News & Insights

Active Fixed-Income Management: What Bond Investors Need To Know About Big Tech’s Borrowing Binge To Build Ai Infrastructure

You may have heard that artificial intelligence is a bit of a hot-button issue across all financial markets and the broader economy. Entire industries like software have fallen out of favor after being market darlings and spending years as favored investments of private equity and venture capital funds.

The world is evolving, and the narrative seems likely to continue changing based on how quickly AI will be adopted and its impacts on companies and their margins. It’s a potential revolution in the very nature of work and employment, and it’s present in the markets through large-scale debt and equity issuances to fund an expensive and rapid buildout of the underlying infrastructure. Whenever things change quickly, it’s helpful to separate the signal from the noise. We’ll give it our best effort, identifying what is fact in the current environment, what is false, and probably most importantly, the known unknowns which remain to be seen.

Facts:

The dominant AI players — Amazon Web Services (AWS), Microsoft Azure and Google Cloud Platform (GCP) plus SpaceX — are augmenting datacenter capacity to capitalize on the AI era. These are the hyperscalers.

They use names you recognize from other services, but they have been providing massive technology, data and cloud compute services for years. The public adoption of AI has brought these services to the forefront. There are others competing in the space at smaller scale including Oracle, IBM and Coreweave among others. Many share large scale exposure to the leading third-party large language model (LLM) providers, primarily OpenAI (ChatGPT) and Anthropic (Claude). AI applications are extremely vast and have massive total addressable markets. As with most emerging technologies, the version we have today is as bad as it will ever be. We do not know if the upper limits of artificial general intelligence matching or surpassing human intelligence is possible. We expect it is likely to improve at an accelerating pace from here.

The proliferation of AI has prompted multiple industries to implement systems they expect will lead to long-term improvement in productivity. These businesses may envision cutting costs by replacing employees with AI agents or expanding businesses with improvements in worker productivity. This trend is on the rise as technology improves and the savings are reflected in the business’ bottom line.

Debt issuance in the space has been massive applying technical pressure to the existing debt of the world’s largest companies and even to things like Treasurys. More than $300 billion in debt has been issued year to date by the hyperscalers and other data center providers. That is likely to increase to more than $400 billion by the end of this year. The market has absorbed this debt but demanded some concessions to previous trading levels which is some combination of just pure supply overwhelming the ability to own it, and some level of fear in the long-term prospects of the industry. This is most clearly shown by the hypserscalers where their curves have steepened, which means that investors have demanded additional compensation to own longer debt (think over 20 years) than other investment-grade companies in different sectors. That is perfectly reasonable, but also not the way the world was even 12 months ago. It is far from a given that big tech will be the most disrupted, especially on a longer time horizon.

Supply bottlenecks, local government, and resident opposition pose real risk to the speed of the datacenter buildout. While many projects to date have been able to stay on schedule and receive required government approvals, there is growing local and state government pushback and resident opposition to new AI datacenters due to concerns about water, electricity, and land use. The opposition is not divided between red or blue states but exists across the spectrum. This is unlikely to abate anytime soon and will be another challenge to navigate. Fewer or slower buildouts, however, imply better pricing for compute that is already built if not under contract.

We are still early on mass market adoption, but the user growth and demand today is massive relative to current online supply. There is a shortage in the current market based on both buyers and sellers of compute. User growth of the LLM chatbots and similar are accelerating in a pattern resembling the hockey stick adoption of many past technologies.

Fictions:

“The return on investment (ROI) of new AI datacenter buildouts is poor as shown by all of the poor/negative free cash flow these hyperscalers have seen in recent quarters.”

This is a fundamental misunderstanding (intentional or not) of basic longer-life projects. Free cash flow (which here using the traditional definition of Operating Cash flow- Capital Expenditure) is a terrible quarterly assessment of a huge buildout like this. All it is telling us is that the hyperscalers are choosing to spend more today to capture what they believe to be a very profitable and scalable business segment. What that actual longer-term ROI is remains to be seen as it is dependent on variables like the useful lives of chips.

“The chipmakers/semiconductor industry is the safest credits; they sell the picks and shovels.”

An interesting dynamic that has emerged in the investment grade universe is that many of the companies in the semiconductor industry have seen their spreads (the risk compensation on top of a risk-free rate that investors demand for owning debt) compress or at least hold steady while the hyperscaler spreads have worsened. As discussed earlier, this is especially prevalent on the longer end of the yield curve (bonds with maturities over 10 years). While this might make sense at first glance, since these are the firms that are generating outsized profits in the current environment, it is worth thinking of second order effects and the risk the market is seemingly pricing into the hyperscalers. The fear is that the market for compute will be oversupplied in the future. That may be because supply is built so fast that demand is outstretched. Perhaps large-scale users like Anthropic and OpenAI will not be able to pay. The technology could advance by leaps and bounds on the efficiency front bringing overall need for incremental supply down. These are all valid concerns and discussed in the “Unknowns” section. What is worth evaluating is who struggles the most in a world where the need for new compute slows down rapidly. It negatively impacts the hyperscalers as marginal prices come down and prospects for growth decline. However, the businesses most impacted are the pick and shovel sellers. The physical suppliers of compute and design services will see the floor fall out of the business. They have likely made their own large capital expenditures to increase the supply of chips and take or maintain market share versus competitors. Semiconductors have historically been a cyclical industry, much like energy that is near and dear to our hearts in Texas. It is a large capital expenditure business with high fixed costs, and cyclical end market demand. Historically that means you should get paid more to own that risky and volatile space, but the debt market is telling us this time is different. Maybe it is, but I think in the downside scenarios the hyperscalers are safer than the underlying equipment sellers. However, it is important to note that many are now building out their own units to compete as they will see marginally worse returns on their investments as the cost of compute comes down as demand wanes. They are not, however, entirely reliant on continued growth and buildout like chipmakers.

