating margin
+9% netEurozone bank credit-standard tightening

“Growth determines the scale of opportunity. Capital productivity determines who captures it.”

The most consequential signal in global markets is a paradox: the AI Supercycle can continue accelerating even as the financial returns on the capital financing it begin to deteriorate.

This changes the investment thesis fundamentally. The principal risk is no longer that AI fails to deliver — it is that AI succeeds technologically while a growing share of the companies financing its expansion fail to capture adequate economic value.

01

The Great Split

Three forces now define it

First, growth and value creation are separating — revenue acceleration and expanding computing capacity do not automatically generate superior free cash flow or returns on invested capital.

Second, economic rents are moving across the ecosystem — scarce semiconductor suppliers capture extraordinary margins while the companies buying their output absorb infrastructure costs.

Third, capital formation is becoming geographically unequal — sovereign credibility, private-credit availability and energy infrastructure increasingly determine where the Next Economy can be financed competitively.

Latest reported results cover October 28–29; early-morning market indications for October 29, 2026 are not closing prices. Context: Fed funds at 3.75–4.00%, ECB deposit rate at 2.50% pre-decision, Brent near $113.91, EU battery-electric share at 22.7% YTD.

02

Five Signals Shaping the Next Economy

Where value is migrating

  1. Microsoft paired $18.6bn free cash flow with 48% Azure growth; Alphabet's cloud revenue more than doubled to $30.71bn yet heavy capex pushed cash flow negative. The market is separating technological, commercial and financial performance — and the risk is the first two accelerate while the third deteriorates. signal: The regime rewards businesses whose marginal returns exceed their marginal cost of capital.

  2. Samsung reported ₩195.3tn revenue and ₩107.4tn operating profit; SK Hynix posted ₩83.3tn on an 80% margin, with contracts covering over half of planned 2027–2028 capacity. Scarce suppliers capture pricing power while customers absorb rising capital costs. signal: Today's supplier profitability can become tomorrow's capital-cost burden for the ecosystem.

  3. The ECB survey showed a net 9% tightening in corporate credit standards against an expected 5%, with loan demand down a net 6%, even as data-centre financing demand rises. The vulnerability is not capability but the cost of deploying it. signal: The split turns damaging when financing gaps become persistent productivity gaps.

  4. The Fed held at 3.75–4.00% in a 10–2 vote; the ECB is expected to hold at 2.50%. AI's infrastructure phase is capital- and energy-intensive, so investment demand can lift prices and hurdle rates before productivity gains arrive. signal: The conditions needed to contain inflation could delay the productivity that would reduce it.

  5. EU battery-electric registrations hit 1.93m units, 22.7% of the market, up from 16.1%. But adoption is an input; productivity is the outcome. Value may rotate toward the productive users of AI — industry, logistics, enterprise services. signal: Advantage shifts from supplying intelligence to converting it into better economics.

03

Three Scenarios — Next 3–6 Months

Judgmental probabilities

Selective Acceleration

AI demand stays strong while investors distinguish cash-generative businesses from debt-dependent infrastructure. Rotation within AI toward scarce semiconductors, productive software and efficient industrial applications.

The Capital-Returns Squeeze

Infrastructure spend keeps rising, cash conversion weakens and financing costs stay elevated. Aggressive repricing of leveraged AI infrastructure and a preference for financial resilience.

The Productivity Breakthrough

Monetization, utilization and efficiency accelerate enough to offset rising costs. Broader earnings expansion and renewed flows toward enterprise AI and industrial productivity.

04

Beyond the Headlines

The central non-consensus risks

AI could experience a financial correction without a technological slowdown: weaker AI equity performance need not signal weaker demand — it may mean value is migrating within the ecosystem. Financial interdependence across manufacturers, cloud providers, banks and private credit could magnify an otherwise manageable adjustment. And Europe could lose the next investment cycle without losing its technological capabilities — financing innovation without capturing a proportionate share of its returns.

Five Takeaways

For investors, executives and policymakers

  • Equity investors — follow the marginal return, not the headline growth rate: incremental operating profit, capital employed and cash conversion.
  • Credit investors — watch the financing of capacity that has not yet generated revenue, against optimistic utilization and concentrated customers.
  • Corporate boards — accelerate productivity, not capex for its own sake; demand measurable gains in margins, utilization and costs.
  • European policymakers — industrial strategy must include the cost of capital, or productive investment will migrate elsewhere.
  • Global strategists — the next split is between value creation and value capture.

The next financial shock may not come from the failure of the AI Supercycle — it may emerge from its success, if technological acceleration creates value faster than the companies financing it can capture that value.

Francesco De Leo Kaufmann · The Velocity Edge

More from The Velocity Edge

Jump straight into another note — no need to head back.

Keep reading

More field notes, from dawn to close.

Day signals and night briefings, published like a journal for board-level AI decisions.

Back to The Velocity Edge