Contributor: Lillian Wu

This article is the second installment in a three-part series adapted from “The Active Side of Indexing,” our 2026 contribution to the CFA Institute Research Foundation.

Great growth companies are not always great growth stocks, and vice versa. Artificial intelligence (AI) may change the world, but that does not mean the leading AI mega caps of today will remain dominant a few years hence, let alone deliver the best future return. By the time everyone agrees on the winner, the stock may already be everywhere, in every index, exchange-traded fund (ETF), model portfolio, and institutional benchmark. The business may be magnificent. And the stock market likely already knows it.

That is the broader lesson of market leadership. Investors assume the largest companies are the strongest growers because they are visible, liquid, richly valued, and institutionally blessed. But the historical record says otherwise. Over the last 35 years, the companies just below the top 500 in market value grew operating cash flow far faster than the giants above them, about 9% a year versus 6.5%. Their cash flow compounded to more than twice the level of the largest-cap cohort. Yet their return advantage was only modest, roughly 0.6% per year, with higher volatility and high tracking error. The giants did not keep up because they grew faster. They kept up because their comparable returns coincided with investors paying more and more for each dollar of their fundamentals.

In 1989, the large-cap valuation premium didn’t exist; it was a discount. Over the next 35 years, it averaged about 17%. By mid-2025, the top 500 sported a price-to-cash flow ratio 80% higher than the next 500. Relative returns followed valuation expansion rather than business growth.

We believe that is the trap. A high multiple can masquerade as growth. Index membership can masquerade as merit. Popularity can masquerade as progress. Cap weighting reinforces the loop: Rising prices create larger weights, larger weights attract more capital, and more capital appears to validate the high price with a rising price. But multiple expansion alone is not growth. We view it as a reflection of investors' expectations that future growth will justify today’s price.

A high multiple can masquerade as growth. Index membership can masquerade as merit. Popularity can masquerade as progress. Cap weighting reinforces the loop.

The better discipline may be to separate price-implied growth from realized fundamental growth. Do not ask only which companies the market has already crowned as winners. Also, ask whether sales, cash flow, dividends, and book value are truly growing, and identify where the market has not already charged full admission.

That brings us back to AI. The obvious mega-cap winners may still be great companies. But the same AI boom may reward a quieter cast of characters: the power behind the data center, the transformer at the substation, the switchgear, the cooling system, the backup generator, or the grid connection. It may even reward the smartest users of AI, in businesses that have nothing to do with AI technology.¹ Same trend. Very different price. Growth is not a big market value or a premium valuation multiple. Growth is growth in fundamentals, no matter what companies deliver it.

Membership Has Its Privileges

To determine whether membership in the market’s largest-capitalization tier is associated with more robust business growth, we compare two mechanically defined portfolios: the “TrueCap 500,” consisting of the 500 largest U.S. companies by market capitalization at each annual reconstitution, and the “Next 500,” those firms ranked immediately below them. Unlike committee-driven indexes, this framework isolates the effects of size-based inclusion itself. As a real-world reference point, we also compare these results with the S&P 500, whose membership substantially overlaps with the largest-cap cohort.

Exhibit 1 shows wealth accumulation and underlying operating cash flow growth for the TrueCap 500, the S&P 500, and the Next 500 from October 1989 through May 2025. The three have delivered similar returns, with the S&P 500 beating the TrueCap 500 by 0.1% per year, with modestly less risk.² The Next 500 beats the TrueCap 500 by 0.6% per year, a nice margin of victory, albeit with higher risk (17.5% vs. 14.9%) and lofty tracking error (7.4%). The performance of the TrueCap 500 and the S&P 500 have been converging toward the Next 500 for over a decade, so large-cap names have been outpacing the Next 500.

Stocks are not businesses. So the more important question is, Did the largest companies earn those returns by growing faster? Not necessarily. Rising valuation multiples, not faster growth in the fundamentals, likely explains this performance. Indeed, the Next 500 has grown cash flow faster than the TrueCap 500, even during a decade-long rout for small-cap and value stocks.

