Behemoth: How U.S. firms became profit giants—and why it may not last

One of the least contentious aspects of any definition of “U.S. exceptionalism” is the dominance of U.S. stocks in global equity markets. In fact, it is a well‑established fact: Over the past century, the U.S. has been the largest‑weighted country in global equity indices apart from a brief interlude in 1989 when Japan took the mantle. Over the last three decades, U.S. dominance of indices has accelerated; at the end of 2024, U.S. stocks made up 63.6% of the MSCI All Country World Index (ACWI), making it by far the largest country component, with Japan the next largest at 5%, followed by the UK at 3.4%.

U.S. stocks have shown superior performance for such a long time, and by such a large magnitude, that all other markets—in fact, all other asset classes—have been overrun by the behemoth that U.S. equities have become. This is neither serendipity nor accident—it has been driven by a quite remarkable rise in U.S. corporate profitability.

Below, we use the S&P 500 as a proxy for the U.S. corporate sector to document this rise, examine the factors behind it, question its sustainability, and discuss some of the investment implications arising from it.

A timeline of U.S. corporate profitability

Since the early 1990s, S&P 500 operating margins have trended up, with meaningful declines occurring only during recessions. The following timeline illustrates the key catalysts:

• 1989–1990s: Operating margins were relatively low at the end of the 1980s but grew steadily throughout the tech‑fueled 1990s bull market.
• 2000–2007: Margins peaked at 12.2% in 2000, dropping sharply after the dot‑com crash to around 9.1% during the early 2000s recession. Corporate profitability and margins subsequently recovered on the back of economic growth and financial engineering—reaching about 14.2% at the 2007 market peak.
• 2008 and 2009: The years dominated by the global financial crisis (GFC) and subsequent recession saw margins contract again, falling to 11.5% in 2009.
• 2010–2019: The years after the GFC saw ultralow interest rates. Operating margins improved lethargically at first but rebounded steadily with nearly every cycle, reaching 14.6% by 2019. The rise of high‑margin tech companies and widespread cost efficiencies were particularly supportive.
• 2020–present: This period was marked by the pandemic and subsequent recovery. Margins dipped briefly to 12.4% at the height of the pandemic in 2020 but recovered sharply, bolstered by the post‑pandemic earnings boom. They rose to 15.7% in 2022, then dropped slightly in 2023 before climbing again to a peak of 16.4% in September 2025.

Cycle to cycle, there has been a structural upshift in the level of market profitability. The key question is: Why?

One reason is lower rates and taxes. According to a 2023 paper by the Federal Reserve Board, from 1989–2019, real profit growth grew at a robust average rate of 3.8% per annum—almost double the rate seen from 1962–1989. Over that time, both interest rates and corporate tax rates declined substantially, significantly boosting corporate profit growth (Figure 2). The reduction in interest payments and corporate tax rates was responsible for more than 40% of the growth in real earnings over the period, the paper says. Or, to put it another way, the uplift in the rate of growth versus the earlier period was entirely due to interest and tax—in fact, earnings before interest and tax (EBIT) declined slightly between the two periods, from 2.4% in 1962–1989 to 2.2% in 1989–2019.

Another factor driving higher profitability has been the changing composition of the S&P 500: New entrants have come in at margins well above the index average, which has changed the math. This has occurred for a number of reasons, including the transition to a service economy, the upweighting of the technology sector, the rise of intangibles, and the proliferation of asset‑light business models. Collectively, these developments have reconfigured the S&P 500.

How the digital era has amplified scale advantage

Our integrated equity team has conducted some revealing research on the impact of these changing index dynamics. It found that the advent of the digital era appears to have brought significant advantages for larger firms, a more concentrated market and an attendant rise in pricing power—and consequently, rising profitability.

Using Organization for Economic Cooperation and Development (OECD) data, the team’s research shows that increasing industry concentration and market power has driven a tremendous separation of winners versus losers from digitalization’s three competitive shifting drivers:

i.  Network effects, both direct and indirect;

ii.  Economies of scale in data collection and analysis and, thanks to this information;

iii. High and increasing levels of price and product differentiation owing to the pervasive power of data analytics.

To summarize: The digital age has conferred greater advantages of scale to technology winners due to the benefits of platform economics and the flywheel of more data leading to better algorithms, and vice versa. These businesses have also benefited from being highly scalable with low marginal cost: While a widget manufacturer needs a costly new plant to increase production, a software provider may be able to add incremental revenue with little additional cost, enabling significant markups. Overall, the OECD data points to the rise of a cohort of elite, highly profitable, scaled businesses between 2001–2014—most of them in the U.S., where tech is so dominant (Figure 3). Our own analysis, looking at net profit margins, shows that the effect has continued from 2014 to the present (Figure 4).

