Wells Fargo Claims AI Spending Is Now America's Growth Engine

AI capital expenditure is now the main driver of growth, exceeding consumer spending. This significant shift is robust enough to sustain economic expansion despite households reducing their spending. This core assertion was outlined in the Wells Fargo Investment Institute's (WFII) midyear 2026 outlook report on the US economy.

The institute forecasts full-year 2026 US real GDP growth of 2.2%, slightly below the average pace since 2009. WFII describes a rotation from consumer-led to investment-led growth, paired with a parallel shift from services to manufacturing-led strength, driven by AI-related capital spending.

Energy-related headwinds tied to the war in Iran are expected to squeeze household purchasing power for the remainder of 2026, hitting lower-income consumers hardest, according to the report.

WFII also points to a widening gap between smaller and larger company finances, noting that larger firms typically have greater pricing power, are less credit-sensitive, and have access to a broader array of low-cost financing, a disparity the institute expects to persist through the rest of 2026.

The report's position is that AI investment will counterbalance that consumer-side weakness rather than the two forces simply offsetting each other into stagnation. WFII's thesis implies that if AI capex continues at its current pace, a soft consumer number does not necessarily mean a soft growth number.

It also directly addresses concerns about AI-related job displacement. WFII states that announced layoffs in 2026 have centered on AI's frontline technology sector, which it says masks net declines elsewhere in the labor market.

The institute notes that massive layoffs have never accompanied periods of technological innovation since the industrial revolution nearly 200 years ago, and that it expects AI's job-enhancement opportunities and revenue-driven hiring gains to exceed job displacement in this cycle.

Why Earnings Growth Becomes the Story

WFII expects 2026 earnings growth for S&P 500 companies to outpace start-of-year estimates at the strongest pace since 1991. The institute is explicit about why this earnings cycle is different from prior bull markets.

Past periods of strong stock performance have often relied heavily on financial leverage or cost-cutting to drive profit growth, according to the report.

WFII frames the current cycle as different in kind. A broad, multiyear investment cycle spanning AI infrastructure, compute capacity, data-center expansion, electrification, and the adjacent industrial and financial ecosystems that support that buildout.

The institute's central thesis is that AI-driven productivity boosts profit margins, leading to corporate earnings growth that more strongly influences stock prices than valuation expansion.

WFII notes that current S&P 500 valuations remain only slightly above their 1980 to 2020 average, meaning further multiple expansion has less room to run, and that rising forward earnings expectations should help compress those valuations over time without requiring stock prices to stall.

Within equities specifically, WFII states a favorable view on US Large Cap and US Mid Cap stocks, citing strong balance sheets, pricing power, diverse global revenue streams, and the ability to self-fund capital investment as reasons these companies are best positioned to absorb commodity-price fluctuations and geopolitical uncertainty.

The institute maintains an unfavorable rating on US Small Cap Equities, citing higher interest rates, greater labor-cost sensitivity, and limited pricing power that it says constrains earnings visibility for smaller firms.

According to Finimize's summary of the report, sectors tied to the AI buildout including Information Technology, Industrials, Materials, and Utilities, could see more pronounced swings depending on whether analysts raise or cut profit forecasts during reporting periods, rather than moving primarily on macro sentiment or rate expectations.

The Hardware-Over-Software Position

Within its favored Information Technology sector, WFII states an explicit preference for hardware over software.

The hardware-related areas include Semiconductors, Semiconductor Materials and Equipment, Communications Equipment, and Technology Hardware. WFII is direct about the limits of that conviction.

The report states it does not yet have the visibility to call for broader participation at the application layer, at least among public companies, even as it acknowledges participation could broaden. That stance has a direct parallel in the way WFII has adjusted its view of the credit market.

The institute notes it has tempered its previously cautious stance toward bonds issued by technology companies building data centers and other AI infrastructure, citing rising yields on those issues and balance sheet improvements among issuers.

WFII separately states that private debt is facing closer scrutiny as investors reassess credit conditions and the potential impact of AI across the economy, leading to increased redemption requests and pressure on asset prices and valuations in that market.

They also noted it has not seen signs of broad-based credit deterioration to date, and that unrealized losses in direct lending portfolios remained relatively low as of the end of 2025, which they say suggests widespread losses may not be imminent.

Where the Capital Is Actually Flowing

The venture capital data in the report claims that deal activity reached a record $267 billion in the first quarter of 2026, and AI and machine-learning companies accounted for more than 89% of that total deal value.

WFII notes that activity was concentrated in a small number of mega deals, indicating that the broader venture market outside AI remains in a more gradual recovery. The capex buildout is also expected to extend beyond the companies typically associated with AI headlines.

WFII points to power generation, grid infrastructure, electrical equipment, and engineering services as adjacent beneficiaries, alongside Industrial Metals tied to the physical build of data centers and the power infrastructure that supports them.

Data Center REITs are named as a favored real estate sub-sector specifically because of AI adoption and the demand for low-latency network connections within data center facilities.

The institute also extends the thesis internationally, noting that because much of the AI hardware supply chain sits outside the United States and the investment theme is expected to spill into Asian emerging-market chip and component makers that supply the semiconductors and electronic components the buildout requires.