The word “hyperscaler” does not appear in our venerable Webster’s Ninth New Collegiate Dictionary. Fairly confident that the term did not refer to cleaning fish or someone in need of an extra-large device for their daily weigh-in, we did some additional research and found the following definition:

Hyperscalers get their name from hyperscale computing, a method of processing data that allows for software architecture to scale and grow as increased demand is added to the system.

Enterprise applications and data demand flexible environments that offer a reliable way to develop and deploy technologies such as artificial intelligence, internet of things (“IoT”), machine learning, and big data analytics. Hyperscalers are able to support these services on a grand scale, using multiple datacenters around the world, with thousands of physical servers running millions of virtual machines.¹

We stumbled across the term hyperscaler while looking for explanations for the tremendous outperformance of the largest global companies relative to their medium and smaller-sized peers. In the three-year period ended October 31, 2024, the MSCI World Index of developed-country stocks produced an average annual total return of 6.4%. MSCI’s World Small Cap Index rose 0.1% per year over the same period. This phenomenon is pronounced in the U.S., where the large cap Russell 1000® Index, led by hyperscalers like Microsoft, Amazon, Meta and Google (and, of course, the hyperscale enabler NVIDIA) returned 8.1% annually over the last three years, while the small cap Russell 2000® Index lost -0.1% per year. Likewise, Goliath has also trounced David in non-U.S. markets, where the MSCI EAFE Index returned 2.7% in U.S. Dollar terms annually over the last three years, while the EAFE Small Cap Index limped home with a -3.0% annual decline. Given the recent outperformance of large caps, the largest two deciles of stocks within the MSCI EAFE Index now include just 20 securities while the smallest two deciles consist of 438.

‍

This result contradicts decades of market data showing that small companies outperform larger ones—even on a risk adjusted basis. Smaller companies are more entrepreneurial, agile, and innovative than sprawling giants and can grow faster off a lower base. Indeed, most of today’s hyperscalers grew faster in their early years of existence than in more recent times. Small stocks may also be less scrutinized than large companies, leading to greater mispricing, more “positive surprises” and strong performance when their success is “discovered” by investors.

What Has Changed?

Why are the largest companies now leading the way? Big companies enjoy some important advantages in the digital age. For example, the network effect describes how a product or service becomes more valuable as more people use it. Facebook would have some utility if a million people used it to keep up with friends, swap recipes and plan parties. But Facebook has an estimated 3.07 billion users, making it the world’s largest social network, almost indispensable to users, and immensely valuable to advertisers. While other networks, like TikTok, may spring up, competition is unlikely to take away from Facebook’s value to customers.

Big companies are also able to invest on a scale that is prohibitive to smaller organizations. Microsoft, for example, spent $13.8 billion on capital investment in a recent quarter, and analysts estimate that its capital spending could top a staggering $60 billion over the next year. For context, the median stock in the MSCI EAFE Index has a market capitalization of just under $15 billion.

Much of Microsoft’s spending will go toward cloud computing resources and artificial intelligence development. NVIDIA’s most powerful AI chips can cost between $30,000 and $40,000 each, illustrating the high cost of training and operating AI models. Clearly, only a behemoth like Microsoft and a handful of other companies can compete in this promising new market.

Novo Nordisk, the Danish maker of the weight loss drug Ozempic, is not a hyperscaler, but it is the largest company by market capitalization in the MSCI EAFE Index of non-U.S. developed companies. Novo Nordisk opened a new pharmaceutical ingredient manufacturing facility in North Carolina in 2021 that cost more than $2 billion. The company announced that it will spend more than $4 billion on an expansion at this site to meet soaring demand for its products. In addition to production costs, research and development expenses and obtaining regulatory approvals mean only the biggest companies can hope to compete in the market for blockbuster pharmaceuticals. These large companies also spend heavily on lobbying to bend regulation and government spending in their favor.

Some of the advantages enjoyed by the very largest companies, such as the network effect, the scale benefits of reaching millions or even billions of customers around the world and the flexibility afforded by a massive balance sheet are, in our opinion, sustainable and valuable. These stocks should not be avoided merely because we have never seen companies this large in the past. Some of them can, and probably will grow even bigger.

Tilting Smaller

Within the MSCI World Index there are six companies—Apple, NVIDIA, Microsoft, Alphabet (Google), Amazon, and Meta—with market capitalizations greater than $1 trillion, a level that seemed unthinkable a decade ago. These companies account for 20.9% of the index of more than 1,400 constituents. A decision not to own even one of these companies amounts to a substantial active bet that it will underperform the rest of the index. Over the last three years, that bet would have created a nearly insurmountable performance obstacle for an active manager seeking to outperform the benchmark. By one line of thinking, managers simply must own these stocks as insurance against poor relative performance.

At TSW we are happy to own the very largest companies at the right price—but not at any price. The considerable strengths of the giants do not mean that they will inevitably outperform small, mid-sized or even the large companies that don’t reach the mega-cap designation. This is especially true when the valuation disparity between mega-, large, and small stocks becomes very wide.

As value investors with a fundamental approach to stock selection, we invest where we are finding the most attractive investment opportunities, regardless—within practical limits—of size. Nevertheless, TSW’s non-U.S. equity portfolio historically has had a bias toward the smaller end of our investable universe.

In recent quarters we have continued to find more attractive risk/reward opportunities in the mid-cap range, edging our portfolio farther from the large end of the benchmark. As of the end of October, the market cap of our median stock holding had declined by just over 10% from the end of 2023. We aren’t predicting the downfall of mega-cap stocks, but rather we are taking advantage of the better value available in the lower market cap deciles of our large cap universe. Even if the biggest companies continue to leverage their advantages to grow, we believe valuation is the best indicator of long-term investment returns and we are excited by opportunities in a segment of the market that has struggled to keep up in the land of the giants.