The comparison is understandable. Technological breakthroughs have a long history of attracting capital, capturing investors’ imaginations, and creating expectations that eventually prove difficult to satisfy. Today, enormous sums of capital are being invested in semiconductors, data centers, power, and the infrastructure necessary to support artificial intelligence. At the same time, many of the companies most closely associated with the AI buildout have become some of the largest and most influential constituents of U.S. equity markets.
But there is an important problem with simply dismissing the current cycle as speculative excess: the earnings are real.
So perhaps the more useful question is not whether artificial intelligence is real, but rather: What happens when the fundamental story is right, but the price begins to assume that it will remain right for a very long time?
One of the most important distinctions in investing is also one of the easiest to forget: a great company and a great investment are not necessarily the same thing.
A company’s revenue can grow. Its margins can expand. Its competitive position can strengthen. Its earnings can exceed expectations. And its stock can still disappoint. Investors do not purchase a company’s past earnings; they purchase a claim on its future cash flows. The price they are willing to pay today therefore incorporates assumptions about what will happen tomorrow.
As expectations rise, so does the hurdle a company must clear. That is particularly relevant when valuations leave less room for disappointment.
Valuation is rarely a useful short-term timing mechanism, but it can tell us something about the margin for error. When investors pay a premium price for an exceptional future, exceptional results may simply be required to justify the price already paid. When expectations are low, a company can create a lot of value simply by doing better than feared. When expectations are high, even very good results can feel ordinary. That is the tension we think matters today.
For the last few years, investors have focused on whether companies would spend enough to build AI infrastructure. We think the next phase is going to be about the return on that spending. Building the infrastructure is one thing. Earning an attractive return on it is another. That is where the conversation gets more interesting for us as investors.
Capital expenditures are not earnings. They are dollars spent today in anticipation of earning more dollars tomorrow.
A company can build an impressive data center, buy an enormous number of GPUs, and generate meaningful revenue from those assets. But the investment question is still the same one it has always been: Does the return on that capital justify the amount of money that was put to work?
But think about how many companies are trying to earn a return from the same ecosystem. Semiconductor companies sell the chips. Cloud providers buy the chips and build the data centers. Software companies add AI to their products. Enterprises buy those services hoping to lower costs, improve productivity, or create new revenue.
Everyone in that chain expects to get paid and earn a return beyond their cost of capital. The question is who ultimately captures the economics.
Does most of the value stay with the semiconductor manufacturers? Do the hyperscalers earn attractive returns on hundreds of billions of dollars of infrastructure? Can software companies consistently charge more for AI-enabled products? Or does competition eventually push a meaningful portion of the benefit down to the end user?
We do not think the answer will be the same across the entire AI ecosystem, and that is exactly why simply being ‘bullish on AI’ may not be enough.
There is another reason we think investors should widen the lens. The companies that won the infrastructure buildout may not necessarily capture all the economics from the adoption phase. The first stage of this cycle rewarded scarcity. Advanced semiconductors were scarce. Computing capacity was scarce. Data center capacity was scarce. In some places, electricity and grid connections have become scarce.
Scarcity is usually good for suppliers, but successful technologies do not stay scarce forever. Capacity gets built. Competitors emerge. Models become more efficient. Costs come down. Customers get more choices. Value can migrate.
In our opinion, the most interesting possibility may be that much of the optimistic AI story proves directionally correct. The spending is real. The demand is real. The earnings are real. The infrastructure buildout is real. Productivity benefits may become increasingly real as well.
However, investing has never required us merely to identify what is real. It requires us to decide what those realities are worth. When expectations are low, being less pessimistic than everyone else can be enough. When expectations are high, being right about the direction of the business may no longer be enough. Investors also must be right about the magnitude, duration, and profitability of that growth. That is the distinction we think matters most today.
Perhaps the biggest risk facing AI investors is not that the technology fails.
It may be that the technology succeeds, and everyone already knows it.
This report has been prepared for informational purposes only. It may include content generated or assisted by artificial intelligence (AI) tools. While the information presented is based on sources believed to be reliable, it should not be construed as personalized investment advice, considered a recommendation or solicitation for the purchase or sale of any security or strategy. Strategy Holdings, Attributions and/or Sectors mentioned are as of the date indicated, subject to change, and should not be relied upon as current. This does not constitute legal or professional advice and is not tailored to the investment needs of any specific investor. Registration of an investment adviser does not imply any certain level of skill or training. Due to rapidly changing market conditions and the complexity of investment decisions, supplemental information may be required to make informed investment decisions, based on your individual investment objectives and suitability specifications. Investors should seek tailored advice and should understand that statements regarding future prospects of the financial market may not be realized, as past performance does not guarantee and/or is not indicative of future results. Content may not be reproduced, distributed, or transmitted in whole or in part by any means without written permission from Globalt. Regarding permission, as well as to receive a copy of Globalt’s Form ADV Part 2 and Part 3, contact Globalt’s Chief Compliance Officer, 3200 Windy Hill Road SE, Suite 1550E, Atlanta GA 30339. You can obtain more information about Globalt Investments and its advisers via the Internet at adviserinfo.sec.gov, sponsored by the U.S. Securities and Exchange Commission. The opinions and some comments contained herein reflect the judgment of the author, as of the date noted.
Globalt Investments LLC (“Globalt” or the “Firm”) was founded in 1990. It has been registered with the SEC as an Investment Adviser pursuant to the Investment Advisers Act of 1940 since 1991. Effective October 1, 2023, Globalt is a limited liability company owned by the employees and succeeding the “Globalt Investments” which had been a separately identifiable division of Synovus Trust Co. N.A. (its former affiliate since 2002). Globalt is no longer affiliated with Synovus. Globalt’s registration with the SEC does not imply any level of skill or training and should not be mistaken for an endorsement.