The shift to services has driven up net margins

At the end of 2024, the S&P 500’s year‑end average net margin was 9.75% compared with one of 5.85% for the period 1989–2015. This improvement can be broken down into distinct fundamental and sector effects. Most of the increase can be attributed to globalization and the transition to a service‑based economy, as reflected in a 450‑basis‑point decrease in the cost of goods sold (COGS) (top chart of Figure 5). Low interest rates and a reduced U.S. corporate tax rate were additional tailwinds to margins, contributing nearly 130 basis points combined.

In the bottom chart of Figure 5, we look at margin expansion through the lens of index sector composition. Expansion in the three technology‑heavy sectors—information technology, communication services, and consumer discretionary—collectively account for over 75% of the nearly 400‑basis‑point improvement we observe in aggregate margin. Our data point to a systemic shift in the economy that has favored knowledge‑based industries and led to historically high margin levels.

The advantages of being “asset‑light”

So far, we have looked only at margins, which do not take account of the capital intensity of a business. But measures of “economic return”—the return of USD 1 invested in a company, sometimes expressed as “return on capital employed” or “economic value add”—tells a similar story. “Cash flow return on investment” (CFROI) is a financial metric that measures a company’s economic return by comparing its operating cash flow with cash invested—in other words, cash out over cash in, a flow compared with a flow. It also considers inflation in its calculations so that investment returns are measured in real terms and not distorted by historical cost accounting, write‑offs, or provisions.

The asset‑light nature (conversely higher asset turns) of technology and technology‑enabled businesses, combined with higher margins, means that the true economic return of investing USD 1 in the U.S. stock market has gone up in parallel with the upweighting in technology. This is illustrated by the significantly higher CFROI of tech firms versus non‑tech firms in recent years (Figure 6).

Further, in Figure 7, we show that the persistence of winners has consistently increased since 1990. Segregating the market by quintiles of profitability (in this case, return on equity (ROE)), the analysis shows that the top quintile remained in that cohort for longer periods in successive decades after 1990. The digital age is all about data, network effects, and scaled digital solutions. This has led to more persistent abnormal (high) growth rates, competitive advantages, and barriers to entry, which in turn have led to non‑mean‑reverting fundamentals.

Persistent abnormal returns distort markets

The economic and market implications are both direct and impactful. The U.S., with its continental scale, abundant data, culture of entrepreneurship, world‑class tertiary education and research institutions, access to risk‑taking venture and scale‑up capital, regulation (although this can be perceived as both a strength and a weakness), and lower cost of equity capital is a hotbed of innovation and commercialized technology. As long as the implicit return of USD 1 invested and reinvested in the U.S. market is superior, a higher multiple is justified. The behemoth will endure and U.S. exceptionalism, at least as it pertains to stock markets, will likely persist.

In capitalization‑weighted indices, the dominance of an elite, highly valued cohort of mega‑caps can have a distorting effect on the market aggregates. While no single valuation metric provides a reliable picture, the current environment is notable for the breadth of indicators showing a highly valued U.S. market:

• The cyclically adjusted Shiller S&P 500 price‑to‑earnings ratio currently hovers just under 40, more than twice its long‑term average and approaching the dot‑com bubble’s peak of 44.
• The narrower forward price‑to‑earnings ratio stands at 22, notably above its long‑term average, as does the price‑to‑sales ratio.
• Warren Buffett’s favored indicator—total market capitalization to gross domestic product—is at an all‑time high, and the U.S. dividend yield of the S&P 500 has slumped to just over 1%.

Meanwhile, the rise of intangibles not captured in balance sheets has almost confined Tobin’s Q ratio, which measures the ratio between a physical asset’s market value and its replacement value—to a relic of history (Figure 8).

Concentrated markets are one thing (the U.S. market is now at its most concentrated in 92 years); concentrated performance adds a very different dimension. The recent period has been highly unusual in that market performance has been highly concentrated in one geography (U.S.) and one sector (technology), and within technology, just seven companies.

It is not new for equity market returns to be driven by a relatively low number of companies—in fact, it has been well documented in Hendrik Bessembinder’s work on 90 years of U.S. stock market history.¹ But it is new for a cohort of rapidly growing and persistently profitable businesses to also be giant companies driving the majority of value created in the market.

One consequence of this has been conspicuous absence of the “small‑cap effect,” whereby small‑caps have historically outperformed large‑caps over long periods of time, justifying the illiquidity premium. This has reversed with the recent material underperformance of small‑caps—and it is a global phenomenon. While the growing advantages of scale and increasing persistence of winners may not wholly explain the recent underperformance of small‑caps, it is very likely one of causes.

