Earnings, AI, and the Second Half of 2026

By Ellen Hazen, CFA®, Chief Market Strategist

Our constructive view on equities is validated by stronger-than-expected earnings growth in 2026. Fiscal policy has been supportive, AI-related capital spending has accelerated dramatically, and corporate profit estimates have risen throughout the year.

While we remain positive on equities, the key question for the second half has shifted from the pace of AI growth to whether that pace will continue or decelerate. Increasing evidence suggests hyperscaler capital spending growth may be approaching a peak, which would have implications not only for AI infrastructure beneficiaries but for broader equity markets.

We continue to favor equities over fixed income and maintain a preference for emerging market equities over developed international markets due to stronger earnings growth. At the same time, elevated valuations, higher-for-longer interest rates, and potential deceleration in AI spending warrant increased vigilance as we enter the second half. See below for what we’re watching in July.

Mid-Year Outlook

We entered 2026 expecting solid US corporate earnings growth, driven by easing fiscal policy from the 2025 tax cuts, easing monetary policy from potential interest rate cuts, and strong artificial intelligence capital spending. Consensus was expecting 14% earnings growth for US large-cap stocks when we entered 2026. This view on corporate earnings growth led to our overweight in US stocks, underweight to fixed income, and within fixed income, overweight to credit.

At the year’s midpoint, fiscal policy has provided support to near-term growth and earnings, monetary policy looks to be neutral rather than stimulative, and AI capital spending has dramatically accelerated. This stronger AI spend has thus mostly offset a newly introduced negative variable: the war with Iran, which increased oil prices by as much as 70% at the peak and cut GDP growth by a few tenths of a percentage point.

In the first half of the year, the scale of the increases in artificial intelligence spending estimates is difficult to overstate. Just the largest four hyperscalers – Alphabet, Microsoft, Meta, and Amazon – have increased estimated capital spending in 2026 from $480 billion to $712 billion over the past six months. This, in turn, has driven large positive earnings estimate revisions, particularly for those companies essential to the AI datacenter buildout: semiconductor companies like Micron and NVIDIA, semiconductor capital equipment companies like Applied Materials and Lam Research, and industrial companies like Vertiv and GE Vernova.

A year ago, as capital spending was increasing, we took comfort in the fact that for the large hyperscaler companies, spending was still easily funded out of cash from operations. With the large increase in expected spending, though, this dynamic has changed materially. Most of these companies are now spending all their cash from operations on capital spending, leaving free cash flow at zero. Cash from operations is, of course, not the only potential source of funds. Companies can fund additional increases through debt issuance, equity issuance, or other funding structures like joint ventures or off-balance-sheet financing. Even with access to additional financing sources, the pace of capital spending growth may be approaching a cyclical peak.

Until recently, many hyperscaler CEOs viewed this as a once-in-a-lifetime, generational opportunity, which led markets to believe that lack of funding simply from existing cash flow would not materially slow acceleration. Meta’s recent announcement that it will lease excess datacenter capacity may represent a modest shift in how hyperscalers evaluate future capacity needs.

As we look at the second half of 2026, we remain optimistic about equities, both domestically and globally. Domestically, because the large AI capital spend continues to drive accelerating corporate profit growth. At the same time, high valuations and decelerating AI capital spending growth may put downward pressure on the infrastructure spend beneficiaries and by extension the entire market.

Within international equities, we view emerging market equities as more attractive than developed market equities because of materially higher growth rates. Emerging market equity performance is increasingly tied to the AI ecosystem. Taiwan Semi alone represents roughly 15% of the MSCI Emerging Market Index, with Samsung Electronics and SK Hynix among the next largest constituents. Currently, earnings growth for the MSCI Emerging Markets index is expected to be 45% in 2026, 24% in 2027, and 15% in 2028. This is well ahead of the 8-10% earnings growth in developed markets as measured by MSCI EAFE, although those developed markets may benefit from inexpensive starting valuations.

Over the past few months, the US Dollar has strengthened, hurting international equity returns. The dollar’s strength has been supported by several factors, including: 1. its traditional safe-haven role and net energy exporter status during the Iran war; 2. higher-than-expected US inflation and US earnings growth relative to other markets given the preponderance of AI spending is occurring in the US; and 3. New Fed Chair Warsh’s emphasis on reducing inflation.

Because inflation has not faded during the first half of the year, we continue to believe that rates will be higher for longer, driving our underweight to fixed income. Rates in general have increased this year: the two-year Treasury yield has increased by over 0.60%, from 3.31% in December 2025 to 3.97% today. Fed Funds futures now reflect the market’s belief that the Fed will increase the fed funds rate in September or October of this year.

