Technological advancement in the United States (U.S.) over the past 50-60 years has been persistent, impressive, and essentially unstoppable. Standout examples include the personal computer, the smartphone, and the internet. These remarkable innovations have enhanced productivity and fundamentally restructured how humans live, work, and play.
With innovation being part of the fabric of American prosperity, it’s then no real surprise that the next big technological revolution is racing through development. While Artificial Intelligence (AI) isn’t brand new1, the technology has advanced quickly in recent years supported by sizeable investments from hyperscalers2, institutional investors, and sovereign wealth funds. The attention from these players along with keen interest from enterprises and consumers has driven AI to become a captivating investment theme globally.
But like prior examples, investors should understand that while new technology platforms can change the world, long-term winners and losers ebb and flow and are not always readily identifiable along the way3. As a result, investors should avoid becoming overexposed to any one investment theme in their portfolios and keep a mix of assets that have different return drivers, balanced characteristics of growth, value, and income, and effective downside protection through market cycles.
There is perhaps no better example of the excitement for AI than the eye-popping ramp of capital expenditure (capex) investments from the 5 largest U.S. hyperscalers (illustrated below in Exhibit 1). Cumulatively, the magnitude of these commitments could exceed $3 trillion for the period 2025-2028, with much of the spending funneled toward data centers, semiconductors, and supporting AI infrastructure.
Exhibit 1: Sheer Magnitude of Capital Expenditure (Capex) Investments
Source: Bloomberg, J.P. Morgan Asset Management. *Data for 2026, 2027, and 2028 reflects consensus estimates.
These aggressive investments have sparked enthusiasm and trepidation in financial markets, as perceived winners and losers are identified, in some cases resulting in extreme short-term market movements.
For perceived winners such as semiconductor companies, some stock prices have indeed “gone parabolic” and are now facing an important question regarding durability of financial results, as these companies have historically been very cyclical. As Exhibit 2 below depicts, in the past, semiconductor stocks have accounted for a modest amount of the overall market but have recently spiked to a sizeable share.
Exhibit 2: Spiked Semiconductor Exposure*
Semiconductors Account for 19% of the S&P 500
Source: Bloomberg, Macrobond, Apollo. *In emerging markets, the exposure is even more acute: three companies alone — TSMC, Samsung and SK Hynix — represent 29% of the MSCI EM Index.
More broadly, Exhibit 3 below highlights that the Technology sector has now grown to account for 37.22% of the overall S&P 500 Index. But, this doesn’t tell the whole story. Alphabet, Meta Platforms, and Netflix are categorized in the Communications sector and Amazon is categorized in the Consumer Discretionary sector. If these companies are considered Technology constituents, as conventional wisdom indicates, then the index weighting for the sector swells to over 50% and is a factor to consider for passive investors and those using it as a performance benchmark.
Exhibit 3: Sector Weightings for S&P 500 Index
Source: iShares by BlackRock
Much of the long-term promise of AI lies in higher labor productivity (output per hour worked), enhanced corporate profit margins, and (even more) future innovation and business formation. If successful, structural market forces will expand financial-market beneficiaries beyond just the Technology sector. Management teams across corporate America will embrace productive new technological capabilities in an effort to protect market share from new competition and enhance corporate profitability.
We fully expect technological advancement and innovation to continue in America and be a driver of U.S. prosperity. But for investors, chasing near-term performance in a hot investment theme can create a portfolio that becomes unbalanced and carries too much concentration risk. Despite the captivating nature of the current AI infrastructure buildout, we advise that investors remain disciplined, long-term oriented, and own assets with differentiated return drivers. This approach preserves the holy grail of appropriate portfolio diversification and keeps an eye on risk management.
1 Early AI arguably dates back to the 1950s, where machine learning was used in board games like checkers and chess. More recently, AI has been used in autonomous driving capabilities, email spam filtering, and content recommendation engines to facilitate advertisements and commerce.
2 So-called “hyperscalers” are large technology companies that own, operate, and lease out large-scale cloud computing infrastructure.
3 Winners in the personal computer era were IBM, Compaq, Microsoft, and Intel while losers were Commodore, Atari, and Tandy/RadioShack. Winners in the smartphone era were Apple, Google (Android), and Samsung while losers were Nokia, Blackberry, and Microsoft. Winners in the internet era were Cisco Systems, Amazon, and Priceline.com while losers were WorldCom, Global Crossing, and Pets.com.
The Alley Company Quarterly Letter discusses general developments, financial events in the news and investment principles. It is provided for information purposes only. It does not provide investment advice and is not an offer to sell a security or a solicitation of an offer, or a recommendation, to buy a security. The statements and opinions contained herein are solely those of Alley Company and are based upon sources and data believed to be accurate and reliable. Additional information regarding Alley Company can be found by accessing the SEC’s website at www.adviserinfo.sec.gov.