The lower lines in Exhibit 1 track the growth of operating cash flow for the companies in each portfolio. Starting from similar base levels in 1989, the operating cash flow associated with a hypothetical $100 investment in the Next 500 grew to approximately $464 by 2025, compared with roughly $203 for the TrueCap 500 and $217 for the S&P 500. That translates into near 9% annualized cash flow growth for the Next 500 versus roughly 6.5% for the largest-cap cohort. In the 35-plus years of our sample, that difference compounds into a very large gap in business scale. Even during the last decade, a dismal one for the Next 500 relative to the top 500, the Next 500 still grew cash flow 2.1% faster than the top 500.

This is the puzzle. The companies just outside the largest-cap tier grew far faster, but their shareholders earned only modestly higher returns. The largest companies grew slower, yet their stocks kept pace remarkably well. Market leadership, at least in this sample, was not proof of superior business growth. It was proof of something else.

That “something else” is where membership begins to matter. The market’s largest companies are not merely companies. They are benchmark constituents, liquidity vehicles, institutional comfort food, and default destinations for vast pools of passive capital. Exhibit 1 does not yet tell us how much those privileges are worth. It does show why we should ask.

When Valuation Outruns Fundamentals

To understand the forces driving these differences, Exhibit 2 compares the relative valuations, relative wealth accumulation, and relative cash flow growth of the large-cap cohort and the Next 500. The green lines show relative valuation multiples (TrueCap 500 vs. Next 500, and S&P 500 vs. Next 500).³ The blue lines track the relative wealth of a $100 investment starting in late 1989 with dividends reinvested, while the red lines show relative cash flow growth. We measure cash flow using five-year moving averages, similar to the cyclically adjusted price-to-earnings ratio (CAPE) developed by Robert Shiller, to smooth economic cycles.

Several patterns stand out. First, valuation differences have been substantial and highly time varying. While relative valuations favored the smaller-cap names in the Next 500 in 1989, over the full 35-year sample, the largest-cap cohort traded at notably higher multiples, with an average premium of about 17% over the Next 500. The top 500 are ostensibly the winners, the towering giants of the U.S. economy. The Next 500 are the also-rans, many ostensibly struggling to turn a profit if not find a path to greatness. The underlying profit growth – slower in the top 500 than in the next 500 – puts the lie to this popular narrative. This premium expanded markedly with the dot-com bubble in the late 1990s, compressed after the burst, and remained subdued for an extended period before widening again in recent years, reaching an unprecedented 80% by mid-2025.

Second, the relationship between valuation and realized business growth shifted over time. Prior to the dot-com bubble, the relative performance and relative valuation multiples of the top 500 and the next 500 moved in near lockstep. In the early 1990s, the largest-cap cohort showed slightly stronger cash flow growth (a rising red line), even with valuations that started the decade at a modest discount. In the late 1990s, top 500 valuations and performance surged to a roughly 40% premium compared to the next 500, before both the premium and the growth advantage faded after the technology bust. That 40% outperformance was almost entirely due to revaluation, not superior growth in profits. By 2011, the largest-cap cohort again traded at a 10% to 15% discount, while relative cash flow growth stayed muted for much of the following decade.

The more recent period marks a notable break from this earlier pattern. The expansion in valuation multiples has not been accompanied by stronger underlying growth, not even recently. The renewed dominance of mega-cap stocks therefore likely reflects rising valuation multiples paired with slower business performance. Some may dispute this, but we can see no other driver for such odd and profoundly inefficient market behavior than the immense flow of capital into index funds.

Third, relative investor wealth closely tracked relative valuation rather than relative fundamentals. Periods of large-cap outperformance generally coincided with expanding valuation premiums, while periods of relative underperformance aligned with valuation compression. Though the wedge between the blue and green lines expands and contracts over time, the short-term behavior is inextricably linked. Relative performance therefore appears to have been driven less by differences in realized business growth than by investors’ willingness to pay ever-higher multiples for the market’s largest companies.

This distinction has important implications for both indexing and growth investing. Modern markets frequently treat high valuation multiples as evidence of robust future growth. Our findings suggest they’re not. Rising prices confirm rising growth expectations. In an efficient market, these rising expectations would be just as likely to underestimate as to overestimate future growth. In practice, the correlation between price-implied growth and the subsequent reality is surprisingly modest. The historical record suggests a far weaker relationship. Our evidence shows that valuation expansion and realized business growth can diverge substantially over very long spans.

