The BlackRock Investment Institute's Q4 2026 Global Outlook1 arrives at a moment when the most consequential forces shaping markets are not moving in the same direction. The AI buildout is accelerating earnings and capital demand simultaneously. Yields have surged, restoring income potential but shrinking the toolkit for capturing it. And the old vocabulary of asset-class labels is struggling to keep up with a world being reshaped beneath it. The team frames all of this under a single, precisely chosen headline: scarcity versus abundance.
The AI Engine, and Its Binding Constraints
The team's first and most forceful conviction is that the AI buildout is not slowing. "The AI buildout is accelerating, bringing binding constraints forward," the authors write, elevating the theme from speculative to structural. What is binding, specifically, is not imagination or capital intention but physical reality: power grids, memory chips, data centers, and the balance sheets required to finance them. Hyperscalers have continued raising capital expenditure guidance, with 2026 capex now expected to reach approximately US$720 billion. Dynamic random-access memory (DRAM) has risen 75% year to date. AI suppliers within the S&P 500 have seen trailing earnings grow to more than three times the January 2023 baseline, while companies with no AI exposure have barely moved.
Against that backdrop, the team finds no warning signs. Credit has decoupled from the broader market without showing stress. Anthropic's tracked annual recurring revenue reached US$74 billion in late July against US$41 billion for OpenAI, confirming that monetization is scaling. The team's assessment remains constructive, though it is watching operating cash flow versus earnings and frontier model economics carefully. The portfolio expression is direct: overweight U.S. equities, focus on AI bottleneck opportunities in power, chips, and data centers, and now upgrade emerging market equities to overweight from neutral.
A Different Kind of Abundance Problem
While AI points toward a productivity future, the team is careful not to conflate possibility with inevitability. "We've long argued that the advent of AI means abundance, including a breakout of the U.S. long-held 2% growth trend, is plausible for the first time," the authors note. The qualifier is significant: plausible. Because the other force the team is tracking, higher interest rates, may divert the potential dividend away from households and traditional capital-intensive industries before it arrives.
The intersection of these forces produces a clear framework. Companies with high AI upside and low exposure to refinancing risk, hyperscalers, semiconductors, data centers, are positioned to benefit. Households, real estate, and traditional capital-intensive industries face rate pressure that could more than offset any productivity dividend in the near term. The tension between these outcomes defines the scarcity-versus-abundance tension the report is named for.
Fixed Income's New Geometry
"Global government bond yields have surged through the summer. Higher yields have restored income opportunities, but only via a flexible approach," the team writes, and the portfolio implications are extensive. Roughly 84% of fixed income assets now yield above 4%, a dramatic reversal from the post-pandemic zero-rate era. The yield curve, nearly flat in 2020, now sits steeply above 4% across maturities. But the team's response is not to chase duration. Long-duration bonds carry high rate sensitivity and have become less reliable diversifiers. The team is underweight long U.S. Treasuries, underweight Japanese government bonds, underweight long-term investment-grade credit, and overweight agency mortgage-backed securities and emerging market local currency debt.
The risk is fiscal. Investors are still financing historically large U.S. borrowing needs for relatively modest compensation. "A jolt to term premium," the team observes, "could spark further surge in yields." That term premium has been rising, and sovereign borrowing is increasingly competing with extraordinary AI-driven private financing needs, pushing prospective capital demand toward levels rarely seen outside crisis periods.
Beyond Labels
The report's third theme is perhaps its most architecturally important. "Asset-class buckets are a less useful starting point in a world shaped by mega forces," the team writes, pointing to infrastructure as the illustration. A data center sits simultaneously in equity markets, bond markets, private credit, and energy infrastructure. It is exposed to AI demand, geopolitical fragmentation, and energy policy at once. Investors who screen it through a single label will miss the exposure they already hold, or fail to see the additional exposure they should build. The team believes some investors can meaningfully boost infrastructure allocation across both public and private channels, calibrated to illiquidity tolerance.
Five Key Takeaways for Advisors and Investors
- Overweight U.S. equities and newly overweight emerging markets, focusing exposure on AI bottlenecks rather than broad index proxies.
- Avoid long-duration government bonds; prefer short- and medium-term Treasuries, agency MBS, and EM local currency debt for income.
- Monitor refinancing risk in AI-adjacent debt, particularly data center ABS and leveraged loans maturing through 2031.
- Rethink portfolio labels: infrastructure spans asset classes and deserves deliberate, not incidental, allocation.
- Stay selective in credit; higher-rated high yield offers better risk-adjusted compensation than investment-grade in the current spread environment.
Footnote:
1 BlackRock Investment Institute. 2026 Global Outlook: Q4 Update, Scarcity vs. Abundance. BlackRock, September 2026. https://www.blackrock.com/us/individual/insights/blackrock-investment-institute/outlook
