From Compute to Capacity: Why AI Infrastructure Spending May Reshape the SMID-Cap Opportunity
How hyperscaler spending is creating a fundamental earnings-cycle case for small- and mid-cap stocks.
The AI opportunity is moving from chips to capacity
Mega-cap technology companies initially dominated AI investment through platform development and model training. However, as AI spending accelerates, the opportunity is broadening toward the physical infrastructure required to bring compute online.
White Paper
From Compute to Capacity: A Real SMID Opportunity
Download the full white paper for detailed analysis on how AI infrastructure spending is creating fundamental earnings-cycle opportunities for small- and mid-cap companies.
Download Our White Paper (PDF)How big is AI infrastructure spending?
Hyperscalers — Alphabet, Amazon, Microsoft, Oracle, and Meta — collectively spend hundreds of billions annually on data centers, chips, networking, power systems, and supporting infrastructure. Approximately 50% of real U.S. GDP growth in 2025 was attributable to AI spending.
The data center vacancy rate sits around 1%, with the pipeline largely pre-committed. AI capital expenditure is flowing into land acquisition, substations, transformers, switchgear, generators, cooling systems, fiber connectivity, construction labor, and grid interconnection.
What are the biggest bottlenecks?
Primary constraints include power and grid access, where interconnection queues stretch multiple years; skilled labor shortages, as every gigawatt requires thousands of workers; and electrical and thermal infrastructure, since the majority of data center investment flows outside GPU and CPU hardware into networking, power distribution, backup power, thermal management, land, and buildings.
Why might SMID-cap stocks benefit more than large caps?
Large-cap indices concentrate in the hyperscalers and semiconductor providers funding AI capital expenditure. SMID indices have greater representation in industrial, electrical, construction, materials, energy, and infrastructure businesses executing the physical buildout.
A $1 billion revenue opportunity creates 20–44% stock-price impact for SMID companies versus approximately 1.6% for large caps. This asymmetry makes the infrastructure phase of AI potentially more significant for small and mid-cap investors.
Are investors already paying peak prices?
No. SMID allocations have declined over two decades, valuations trade at significant discounts to large caps, and fund flows remain historically depressed. This creates a compelling setup for active managers with deep expertise in the space.
Conclusion
“The first phase of AI rewarded ownership of compute. The next phase may reward ownership of capacity.” This shift potentially creates a favorable environment for SMID companies, with opportunity broadening from models to infrastructure.
Read the full analysis
Download our white paper for the complete research behind this thesis.
Download PDFGet insights from the 5Perspectives Growth team.
Receive commentaries, perspectives, and research from this investment team directly in your inbox.
The views expressed represent the opinions of Polen Capital and are not intended as a forecast or guarantee of future results. Statements are current as of the date of this material only. Past performance does not guarantee future results. Investing involves risk, including the possible loss of principal. Small-cap and mid-cap companies may be more volatile and less liquid than larger companies. References to specific securities, sectors, or market themes are for illustrative purposes only and do not represent all securities purchased, sold, or recommended. This material is provided for informational purposes only and should not be considered investment advice or a recommendation to buy or sell any security. Polen Capital reserves the right to amend this material at any time. This information may not be redistributed without the prior written permission of Polen Capital.