By Gary Aiken | September 10, 2026
Technology stocks had a rough summer. A major sector rotation saw technology, especially semiconductors, underperform as out-of-favor sectors like healthcare and consumer discretionary stocks led the market, which continued to make all-time highs.
As August rolls into September, we have started to see tech companies regain some lost ground. Revenues and earnings continue to move higher across the AI ecosystem. This growth is contingent on dollars continuing to flow to AI investments. Too much money flowing to uneconomic investments could form a “bubble”. Until OpenAI and Anthropic go public, potentially this fall, part of this landscape is hidden. However, signals from token and electricity usage, public company reporting, and commitments ending up in real cash flows and physical construction continue to show the AI boom is not busting in the immediate future.
At Concord though, one of our investment dictums is “focus on the risk, and returns will follow”. As Tech’s outperformance and overweight heads into its fourth year, it’s important to review the risks to our AI thesis.
The first risk group is concentration. A few names are a large part of the portfolio who all share a similar set of business risks. Circular or vendor financing is a common concern with firms being each other’s financiers, suppliers and customers. Analysts have been raising some accounting questions around how firms are capitalizing expenses and how depreciating those capitalized expenses impacts net income and the balance sheet. If accounting risk is something we can see, then off-balance-sheet financing risk is the unseen twin. Companies are using special purpose vehicles they fund with debt and equity to enter circular financing like arrangements, but these vehicles are often built with parent company guarantees, making defaults an on-balance-sheet risk.
Execution risks abound as well. Power, water and supply chain risks are now in the news more. Grid capacity, transformers and turbine lead times, water availability for cooling, and advanced-packaging capacity are physical constraints that can gate the buildout regardless of capital availability or demand. A growing number of local and state jurisdictions are weighing or enacting moratoriums on new data centers over power and water concerns — a real-world throttle on the pace of deployment. Politics doesn’t end at the water’s edge here either. Export-control changes (China chip restrictions), antitrust scrutiny of the mega-cap platforms, AI-specific regulation, tariff policy, and Taiwan geopolitical risk (given TSMC’s concentration of leading-edge capacity) can each reprice the theme independent of fundamentals. The 2026 midterm elections are a specific, dated point where the policy backdrop for the next two years gets set.
Then there’s the set of risks that are reflected prominently in interest rates. High price-to-earnings multiples are like long-duration bonds. AI names have been particularly sensitive to interest rates in the past. If there’s a slowdown in the economy, businesses and consumers might slow their AI spend, making it harder to justify the capex. Finally, these companies, especially for off-balance-sheet financing, have become dependent on private markets finance. The ability to IPO at a particularly high valuation is very important to the sustainability of the entire AI ecosystem both public and private.
The chart of the month shows Concord’s characterization of the S&P 500’s exposure to AI. Each bar is the combined S&P 500 weight of every company in a category of exposure. Many companies fall into multiple columns, so they don’t add to 100%. For example, Nvidia is a chip developer, a data-center supplier, a frontier-lab counterparty, and a direct-AI business all at once.
S&P 500 Exposure to AI by Market Cap
Source: Concord Asset Management, Bloomberg Finance LP
Direct AI represents about 31.7% of the index. AI is the thing being sold. These include frontier-model owners and hyperscale AI clouds, accelerator designers, AI-native compute and software, plus the names where autonomy and robotics are the stated product roadmap.
Data-center companies sell into the rack or the building. They include power distribution and cooling, networking and optics, storage and servers, engineering and construction, and the data-center landlords.
Microchips include semiconductors and semicap group. This category includes silicon exposure regardless of end customer, so it spans both AI accelerators and chip names with little AI content, plus the semicap toolmakers.
Indirect AI is one step further removed. We include the inputs that make the buildout possible and that are sold in the same form they always were. Utilities and independent power, pipelines and LNG, heavy equipment and generators, construction materials, industrial gases and chemicals, copper, and networks and towers are crucial to the buildout, but not only useful to AI.