Modern markets frequently treat high valuation multiples as evidence of superior future growth. They’re not. Rising prices confirm rising growth expectations.

Why Membership Matters

Why does this gap between valuation and realized growth persist? Part of the explanation may well lie in the mechanics of modern indexing itself. Cap-weighted indexes systematically allocate more capital to companies whose prices have already risen enough to place them at the top of the market-cap hierarchy. As passive investing has increased, membership in the largest-cap cohort has become economically meaningful in its own right.

Inclusion in major indexes now brings visibility, liquidity, benchmark demand, and persistent passive ownership flows. These forces can reinforce elevated valuations even when the underlying business fundamentals fail to keep pace. Membership carries privileges, specifically a right to carry a higher valuation multiple and a right to upward pressure on valuation multiples every time an investor puts money into an index fund. Those privileges stem less from superior growth than from the structural dynamics surrounding index inclusion itself.

Consider that the S&P 500 currently spans about 80% of U.S. stock market capitalization, and owns about one-fourth of the market value of the member companies. This means that every $100 invested in an S&P 500 index fund requires selling an average of $20 of non-S&P companies, to buy $20 of S&P member stocks.⁴ In an efficient market, this should not matter even one iota to the pricing of members or non-members. In the real world, it’s hard to dismiss the idea that the S&P 500’s outperformance relative to the Next 500 may be a direct consequence of money pouring into index funds, especially when the cash flow profits of the S&P 500 have lagged those of the Next 500.

This dynamic may also help explain why valuation expansion has increasingly dominated realized business growth in driving relative returns. Higher prices attract greater index weight and additional passive capital, reinforcing market leadership even after the underlying growth advantage has faded. At the margin, the reconstitution process amplifies these effects: Companies typically enter the largest-cap cohort after prolonged price appreciation and exit after extended weakness, embedding a momentum-like migration process within cap-weighted indexing.

None of this implies that the market’s largest companies are poor businesses. Many are extraordinarily profitable and innovative enterprises. But the evidence suggests that their valuation premiums reflect market structure and the flow of capital into index funds more than underlying fundamentals. In modern markets, index membership itself has become part of the investment story.

Rethinking “Growth”

The evidence presented so far challenges one of the market’s most expensive habits: treating a high price as proof of robust future growth. Historically, that assumption has been unreliable at best. The market’s largest and most richly valued companies have often delivered slower realized fundamental growth than the firms immediately below them. Yet investors continued to reward them with higher and higher multiples. The market did not merely pay for growth. It often paid for past recognition.

Today’s AI boom offers a useful live example. NVIDIA may be a magnificent business; it is also an obvious shorthand for AI. The harder question is not whether AI is real but whether the obvious AI winners already charge too much admission. The same boom may also feed less celebrated parts of the ecosystem: power management, thermal systems, substations, grid construction, backup generation, and interconnection, as well as the vast array of eventual AI users and customers, often well outside of the tech arena. Think Nvidia for the headline, Vertiv for critical power and cooling, Eaton for power management, and Quanta for the electric infrastructure beneath it all. Same growth story. Different place in the value chain. Different price.

This disconnect reveals a potentially crucial distinction between price-implied growth and realized fundamental growth. Price-implied growth asks what investors are already willing to pay. Realized fundamental growth asks what the business actually delivers. The two can be very different. Prices can reflect benchmark demand, passive flows, investor enthusiasm, career risk, and extrapolation of recent success. A stock can rise because the business is improving or because the crowd is more willing to pay for a consensus expectation that is amply reflected in the share price.

Exhibit 3 illustrates this point directly. Companies sorted on realized fundamental growth, measured using subsequent growth in sales, profits, or R&D spending have outperformed portfolios constructed using traditional price-based growth definitions. The distinction is subtle but important. One approach assumes that high valuations forecast future growth. The other assumes that past growth forecasts future growth. In an efficient market, both may be true, but high valuations must be every bit as good as past growth in forecasting future growth. The market is self-evidently not efficient: We believe valuation multiples predict future growth poorly, while observed past growth forecasts future growth rather nicely.

