AI Investing Is Evolving. Is Your Portfolio?

Artificial intelligence has quickly moved from a promising technology to one of the most consequential forces shaping today’s economy and financial markets.

The investment behind it has been extraordinary. Microsoft, Alphabet, Amazon, Meta and other technology leaders are committing hundreds of billions of dollars to the data centers, chips, energy infrastructure and computing capacity required to support AI.

That spending has helped fuel earnings growth and significant market gains among many of the companies at the center of the AI buildout. It has also raised important questions for investors:

Is all this spending paying off?
Has the market become too dependent on a handful of companies?
If AI transforms the economy, who ultimately benefits?

The first phase of the AI investment cycle has largely rewarded the companies building the technology. As adoption broadens, the benefits may increasingly extend to companies that figure out how to use it.

For long-term investors, that distinction matters.

Is All This AI Spending Paying Off?

The first several years of the AI boom have been dominated by infrastructure. Semiconductors, data centers, cloud computing and power infrastructure have all benefited as companies raced to build AI capacity.

The companies leading that buildout have an important advantage: many are among the world’s most profitable businesses and generate substantial cash flow from established operations.

The spending on AI has been enormous and is expected to grow substantially in future years.

AI capital investment has ramped up sharply over the last few years as shown in the graph below.

Source: Charles Schwab, “Navigating the Global Earnings Boom”

Charles Schwab and Bloomberg. Estimates from 1/1/2021 through 12/31/2027, data as of 7/8/2026.

Forecasts contained herein are for illustrative purposes only, may be based upon proprietary research and are developed through analysis of historical public data. All corporate names and market data shown are for illustrative purposes only and are not a recommendation, offer to sell, or a solicitation of an offer to buy any security. Past performance is no guarantee of future results.

But the ability to spend does not guarantee an adequate return on that spending.

The next test is whether that investment translates into higher revenues, stronger margins, productivity gains and sustainable earnings growth. So far, the fundamentals remain supportive: RBC estimates earnings for its group of eight large AI-related technology companies will grow about 22% in 2026, versus roughly 13% for the S&P 500. If that earnings advantage continues, it could help support spending levels and valuations.

However, there are significant risks. Competition could drive down the price of AI services faster than it reduces costs. AI adoption could take longer than investors currently expect. Compliance costs and regulations could weigh on future growth potential.

A recent research paper from a Columbia University finance professor found that to justify the current and future level of AI spending, American businesses and consumers would need to spend roughly 9% of GDP per year on the technology in 2032. Roughly the same amount as on food. While it’s possible that could be achieved in the future, that is a significant hurdle to meet to see a return on the investment.

Other Sources:

RBC Wealth Management, research report

The Wall Street Journal, “Will America Spend 9% of Its GDP on AI?”

Are Investors Becoming Too Concentrated?

AI-related companies have become an increasingly large part of the U.S. stock market, and their strong performance has increased investors’ exposure to the AI theme—often without investors making an intentional decision to do so.

The 10 largest companies in the S&P 500 now represent nearly 40% of the index, compared with roughly 18% a decade ago. Most of those companies—including Nvidia, Microsoft, Alphabet, Apple, Broadcom, Micron Technology, Amazon and Meta—are also closely tied to the AI investment cycle.

Exhibit 1: The top 10 companies represent almost 40% of the S&P 500
% of market capitalization, % of last 12 months’ earnings

Source: FactSet, Bloomberg, Standard & Poor’s, J.P. Morgan Asset Management. Guide to the Markets – U.S. Data as of March 31, 2026.

However, history shows that market leadership changes significantly over time, with relatively little overlap among the largest companies from decade to decade. Today’s market leaders are not necessarily tomorrow’s.

Exhibit 2: Top 10 S&P 500 companies by market capitalization
Percent of S&P 500 market capitalization as of the first day of the indicated year

Other Sources:

RBC Wealth Management, research report

The Wall Street Journal, “Will America Spend 9% of Its GDP on AI?”

Concentration itself is not necessarily a reason to avoid successful businesses. Many of today’s largest companies are generating substantial earnings and profit margins. But when such a large share of the market is concentrated in a relatively small number of companies that are benefiting from—and in some cases dependent on—the same investment cycle, diversification becomes increasingly important.

