Navigating the AI market: Narrow leadership and transformative potential

AI innovation is advancing at a remarkable pace. There is tremendous enthusiasm around AI’s scale and its transformative potential, and the long-term opportunity could prove to be generational. At the same time, we remain in the early stages of the infrastructure buildout, and the use cases, profitability, and returns on the capital being invested remain to be determined.

From a thematic investment standpoint, investors are effectively navigating two markets today: an AI market shaped by the infrastructure arms race, momentum, and rapid valuation changes, and a broader market driven more conventionally by fundamentals, earnings, economic growth, rates, and policy. The extraordinary returns of a very small group of AI beneficiaries have widened the divide between these two markets.

AI’s reach continues to expand

AI’s influence is extending across markets and the broader economy. It has revived the IPO market after six years of minimal issuance. A few companies that existed less than three years ago are coming to market at record valuations. At the same time, the enormous capital required to fund AI infrastructure buildout has pushed companies toward the bond market, contributing to elevated issuance.

The impacts are being felt well beyond capital markets. AI is driving the largest boost to GDP in history. As a share of overall economic activity, the capital being invested in the AI buildout is set to far surpass the internet-driven fiber optic buildout and even the railroad construction era.

Exhibit 1: Economic boost from the AI buildout set to surpass other historical episodes

Data as of March 20, 2026. Source: “Financing the AI Buildout.” Stijn Van Nieuwerburgh

It’s beginning to influence economic growth, labor markets, supply chains, productivity, inflation, and potentially the path of monetary policy. Over time, it could create entirely new jobs and industries while reshaping existing ones, and its impact will likely have industrial applications and revolutionize areas like healthcare. Across markets, these expectations are already visible in investor sentiment, shifting narratives, and significant valuation changes.

The long-term opportunity is still taking shape

We are still in the early innings. The potential is massive, and this could prove to be one of the most transformative technology shifts we have ever navigated. The speed of innovation is striking: it took OpenAI’s ChatGPT about three years to reach a $10 billion annual sales run rate, versus 11 years for the Apple II computer.

At the same time, we are still very early in the infrastructure buildout. We do not yet know all the use cases or potential channels for adoption, and many of the most important ones may not be obvious today. Inference and agentic AI use cases are likely to be profound. We believe there’s reason to be excited about the potential opportunities.

The promise is clear, but the winners are not

History offers many examples of transformative technologies that produce challenging investment environments. The question is less about whether AI will matter and more about who the ultimate beneficiaries will be and whether they will become profitable. AI capital needs are staggering, resulting in $700 billion in capital expenditures that will be spent in 2026 alone, with more expected to follow in the years to come.¹

However, we do not know how AI adoption will ultimately translate into monetization and profitability. Much of the AI enthusiasm seen in the market is the result of a broad investor desire to participate in this opportunity. There are massive bets seeking to extrapolate AI’s scale, scope, and range of outcomes, resulting in areas of the market being driven by narrative rather than fundamentals. Narrow leadership casts a long shadow What is clear, for the moment, is that computing power—specifically processing and memory—is the early winner. A sudden, massive increase in demand has overwhelmed supply, leading to an astronomical repricing of a handful of chipmakers. Year to date through June, six memory and semiconductor companies that represented approximately 2.2% of the S&P 500 at the beginning of the year contributed 5.2 percentage points of the index's 10.2% return—roughly half of the total. By July, their contribution had declined to 3.2 percentage points of a 10.1% year-to-date return, or about 30%, illustrating both the magnitude and volatility of the narrative.

Exhibit 2: From 50% to 30% of S&P 500 Return

Contribution of Top 6 contributors ex-Nvidia in S&P 500*

*Top 6 contributors include Advanced Micro Devices, Applied Materials Inc, Intel Corp, Lam Research Corp, Micron Technology Inc, and Sandisk Corp. Nvidia excluded from Top 6 as its size accounted for its contribution rather than its performance. Data as of July 31, 2026. Source: Bloomberg

Exhibit 3 reinforces the point about shifting sentiment: semiconductor returns separated sharply from the S&P 500, the S&P 493, software, and even the Magnificent Seven. Unlike the Magnificent Seven, this was not just a story of the index's largest companies dominating because of their existing weights, but an unusually large return contribution from a small corner of the benchmark whose price moves went parabolic.

Exhibit 3: Rotation and a wide performance range

Total returns across indices

December 31, 2025 to July 31, 2026. Source: Bloomberg

Expanding the investment opportunity set

AI has quickly become a separate and distinct market within the broader market, and this is being driven by an evolving narrative rather than solely by fundamentals. This single theme has taken on a life of its own apart from other macro, geopolitical, and policy forces currently at play.

The AI investment landscape is interesting, unique, and dynamic, and it’s moving at unprecedented speed. Investors are trying to assess the investment implications, resulting in thematic momentum, sharp valuation changes, and a wide range of performance between these highly volatile stocks and the rest of the market, creating a challenging environment for investors. This dynamic is characteristic of periods of transformative change.

While the ultimate outcome remains unknown, AI will continue to evolve, and its impact is likely to extend well beyond what we see today. With this backdrop in mind, we believe opportunities lie beyond the select few AI-related names that drove markets higher, which is positive for diversified long-term investors, particularly as the focus shifts from building AI infrastructure toward adoption, implementation, productivity gains and eventual monetization. That broadening should create a more diverse opportunity set for long-term investors focused on fundamentals, durable earnings growth, and disciplined valuations.