American markets have experienced concentrated periods before — the 1920s, the Nifty Fifty, the dot-com peak. Each ended in a way that reshaped what "diversified" meant for a generation. Cook Pierce examines what the historical parallels reveal and where 2026 is genuinely different.
American markets have been here before. Not exactly here, but close enough that the historical parallels are worth understanding.
Three prior periods stand out for concentration at the top of the market. Each ended in a way that reshaped what diversification meant for a generation of investors. And each is different from the current moment in ways that are worth naming precisely.
By 1929, roughly a dozen holding companies and industrial trusts dominated American markets. Utilities, railroads, and early consumer products giants like RCA and General Motors accounted for a disproportionate share of total market value. The concentration was built on genuine industrial transformation — electrification, automobiles, radio — combined with speculative capital flows amplified by leverage.
When the crash came, the concentrated top absorbed the largest losses. RCA fell 98 percent from peak to trough. Broad diversification did not protect investors because "broad" at the time was itself concentrated. The 1930s reshaped what American investors understood about the difference between owning many companies and owning genuinely diversified risk.
In the early 1970s a different form of concentration took hold. The "Nifty Fifty" — Xerox, Polaroid, Kodak, IBM, Avon, McDonald's, and roughly forty other blue-chip American firms — came to be seen as one-decision stocks. Buy them and hold them and the growth would continue. The concentration was based on genuine earnings quality and defensible business franchises.
By 1972 these fifty stocks traded at price-to-earnings ratios of 40 to 90 times, while the broader market traded at 15 times. The concentration was in valuation rather than in market cap weighting alone, but the effect on portfolios that held them heavily was the same. Between 1973 and 1974 the Nifty Fifty fell dramatically, and many of the individual names never fully recovered their prior positions. The lesson repeated: high concentration in what looked like the safest holdings meant the largest exposure when the valuations adjusted.
By March 2000, technology stocks made up nearly 33 percent of the S&P 500 by weight — up from about 6 percent a decade earlier. The concentration was built on the genuine transformation of the internet economy combined with speculative capital flows into companies whose fundamentals did not yet justify their valuations.
The subsequent unwind cost the Nasdaq more than 75 percent of its value from peak to trough. Investors who believed their index fund was diversified discovered that index weighting had concentrated their exposure precisely where the correction happened. It took the Nasdaq fifteen years to reclaim its 2000 peak.
Today's concentration exceeds each of the prior peaks in absolute terms. The top ten percent of American listed companies now hold more than three-quarters of total market capitalization — higher than 1929, higher than 1973, higher than 2000.
But there are two differences that matter.
The first is that the largest firms today are, by most measures, extraordinarily profitable. Alphabet, Amazon, Meta, Microsoft, and Nvidia are not the RCAs of 1929 or the pets-dot-coms of 2000. They generate significant real cash flows from operating businesses whose customers pay real money for real services. The concentration is built on more genuine economic activity than the prior peaks were.
The second is that the largest firms are borrowing more to fund their capital expenditure than they used to. Amazon's ratio of debt to annual free cash flow has moved from below the S&P 500 average in 2019 to more than four times that average today. The AI infrastructure buildout is being paid for increasingly with borrowed money against future revenue projections rather than with cash on hand.
Together these differences suggest that the current concentration is neither as fragile as 1929 nor as pure a speculation as 2000. But it is also not immune to the pattern that all three prior periods established: when concentration reaches an extreme, the eventual adjustment is proportional to the concentration itself.
None of the prior concentrated periods ended because the concentration was identified in advance as unsustainable. They ended because specific events — a Fed decision, a credit shock, a business failure at the top — triggered a repricing that the market's structure amplified.
For a well-ordered retirement plan, the historical pattern does not require predicting when or how the current concentration adjusts. It requires holding the market-based portion in a position where the adjustment, whenever it comes, does not force a change in the retirement itself. That position is Surplus. When market-based assets carry the weight of retirement income directly, concentration at the top of the market becomes concentration at the base of the retirement plan.
The three prior periods each rewrote what diversification meant for the generation that lived through the correction. The plan built to survive the next such rewrite is not the one that predicts it. It is the one that does not need to.