What happens when mean reversion disappears

When mean‑reverting fundamentals fade, there are clear implications for style approaches such as relative value investing. Economic theory teaches us that competition drives abnormal returns toward the cost of capital, and value investors historically have benefited from understanding these “base rates” and not over‑extrapolating outsized strong or weak performance into the future. However, while this made sense historically, it turns out that, more recently, extrapolating winners has been the right thing to do. While past performance is no guarantee of future results, winners have continued to win, and losers have not been able to catch up to their scaled superiors. We quantify the diminishing returns to value as a factor over the recent decade (Figure 9).

This has challenged the modus operandi of stock pickers and asset allocators alike. As Sébastien Page, our head of Multi‑Asset, puts it in his paper, “When Valuations Fail”: The apparent disappearance of the value premium has challenged tactical asset allocators who seek to use relative valuation metrics to overweight cheap and underweight expensive asset classes.

 A repeat of the dot‑com bubble?

We have described how the economics of scale and competition have changed in the digital era, how the U.S. is fertile territory for the creation of scaled technology companies, and how this is increasing the persistency of market winners. The question now is: What, if anything, can change these dynamics?

Having invested in the dot‑com and artificial intelligence (AI) technology cycles, we understand why comparisons are often drawn between the two. In both eras, for example, the world understood that we were at the onset of a technology breakthrough that would revolutionize the way we work without knowing exactly how. In both, incumbent businesses realized they needed to adapt their business models or face existential challenges. The extensive buildout of infrastructure in the form of fiber cables and switches in the late 1990s and early 2000s bears a striking similarity to the AI‑driven heavy investment in silicon chips and data warehouses. Another common factor between the two periods is the presence of richly valued firms that are encouraged by investors to spend and invest as much as they can.

There are also clear differences, however. The dot‑com era, at least at its 1999–2000 dénouement, was led by unprofitable stocks. The market traded on fumes, and businesses were created and IPO’d in the wave of an oily rag. Investors bought into a “concept” or a “story,” while the rules of economics and financial discipline were parked. The focus was on pre‑monetization metrics such as “price to eyeballs.” Less price to earnings, more “price to yearnings.” In the final phase, the share prices went parabolic, but the apex was pointy and the descent equally steep. The net effect was misallocation of capital on a mass scale. Much of the capital expenditure had been financed by debt, and both credit and equity holders lost a lot of money.

We do not yet see capital misallocation on this scale in the AI era. Further, the capex boom is not (yet) being financed by credit but out of the cash flow generated by a set of highly profitable mega‑caps—the likes of which we have never seen—that can afford it. The inference from this is that, because this capex cycle is being funded from cash flow and the market is in “the more, the better” mode, this boom could go on for some time. Either that or we are at a similar point at which former Fed chair Alan Greenspan warned of frothy valuations in his “irrational exuberance” in 1996—when the irrationality did not fade for another four years.

The key to any change in this dynamic lies in the description “capital light.” When incumbent winners face a technological step change, they have no choice but to invest in that change. In game theory, this is the prisoner’s dilemma—and the sums are staggering. Individual companies’ capex line items in annual budgets now exceed USD 100 billion in some cases. At this stage, return on investment on these dollars will not, or cannot, be called in. The investment in computing power alone may generate a good return, but investment in adjacent assets such as dedicated electricity facilities is questionable and likely dilutive.

As David Giroux, portfolio manager and Chief Investment Officer of T. Rowe Price Investment Management, has explained, the consequences of poor capital allocation will become increasingly evident in a high‑stakes capex environment. He says: “Academic research highlights just how difficult and counterintuitive good capital allocation can be…the tendency to dial up these activities when times are good can make it harder to generate favorable returns.…Technological innovation is accelerating [and] stakes for capital allocation are high. Some companies will flourish because of these decisions, and some will falter.”

Historically, capital deployments of this magnitude failed to deliver the anticipated returns, and we do not believe this cycle will be an exception.

The word “bubble” needs to be used with caution as bubbles can only be validated with hindsight. Moreover, each new bubble varies in duration and amplitude. We are not yet convinced that we are in bubble territory given the fundamental strength of the companies leading the market. However, between the unprecedented capex and other flagposts of frothy investor behavior, we think investors should balance their enthusiasm with an awareness of how and when it might end.

As always, the dilemma is: Can you afford to be out of it and when do you get off? One of the great paradoxes of investing is that you can make the most money when others are misallocating capital. To paraphrase famous investor Sir John Templeton: “This time is no different, just not the same.”