For more than a decade, seven large cap companies have dominated the stock market. Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta, and Tesla have grown so large that they account for roughly 34% of the entire S&P 500 (1). That kind of concentration was the result of an entire decade’s worth of large cap dominance.
What was driving their performance? The Magnificent Seven generated enormous free cash flow, paid down debt to nearly nothing, bought back stock, and initiated dividends. No wonder investors liked them.
For the past number of months, though, the Mag Seven’s fundamentals have begun to change. What were once the greatest cash generators in the world have now become the greatest spenders in the world — plowing capital into the data centers, chips, and power infrastructure needed to build out artificial intelligence. These same companies that spent a decade buying back stock are now issuing it. Companies that spent a decade paying down debt are now issuing bonds. It’s a major reversal, and one that I believe has heralded a shift to a new cycle: one that favors small caps.
When the Cash Machines Stop Generating Cash
If you’re searching for a sign of how things have changed, look no further than Alphabet: The company posted its first negative quarterly free cash flow since going public in 2004 (2), burning through $5.9 billion in the second quarter as capital expenditures hit nearly $45 billion. Meta’s free cash flow, meanwhile, collapsed roughly 91% year-over-year (3) in the same quarter, down to under $800 million, as its own capex climbed above $30 billion.
These aren’t struggling companies — they remain immensely profitable. But investors are starting to look at them differently, because the calculus has changed. As one investor told the Wall Street Journal, “You can’t be offloading this much cash and not talk about it (”4).
The question investors are posing isn’t whether AI will transform the economy; it’s whether all this capital expenditure will actually pay off in the long run. With Google, OpenAI, Anthropic, Microsoft, and others all racing for the same prize, nobody yet knows who the ultimate winners will be — or whether there’s room for more than one or two. That uncertainty makes it far harder to swallow today’s valuations on the companies doing the spending, which may be why we’re starting to see money flow elsewhere.
A Trillion-Dollar Private Stimulus
One way to think about the hyperscalers’ spending spree is that it functions like a large, privately funded stimulus package for the rest of the economy. Citigroup now estimates global AI capital expenditure could reach $8.9 trillion between 2026 and 2030 (5). And all of that new infrastructure needs to be built by someone.
This is how small cap companies may stand to benefit immensely. All of this money is eventually flowing to the companies constructing this new AI infrastructure. The scope is large. Electricians wiring new data centers. Engineering and construction firms that are building them. Firms supplying the raw materials that power the AI infrastructure and connect it to the electrical grid. Even regional banks making loans to support all this new building stand to benefit.
What do many of these companies have in common? They’re small caps. As large cap corporations continue to spend — and investors begin to grow wary of the stocks — I believe small cap corporations are posed to begin to compete once again.
Small Caps Have Already Started to Perform
A small cap revitalization isn’t just a hypothesis. The data shows it may already have arrived. The Russell 2000 small cap index has climbed more than 30% over the past year (6), sharply outpacing the broader large cap market, while the Magnificent Seven have increased 17% over the same period (7). Meanwhile, market breadth is shifting: the other 493 companies in the S&P 500 have started outperforming the mega-caps, in a genuine broadening of the market (8).
It’s worth thinking about how the current spending on AI infrastructure bears some similarities with another recent capital expenditure: the boom in U.S. energy infrastructure that began in the mid-to-late 2000s. From 2008 to 2019, production of American oil surged from five million barrels a day to more than 12 million (9). That was a major boon for the entire economy — not unlike what we may see with the hyperscalers’ nearly $9 trillion investment.
But here’s the other side of that story: Roughly 275 oil and gas producers went bankrupt between 2015 and 2021 (10), according to an analysis by law firm Haynes and Boone — not because the oil boom wasn’t real, but because the producers didn’t receive enough return on their investment. The lesson that seems most relevant: Not everyone juicing their capex is guaranteed to be a winner.
I’ve been keeping that historical parallel in mind when it comes to today’s AI buildout. The technology’s benefits — efficiency gains, new jobs, faster growth — will likely accrue broadly across the economy. But the companies spending the money to build it may not be the ones who ultimately capture the returns. If history is any guide, that’s exactly the setup that tends to launch a small cap cycle, in my opinion.
For more information, email: info@penncapital.com
- https://www.fool.com/research/magnificent-seven-sp-500/
- https://finance.yahoo.com/markets/stocks/articles/sundar-pichais-alphabet-reported-negative-110500525.html
- https://finance.yahoo.com/markets/stocks/articles/meta-stock-plunges-9-1-201045595.html
- https://www.wsj.com/tech/ai/alphabet-googl-q2-earnings-report-2026-stock-f6bdd223
- https://www.reuters.com/technology/citigroup-raises-ai-capex-revenue-forecasts-amid-rapid-enterprise-adoption-2026-03-10/
- https://www.investing.com/news/stock-market-news/small-caps-remain-close-to-alltime-high-buy-hold-or-take-profits-93CH-4865132
- https://finance.yahoo.com/quote/MAGS/performance/
- https://www.nationwide.com/financial-professionals/blog/markets-economy/articles/recent-mag-7-weakness-hasnt-hurt-the-broader-market
- https://www.eia.gov/dnav/pet/hist/leafhandler.ashx?n=pet&s=mcrfpus2&f=m
- https://naturalgasintel.com/news/north-american-oil-natural-gas-bankruptcies-plunge-in-2021-as-industry-bounces-back-from-downturn/
Disclosure:
Investing in the stock market involves gains and losses and may not be suitable for all investors. Investing in equity securities is speculative and involves substantial risk. Small and mid- cap securities may involve more risk than larger-cap securities. Investments concentrated in sector or specific industry may involve greater risk and volatility than more diversified investments. Past performance does not guarantee future results and there is no assurance the strategy will achieve similar performance under future market conditions.