Key Takeaways
- • J.P. Morgan Private Bank's Global Investment Strategy team (Clay Erwin, Grace Peters, Steve Parker) identifies three megatrends for 2026: artificial intelligence, global fragmentation, and persistent inflation.
- • Steve Parker, Co-Head of Global Investment Strategy, says AI-related companies have driven 75% of S&P gains, 80% of earnings, and 90% of capital spending growth since the launch of ChatGPT in 2022.
- • The team sees AI bubble risk as manageable: data center vacancy rates are near historic lows at 1.5%, with about 75% of new capacity already preleased, versus the late-1990s fiber overbuild where about 1% of installed fiber was in use by mid-2001.
- • Leverage is flagged as a "yellow" warning sign as AI spending shifts from free cash flow toward debt and equity financing, though credit spreads and cash generation do not yet cause concern.
- • The strategy team expects AI-linked earnings growth of around 20% per year over the next couple of years, favoring large-cap hyperscalers and power infrastructure as the next investment theme.
Summary
J.P. Morgan Private Bank's Global Investment Strategy team, comprising Clay Erwin, Grace Peters, and Steve Parker, outlined its 2026 capital market outlook under the theme "Promise and Pressure." The team characterized 2025 as "comfortably uncomfortable," a year marked by shifting geopolitical forces, record equity market highs, a declining dollar, and lower interest rates. For 2026, it expects a similar environment of shifting macro forces and record equity levels, a stable dollar, lower rates, and persistent inflation. Its outlook centers on three megatrends: artificial intelligence, global fragmentation, and persistent inflation.
On AI, Parker said the technology has been the dominant driver of market performance. Per the discussion, AI-related companies have accounted for 75% of S&P gains, 80% of earnings, and 90% of capital spending growth since ChatGPT's launch in 2022, and AI-related spend contributed more to US growth than consumer spending in the first half of the year. Parker argued that a market view cannot be formed without an AI view.
On bubble risk, the team used the late-1990s internet bubble as a reference point. It noted that nearly 40 million miles of fiber were installed in the late 1990s, with only about 1% in use by mid-2001. By contrast, data center vacancy rates are at historic lows of about 1.5%, and roughly three-quarters of new construction capacity is already preleased. The team rates leverage as a yellow flag: most AI spending has been funded from free cash flow, but debt financing is rising. Circular financing, in which chipmakers finance their customers, is viewed as near-term positive but as increasing ecosystem connectivity and potential volatility. On valuations, Parker described sentiment as "more excitement than euphoria," noting that AI-related multiples have declined in recent years as earnings growth outpaced price appreciation.
On investment implementation, the team favors public and private markets. Its four-point approach begins with hyperscalers such as Amazon, Google, Meta, and Microsoft, where the team believes the market has consistently underestimated capital expenditure and earnings growth, with earnings growth of about 20% per year forecast over the next couple of years. Power infrastructure is identified as the next theme, given rising electricity demand from data centers and AI model usage, which the team says is straining aging grid infrastructure.
On labor, the team expects a two-phase transition: near-term automation pressure on roles such as customer support and coding, followed by job creation in power and cybersecurity infrastructure and longer-term growth in new categories.
On AI, Parker said the technology has been the dominant driver of market performance. Per the discussion, AI-related companies have accounted for 75% of S&P gains, 80% of earnings, and 90% of capital spending growth since ChatGPT's launch in 2022, and AI-related spend contributed more to US growth than consumer spending in the first half of the year. Parker argued that a market view cannot be formed without an AI view.
On bubble risk, the team used the late-1990s internet bubble as a reference point. It noted that nearly 40 million miles of fiber were installed in the late 1990s, with only about 1% in use by mid-2001. By contrast, data center vacancy rates are at historic lows of about 1.5%, and roughly three-quarters of new construction capacity is already preleased. The team rates leverage as a yellow flag: most AI spending has been funded from free cash flow, but debt financing is rising. Circular financing, in which chipmakers finance their customers, is viewed as near-term positive but as increasing ecosystem connectivity and potential volatility. On valuations, Parker described sentiment as "more excitement than euphoria," noting that AI-related multiples have declined in recent years as earnings growth outpaced price appreciation.
On investment implementation, the team favors public and private markets. Its four-point approach begins with hyperscalers such as Amazon, Google, Meta, and Microsoft, where the team believes the market has consistently underestimated capital expenditure and earnings growth, with earnings growth of about 20% per year forecast over the next couple of years. Power infrastructure is identified as the next theme, given rising electricity demand from data centers and AI model usage, which the team says is straining aging grid infrastructure.
On labor, the team expects a two-phase transition: near-term automation pressure on roles such as customer support and coding, followed by job creation in power and cybersecurity infrastructure and longer-term growth in new categories.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.