On Wednesday, Robert Armstrong, a columnist for the Financial Times, published an in-depth article examining the high concentration of the US stock market in the hands of a few tech giants, debating whether this concentration genuinely signals increased market risk. Armstrong began by depicting a dramatic market day: following the announcement of a new legal work automation tool by AI firm Anthropic, shares of software and professional services companies plummeted by over 3%. However, the tech behemoths were not spared either, while traditional economic sectors like energy, telecommunications, consumer staples, and materials emerged as the winners. This example illustrates a shifting market landscape. Armstrong's central argument is that market concentration itself is not the primary concern; the real danger lies in excessive valuations. Although the Magnificent 7 tech stocks no longer dominate the market as they did a few months ago, the concentration within the S&P 500 index remains remarkably high. More importantly, historical data reveals that such concentration is not an anomaly, with a similar scenario having occurred back in 1932.

Market Concentration: Six Companies Account for One-Third

Just how concentrated is the current US stock market? Armstrong provides detailed data. A mere six companies account for one-third of the S&P 500's total market capitalization, with NVIDIA alone representing 7%. Looking at the top 62 largest companies, they collectively constitute two-thirds of the entire index's market value. From a net profit perspective, the situation differs slightly but is similar in essence. Those six giants contribute 27% of the net profit, while the top 62 companies account for 63%. Armstrong points out that this implies the largest companies typically possess "higher valuations—price-to-earnings ratios" compared to their smaller counterparts. This level of concentration raises a question many analysts are asking: does extreme market cap concentration make the market more precarious? Should investors reduce their exposure to the largest companies or sectors relative to their market weight?

History Tells Us: Concentration is the Norm

Armstrong highlights that a research paper by scholars Per Bye, Jens Soerlie Kvaerner, and Bas Werker provides compelling historical evidence. They analyzed data from all companies traded on major US exchanges since 1926, including market capitalization and various profitability metrics. The findings are quite surprising. The paper's authors wrote:

The relative importance of the Mag7 is not particularly unusual. For instance, from the 1930s to the 1960s, seven companies held a similar share. "The peak occurred in May 1932, when AT&T, Standard Oil, New York Consolidated Gas, General Motors, DuPont, Reynolds Tobacco, and United Gas Improvement Company collectively accounted for approximately one-third of the total market capitalization.

This indicates that market concentration is not a phenomenon unique to the modern era or a special product of the tech age. The same pattern of concentration existed during the industrial era. In other words, a market dominated by a handful of large companies appears to be a常态 of capital markets. The research also uncovered a significant finding: company fundamentals (such as revenue and profits) do tend to follow changes in market cap concentration, but this correlation is loose and exhibits cyclical fluctuations. The paper further noted that when market cap concentration reaches extreme levels (i.e., the number of companies comprising the top third of total market cap hits a historical low, currently below 0.5%), the share of total revenue, profits, and cash flow accounted for by these largest companies is actually at a historically low level, around one-fifth.

So, does high concentration also predict low long-term returns? The paper states:

In isolation, market concentration does have a negative predictive effect on returns (high concentration, low subsequent returns), but this is not the case when controlling for valuation... Holding valuation constant, higher concentration is actually associated with higher future returns!

Mathematical Model Confirms: Concentration is a Natural Outcome

The second part of the paper uses a mathematical model to further bolster the argument that market concentration is a normal phenomenon. The study employed a model "based on a standard geometric Brownian motion diffusion process," incorporating "common market factors and firm-specific shocks." Specifically, in this model, returns follow a stochastic process. One can imagine a company's market capitalization as being constantly subjected to various shocks—productivity and innovation shocks, good or bad leadership, luck and misfortune. Over time, most companies remain small (their positive and negative shocks cancel each other out), while a few companies receive a multitude of positive shocks and grow into giants. Using an analogy, the stock market is like an ongoing dice-rolling game. Most players' results will tend toward the average, but there will always be a few players with exceptionally good luck, rolling high numbers consecutively. This is not surprising but a natural outcome of probability. By the same logic, market concentration is a natural result of the market mechanism's ordinary functioning.