Bottom Line

Strong earnings growth remains the primary support for our positive view on equities. Corporate profits have exceeded expectations, AI-related investment continues to drive substantial earnings revisions, and equity markets have generally benefited from that backdrop.

Looking ahead, the most important question is whether AI investment continues to accelerate at the pace investors currently expect. Capital spending growth remains robust, but signs are emerging that hyperscalers may be placing greater emphasis on efficiency, utilization, and funding discipline.

We continue to favor equities over fixed income and maintain a constructive outlook for both US and emerging market equities. Earnings momentum remains healthy, although elevated valuations and higher real interest rates suggest that markets may become increasingly sensitive to changes in growth expectations during the second half of the year.

What We’re Watching in July

  • Earnings growth remains the market’s foundation. First-quarter S&P 500 earnings grew by 21%, while second-quarter growth is currently projected at 22%. Full-year 2026 earnings expectations have risen from 14% at the start of the year to 25% today. Much of that improvement has been driven by AI-related spending, particularly in semiconductors
  • As earnings season unfolds, investors will focus on a simple question: are companies generating measurable returns on their AI investments? Commentary around productivity gains, revenue opportunities, cost savings, and future spending plans will be closely scrutinized. A meaningful portion of this year’s earnings acceleration can be traced directly or indirectly to AI-related capital spending. If companies (and investors) increasingly question the return on investment for AI spend, leading to spending deceleration or even decline, S&P 500 earnings growth could be negatively impacted
  • Earnings growth has broadened – or has it? Investors have welcomed signs that earnings growth is expanding beyond the Magnificent Seven stocks that drove index returns over the past several years. Many of the sectors with the largest positive earnings revisions though, are still driven by AI spending. The largest positive revisions have been in the Energy sector, due to the Iran war, followed by the Technology, Materials, and Communications sectors. Earnings season should provide a clearer picture of whether profit growth expands beyond AI-related beneficiaries
  • Equity supply and demand dynamics are shifting. In June, SpaceX raised $86 billion in the largest IPO in history, while AI companies Anthropic and OpenAI have also filed confidentially with the SEC for IPOs estimated to be $60 billion each. The combination of large equity issuances – like these IPOs – and the reduction or cessation of share buybacks by large hyperscaler companies increases the supply of public equities while reducing a historically important source of demand
  • Can AI infrastructure maintain leadership? Market leadership has increasingly shifted from the companies funding AI infrastructure to the companies benefiting from that spending. Semiconductor, power, cooling, and electrical infrastructure companies have been among the largest beneficiaries
  • The Fed faces another inflation test. Despite the Fed holding the federal funds rate steady at 3.50-3.75% at its mid-June meeting, the combination of the upward movement of the “dot plot” (each Fed members’ estimate of where rates will be at various points in the future) and Warsh’s hawkish commentary means the market is now pricing in an interest rate increase by year-end 2026. At the start of the year, investors anticipated rate cuts. Today, markets are increasingly contemplating the possibility of higher policy rates. Warsh has also created five task forces to study Fed communications, the Fed’s balance sheet, enhanced and potentially real-time data sources, productivity and jobs (AI), and the Fed’s inflation framework. He expects early reports from the task forces by fall and conclusions by yearend. Of particular interest will be conclusions related to the Fed’s inflation framework and balance sheet policy, both of which could have significant implications for financial markets
  • Is inflation finally turning lower? (CPI will be reported July 15 and PCE on July 30). Inflation has remained above the Fed’s 2.0% target for over five years. Several developments could contribute to lower inflation readings later this year. Tariff-related increases will begin rolling out of year-over-year comparisons, lower oil prices will reduce pressure on energy-sensitive categories, and upcoming changes to PCE methodology may modestly lower reported inflation. Counterbalancing these forces are persistent services inflation, rising technology hardware costs (see Apple’s recent price increase announcement for a salient example), and ongoing efforts by countries and corporations to build more resilient supply chains
  • High real rates are reshaping markets. Most of the increase in interest rates this year has been due to real rates (i.e., inflation-adjusted rates) increasing, rather than to higher inflation. Currently, the real 10-year rate is 2%, similar to its level during the dot-com boom, and well above the near-zero levels seen between 2012 and 2022
  • Labor market continues to moderate. New jobs created in June were weaker than expected, at 57,000 vs. consensus expectations of 113,000. New jobs created in May were revised lower; together, these numbers paint a picture of a labor market not quite as hot as the initial May report suggested. At the same time, the unemployment rate dropped from 4.3% in May to 4.2% in June, as 720,000 people left the labor force. These figures bring the three-month moving average of new jobs created to 111,000, which is above the 50,000 estimate of jobs needed to keep the country at full employment