This is why the AI example matters. The investor’s task is not to deny the obvious winner but to ask whether the obvious winner still represents the best use of capital. Sometimes it will. Often, the better option is one step removed from the headline: the supplier, the bottleneck, the capacity constraint, the unglamorous enabler, or even a new technology’s happy end customer.

The difference is material because markets are noisy, especially over shorter horizons. Prices can move far ahead of business fundamentals, particularly when narratives, momentum, or benchmark flows dominate investor behavior. By anchoring growth measurement to realized business expansion rather than price-implied expectations, investors can separate durable economic growth from temporary valuation enthusiasm.

By anchoring growth measurement to realized business expansion rather than price-implied expectations, investors can separate durable economic growth from temporary valuation enthusiasm.

One shocking takeaway: As in Exhibit 2, where the top 500 and next 500 track almost identically until 1999, Exhibit 3 shows the 500 highest valuation multiple stocks and the 500 fastest prior growth stocks follow a similar pattern. What changed? The indexing hockey stick went parabolic around the turn of the century. Indexing was about 3% of U.S. market cap in 1990, 10% in 2000, 25% in 2010, and more than 50% by 2025. This is not an argument against indexing, large-cap investing, or growth investing – all serve their purpose and all will have their days in the sun – but an argument against incomplete definitions. Growth is not a high multiple, a rising price, or index membership. Growth is the expansion of the underlying business.

Once indexes became investment vehicles rather than mere market barometers, price and popularity began to shape outcomes. Rising prices create larger weights, larger weights attract more capital, and more capital can reinforce the rising price. That loop may reward investors handsomely for a time. But it should not be confused with business growth. The active opportunity is to break that confusion, to look past the crowned winners, and ask where fundamentals are growing faster than expectations.

Once indexes became investment vehicles rather than mere market barometers, price and popularity began to shape outcomes.

Membership has its privileges, a higher valuation multiple – arguably directly due to inclusion in an index – and a steady flow of new investment capital that can propel those multiples higher. But the flip side should never be overlooked: Faster future fundamental growth, paired with a lower starting valuation multiple has not historically failed to deliver a higher internal rate of return (IRR) to that vanishing breed, the patient long-term investor.

End Notes

1. A fun example is Ralph Lauren (RL), which added an “Ask Ralph” bot to its website that gives you fashion advice directly from the master, or at least an AI interpretation of him. In the three years ended April 30, 2026, RL has delivered a total return of 228%, far ahead of the 80% and 58% return respectively, of the SPY S&P 500 ETF and the IWD Russell 1000 Value ETF. For each technological innovation, there must be happy customers, monetizing their use of the technology.

2. The S&P 500 has also done better than the TrueCap 500 at identifying growing companies, with 0.20%-faster annual cash flow growth. The S&P Index Committee’s subjective judgment would appear to add modest value, not just by excluding stocks that may be mere froth due to a lack of demonstrated earnings, but also by choosing those with, on average, slightly better performance. In a market where individual basis points (bps) matter, an annual return difference of 13 bps is economically meaningful, even if it lacks statistical significance, leaving investors nearly 5% wealthier after 36 years.

3. Some readers might wonder whether we compared the S&P 500 with its own Next 500, in other words, the 500 largest market-cap names that are not in the S&P 500. We did and realized the results were too strange to merit inclusion in this paper. On average, 120 TrueCap 500 names are not in the S&P 500, and vice versa. So, an “S&P Next 500” will include 120 large-cap names, on average, that are not in the S&P 500, which will utterly dominate the performance and relative valuation metrics for a Next 500 index. For example, when Tesla joined the S&P 500 in late 2020, it was among the 10 most valuable stocks in the world. Prior to Tesla’s move into the S&P 500, an S&P Next 500 might have been described as “Tesla and the 499 dwarfs.”

4. The cliché is that prices go up because there are more buyers than sellers. This is self-evidently incorrect, as every buyer must find a seller, and vice versa. But the cliché has merit. Prices go up because of more eager buyers than sellers. For every $100 added to an S&P 500 index fund, the resulting $20 of index member purchases must find a willing seller from the diminishing roster of non-index-fund investors. And the $20 of non-member sales must find a willing buyer, also from the non-index-fund investors. This drives the prices of the former higher and the prices of the latter lower.