This is one reason our investment approach seeks to reduce concentration risk by diversifying across market capitalizations, and across the globe, rather than allowing the largest recent winners to dominate portfolio construction.

The question isn’t simply whether AI will succeed. It’s how much future success today’s prices already assume—and whether the portfolio is diversified enough if those expectations are not fully realized.

Source: J.P. Morgan Asset Management, “It Is an Active Decision to Go Passive in Equities”

Where Will the Economic Benefits of AI Appear?

This is where the AI investment story gets particularly interesting.

AI has characteristics of a general-purpose technology—an innovation capable of improving productivity across many industries, not just creating one new product or market.

History suggests that the companies developing transformative technologies are not necessarily the only long-term beneficiaries. Vanguard’s research on innovations ranging from railroads and electricity to automobiles and personal computers suggests that as transformative technologies become more widely adopted, their economic benefits can spread across industries. The analysis found that value stocks tended to perform better relative to growth stocks during periods when transformative technologies moved from early adoption toward broader economic use.

Figure 1: What history suggests about transformative technologies
Annualized return

Sources: Vanguard calculations, based on data from the Yale School of Management and the Ken French Data Library, as of March 31, 2026.

Disclosure: Past performance is no guarantee of future returns. The performance of an index is not an exact representation of any particular investment, as you cannot invest directly in an index.

Notes: Equity returns are analyzed across historical general purpose technology (GPT) cycles using a regime based framework. Each quarter from 1900 to 2023 is classified into one of three regimes—GPT emerging, GPT spreading, or other—based on the behavior of three megatrend variables: transformation, efficiency, and augmentation. These variables are calculated by the Vanguard Megatrends model. A GPT emerging period is identified when the transformation variable is rising, signaling the early adoption and deployment of a new GPT. A GPT-spreading period is identified when transformation momentum has peaked, but efficiency and augmentation continue to improve, reflecting broader diffusion across the economy. Regime classification is based on average changes in these variables over the prior five years. Market refers to Fama-French market factors and value-growth returns refer to Fama-French value factors. Industry concentration is calculated as the Herfindahl-Hirschman Index (HHI) of the Fama-French twelve industries.

AI could follow a similar path:

Lower-priced, or value stocks—companies trading at lower prices relative to their earnings and other fundamentals—could be among the beneficiaries as AI adoption broadens resulting in improved operating efficiencies and higher profit margins.

The opportunities could extend across industries. Health care companies may use AI to improve clinical workflows and automate administrative work. Financial institutions can apply it to underwriting, fraud detection and risk management, while industrial companies can use AI for logistics, automation and predictive maintenance.

That doesn’t mean investors should move out of technology or growth stocks in anticipation of an AI productivity boom. We believe investors should invest across investment styles and markets rather than trying to time when this transition could occur.

Source: Vanguard, Portfolio Perspectives

Do Investors Need to Find Tomorrow’s AI Winners Today?

We don’t think so. Let’s consider Nvidia as an example.

Vanguard calculates that an investor with $100,000 in the S&P 500 at the beginning of 2003 had only about $25 invested in Nvidia. Over the following 20 years, that small position grew to more than $60,000.

Few investors in 2003 could have predicted Nvidia’s eventual role at the center of the AI revolution.

They didn’t need to. A diversified portfolio already provided exposure as the company grew. As Nvidia became a larger and more important company, its weight in the market, and therefore in a diversified portfolio, grew with it.

That illustrates an important benefit of diversification:

Diversification doesn’t just reduce the consequences of picking the wrong winner. It can provide exposure to tomorrow’s winners before anyone knows who they will be.

The next major beneficiaries of AI may be today’s leaders. But they may also emerge from lower-priced companies, international markets or industries where AI’s most productive applications haven’t yet been developed. For investors, that reinforces the case for maintaining broad exposure to the market rather than trying to predict where the next winner will emerge.

Source: Vanguard, Portfolio Perspectives

Should This Change How You Invest Your Portfolio?

At Raffa, we don’t believe investors need to choose between participating in innovation and maintaining investment discipline.