Unknowns:

There is no denying that things are changing fast. That brings risk. A widely adopted revolutionary technology does not guarantee the buildout and investment are good investments. There are multiple historical examples of the opposite; from railroads to the internet to the dot-com bubble. It’s hard to be certain about what will happen next. It is especially noticeable from a debt perspective where upside is capped at interest and principal. Equity can capture the vast upside in the space if the right names are involved. At the same time, given the massive supply and wider spreads in the space, it is important to identify opportunities where risk compensation, even on the debt side, is adequate.

How much compute is needed? One core question around the AI buildout is how much of it is exploding demand that will taper? That would be especially bad for the chipmakers. How much of it is recurring buildouts, especially as it concerns the need for new state-of-the-art equipment? This is more of an equity story than a debt one, but we need much more data for a firm view on how much growth there is in the space and how sustainable it is.

When will technical pressure abate? While funding needs are at least reasonably well known — though growing with each quarter as the hyperscalers continue raising capital expenditure guidance — the real question is if there is a limit for how much the market is willing to absorb. Once 5%, 10% or 20% of a portfolio is tied to AI buildout is there an appetite for more? Companies are getting creative with deal structures through the asset-backed securities market, private credit, and even equity raises. However, given the scale of the cost of the buildout it is a more than fair question without a clear answer.

Will I have a job in five years? I don’t know. I sure hope so.

Important Information

The following companies mentioned in this commentary are held in the Frost Growth Equity Fund:

Amazon, Microsoft, Oracle, (Alphabet) Google, and SpaceX

This commentary is as of Sept. 10, 2026, for informational purposes only and is not investment advice, a solicitation, an offer to buy or sell, or a recommendation of any security to any person. Managers’ opinions, beliefs and/or thoughts are as of the date given and are subject to change without notice. The information presented in this commentary was obtained from sources and data considered to be reliable, but its accuracy and completeness is not guaranteed. It should not be used as a primary basis for making investment decisions. Consider your own financial circumstances and goals carefully before investing. Certain sections of this commentary contain forward-looking statements that are based on our reasonable expectations, estimates, projections, and assumptions. Forward-looking statements are not indicators or guarantees of future performance and involve certain risks and uncertainties, which are difficult to predict. Past performance is not indicative of future results. Diversification strategies do not ensure a profit and cannot protect against losses in a declining market. All indices are unmanaged, and investors cannot invest directly into an index. You should not assume that an investment in the securities or investment strategies identified was or will be profitable.

Investment Products Are: Not FDIC Insured | Not Insured by Any Federal Government Agency | Not a Deposit or Obligation of, or Guaranteed by, Frost Bank, or Any of its Affiliates | Subject to Investment Risks, Including Possible Loss of the Principal Amount Invested

About Frost Investment Advisors, LLC

Frost Investment Advisors, LLC, a wholly owned subsidiary of Frost Bank, one of the oldest and largest Texas-based banking organizations, offers a family of mutual funds to institutional and retail investors. The firm has offered institutional and retail shares since 2008.

Frost Investment Advisors' (FIA) family of funds provides clients with diversification by offering separate funds for equity and fixed income strategies. Registered with the SEC in January 2008, FIA manages more than $5.1 billion in mutual fund assets and provides investment advisory services to institutional and high-net-worth clients, Frost Bank, and Frost Investment Advisors’ affiliates. As of July 31, 2026, the firm has $5.4 billion in assets under management, including the mutual fund assets referenced above. Mutual fund investing involves risk, including possible loss of principal. Current and future portfolio holdings are subject to risks as well.

To determine if a fund is an appropriate investment for you, carefully consider the fund’s investment objectives, risk, charges, and expenses. There can be no assurance that the fund will achieve its stated objectives. This and other information can be found in the Class A-Shares Prospectus, Investor Shares Prospectus or Class I-Shares Prospectus, or by calling 1-877-71-FROST. Please read the prospectus carefully before investing.

Frost Investment Advisors, LLC (the "Adviser") serves as the investment adviser to the Frost mutual funds. The Frost mutual funds are distributed by SEI Investments Distribution Co. (SIDCO) which is not affiliated with Frost Investment Advisors, LLC or its affiliates. Check the background of SIDCO on FINRA's http://brokercheck.finra.org/.

Frost Investment Advisors, LLC provides services to its affiliates, Frost Bank, Frost Brokerage Services, Inc. and Frost Investment Services, LLC. Services include market and economic commentary, recommendations for asset allocation targets and selection of securities; however, its affiliates retain the discretion to accept, modify or reject the recommendations. 

Frost Bank is a bank subsidiary of Cullen/Frost Bankers Inc. (NYSE: CFR). Brokerage services are offered through Frost Brokerage Services, Inc., a broker-dealer registered with the SEC and a member of FINRA and SIPC. Investment advisory services are offered through Frost Investment Services, LLC, an investment adviser registered with the SEC. Both companies are subsidiaries of Frost Bank.