Exposure to frontier labs indicates a customer-concentration flag rather than a business-model one. If a company discloses revenue, supply agreements or partnerships with the frontier labs, it fits here. This category captures OpenAI, Anthropic, Meta, Oracle, and SpaceX’s counterparties and data licensors. It also overlaps significantly with Direct AI and Data Center, but often investors want to know who has direct exposure to the frontier labs themselves.
It’s not all potential downside risk. We’ve identified 59.9%, 249 names, that are likely to be helped by AI. AI is used inside the company to lower costs or improve the existing product but isn’t sold as AI. It’s the largest bar because it spans whole sectors: banks and insurers, health care and pharma, retail and restaurants, defense, and most enterprise software and payments. This is why we think AI is an opportunity like railroads, the interstate highway system, and the internet. It’s hard to think of a company today that hasn’t been positively impacted by those capex booms.
At the same time, a limited number of companies are hurt by AI. Business models where AI substitutes for the thing being sold rather than assisting it: advertising agencies, IT services and consulting, seat-priced software where an agent can replace a licensed user, syndicated research, and recorded music.
AI Indifferent companies represent about 11.8%, 135 companies. Neither the demand driver nor the cost base changes much either way: consumer staples, oil and gas production, autos and homebuilders, travel and leisure, residential and retail REITs, waste services and railroads. These old-line businesses could prove a respite if the AI trade collapses.
In the end, this is just a part of our analysis, scratching the surface of a plan to control risk in client portfolios. Our base case remains that we are in the mid innings of a game that probably goes into extra innings. Prudence requires that we examine the assumptions that underpin our thesis and risks to the outlook as data emerges.
Author
Gary Aiken
Chief Investment Officer
Concord Asset Management
Gary Aiken is the Chief Investment Officer for Concord Asset Management and is responsible for macroeconomic analysis, asset allocation, and security selection, as well as trading and investment operations.
Gary has over 23 years of investment experience and holds an undergraduate degree in economics from the University of Maryland and an MBA from The George Washington University School of Business.
—
Please remember that past performance may not be indicative of future results. Different types of investments involve varying degrees of risk, and there can be no assurance that the future performance of any specific investment, investment strategy, or product, including the investments and/or investment strategies recommended or undertaken by Concord Asset Management, or any non-investment related content made reference to directly or indirectly in this article, will be profitable, equal any corresponding indicated historical performance level(s), be suitable for your portfolio or individual situation, or prove successful. This content is provided for informational and educational purposes only and should not be construed as personalized investment, legal, tax, or accounting advice. Any forward-looking statements, opinions, or projections expressed herein are based on information available at the time of publication and are not guarantees, predictions, or assurances of future events, performance, or results. Actual outcomes may differ materially due to market conditions, economic developments, regulatory changes, changes in applicable laws or regulations, and other factors.
Due to various factors, including changing market conditions and/or applicable laws, the content may no longer be reflective of current opinions or positions. Moreover, you should not assume that any discussion or information contained in this article serves as the receipt of, or as a substitute for, personalized investment advice from Concord Asset Management. To the extent that a reader has any questions regarding the applicability of any specific issue discussed above to his/her individual situation, he/she is encouraged to consult with the professional advisor of his/her choosing.
Concord Asset Management is neither a law firm nor a certified public accounting firm, and no portion of this content should be construed as legal or accounting advice. A copy of Concord Asset Management’s current written disclosure Brochure discussing our advisory services and fees is available upon request or at www.concordassetmgmt.com. Please Note: If you are a Concord Asset Management or Concord Wealth Partners client, please remember to contact the firm in writing if there are any changes in your personal/financial situation or investment objectives for the purpose of reviewing, evaluating, and/or revising our previous recommendations and/or services, or if you would like to impose, add, or modify any reasonable restrictions to our investment advisory services. Concord Asset Management and Concord Wealth Partners shall continue to rely on the accuracy of information that you have provided.
Please Note: If you are a Concord Asset Management or Concord Wealth Partners client, please advise us if you have not been receiving account statements, at least quarterly, from the account custodian.