Artificial intelligence: implications for the economy and equity and fixed income markets

AI is shifting from an infrastructure scarcity trade into a multi-year investment cycle where monetisation, returns on capital and balance-sheet discipline will determine the winners.

• Spending by the largest cloud and technology companies on AI infrastructure is expected to exceed $750 billion by the end of 2026, up approximately 70% year-on-year, with a further 24% increase projected for 2027; cloud revenue is expected to accelerate into the mid-40% range in 2026/27.
• AI-related investment could add 40–50 basis points to US economic growth in both 2026 and 2027, while data-centre demand may increase US power requirements by approximately 3–4% annually through 2030.
• The value chain is broadening from semiconductors, memory and cloud platforms to networking, power equipment, grid modernisation, cooling, data-centre real estate and software adoption; data-centre construction loans and asset-backed securities are emerging as new fixed-income segments.
• Semiconductor demand remains constrained: data-centre revenue is growing approximately 100% year-on-year, committed sales through 2026 total $350 billion, and high-bandwidth memory requires roughly three times the manufacturing capacity of traditional memory, with supply relief not expected until late 2027.
• Labour-market disruption is currently gradual, with AI-related cuts representing approximately 40% of announced layoffs in May 2026, while hiring restraint is more significant than recorded redundancies; productivity gains could eventually reach 6–16% over a decade, although the estimates remain speculative.
• The US captures much of the platform and cloud economics, Asia dominates hardware and manufacturing, and Europe faces weaker platform leadership and regulatory headwinds; the paper also highlights concentration risk and separates AI exposure into infrastructure, enabling assets and productivity adopters.

Takeaway: Investors should treat AI as a multi-year, multi-asset investment wave rather than a single technology trade, favouring active selection across equities, infrastructure and credit while remaining alert to valuation, concentration, monetisation and execution risks.