Most investors likely already have significant exposure to AI through the largest companies in the U.S. Rather than adding exposure specifically around the theme or trying to predict which companies will ultimately benefit most, we believe the focus should remain on diversification, valuation and portfolio resilience.

Diversification: Maintain exposure across market capitalizations, sectors and stock markets outside the U.S., where future beneficiaries may emerge as AI adoption broadens.
Valuation: Maintain exposure to stocks trading at more attractive valuations. While we believe lower priced stocks will perform well over time, they could benefit as AI adoption broadens across industries and companies.
Resilience: Build a portfolio that can remain aligned with its objectives whether today’s AI leaders continue to outperform, new beneficiaries emerge, or AI takes longer than expected to produce economic returns.

Our investment approach is designed to reduce concentration risk by limiting reliance on the largest companies at the top of the market. We also maintain exposure to stocks with more attractive valuations, reducing exposure to more expensive stocks, while maintaining broad participation in the market.

That allows investors to participate in innovation without requiring their portfolio’s success to depend on correctly predicting how that innovation unfolds.

The Bottom Line

AI may prove to be one of the defining economic developments of our generation. But the investment opportunity is likely to evolve as the technology does.

Today’s AI leaders have already created enormous value. As AI moves from infrastructure buildout toward broader adoption, its benefits may increasingly appear in companies that use the technology to improve productivity, margins and growth.

Investors don’t need to know exactly where those winners will emerge.

Investors don’t need to predict how AI will reshape the economy. They need a portfolio that is positioned to participate if the technology delivers on its promise, while remaining resilient if the benefits emerge differently, or more slowly, than expected.

AI investing is evolving. A sound investment process should be positioned to participate in that evolution—not chase it.

About the Author

Mark Murphy, CFA 

Chief Investment Officer

Mark is the Chief Investment Officer at Raffa Investment Advisers and has more than 15 years of direct experience serving as an OCIO and lead portfolio manager for nonprofit organizations and membership associations. As Chief Investment Officer, Mark works closely with investment committees and staff to implement policy-driven investment programs designed to balance liquidity needs, risk management, and long-term financial sustainability. Mark is an active participant in the nonprofit and association community and has spoken on governance, fiduciary oversight, and nonprofit investing for organizations including the Greater Washington Society of CPAs, the AICPA, BoardSource, and the American Society of Association Executives.

View Mark's Bio

Disclosures:

Raffa Wealth Management, LLC dba Raffa Investment Advisers (“Raffa”) is a registered investment adviser with the U.S. Securities and Exchange Commission (SEC). Registration does not imply a certain level of skill or training. This material is for informational purposes only and should not be considered investment or financial advice. Past performance is no guarantee of future results. Forecasts and opinions are subject to change without notice. For more information, please see Raffa’s Form ADV Part 2, available at raffaadvisers.com/disclosures.

This material is provided for informational and educational purposes only and does not constitute investment advice or a recommendation to buy or sell any security. The views expressed are those of Raffa Wealth Management, LLC dba Raffa Investment Advisers (“Raffa”) as of the date indicated and are subject to change without notice.

Investing involves risk, including the possible loss of principal. Past performance is no guarantee of future results. No assurance can be given that any investment strategy will be successful or achieve its objectives.

Certain information contained herein reflects historical data and may include references to broad market indices. Index performance is provided for illustrative purposes only, does not reflect the deduction of fees or expenses, and is not available for direct investment.

Any hypothetical, back-tested, or model-based results presented are for illustrative purposes only and do not represent actual client performance. Such results are based on assumptions that may not reflect actual market conditions or investor behavior and are subject to inherent limitations. Actual results may differ materially. These materials are not intended for use by, and may not be suitable for, all investors.

The information presented has been obtained from sources believed to be reliable; however, Raffa does not guarantee the accuracy, completeness, or timeliness of such information. Any forward-looking statements or expectations are hypothetical in nature and are subject to change.

This material is not intended as a substitute for personalized investment advice. Investors should consult with a qualified financial professional before making any investment decision, taking into account their individual objectives, financial situation, and risk tolerance.

For additional information about Raffa Investment Advisers, including a description of services and fees, please request a copy of our current Form ADV Part 2A.