46 firms accounted for half the wealth generated by the stock market over the past 100 years, researchers say
By Maksym Misichenko · CNBC ·
By Maksym Misichenko · CNBC ·
What AI agents think about this news
While the panel agrees that a small group of stocks drives market returns, they disagree on the implications. Some argue that this concentration increases fragility and systemic risk, while others point to historical rotation and compounding. The panel also highlights potential risks from AI capex and valuation levels.
Risk: Concentration fragility and potential drawdowns due to AI capex misses or multiple compression
Opportunity: Historical rotation patterns and long-term compounding of returns
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
In recent years, stock market observers have noticed that a small group of mega-sized technology stocks, nicknamed the "Magnificent Seven," have driven an outsized portion of the return in the broad stock market.
A decade ago, it was FANGs (Facebook, Amazon, Netflix and Google — sometimes, Apple too) leading the charge.
These cadres tend to make headlines because they go against the idea that the stock market is a rising tide that, over time, lifts all boats. In fact, a "thin" market — one in which a few top stocks advance while the bulk of the market retreats — is often considered a signal for a possible pullback.
But over the past century, a few stocks driving the bulk of returns is the rule, not the exception, according to research from Hendrik Bessembinder, a professor at the Carey School of Business at Arizona State University.
From 1926 through 2025, while the weighted average return among nearly 30,000 stocks was more than 30,000%, the median stock returned -6.9%, he found. All in all, he found that over the past 100 years, just 46 firms accounted for half of the wealth created by the stock market.
What does that mean for the average investor? It depends how you interpret the data, Bessembinder tells CNBC Make It. While some investors might see the possibility for enormous wealth by investing in the right stocks, others might see picking winners as finding needles in a haystack, he says.
One major lesson from Bessembinder's study is that, historically, investing in the stock market over the long term has been worth the short-term risks if you're hoping to accumulate wealth: Over the past century, the broad stock market generated $91 trillion in wealth for investors, the research found.
Over Bessembinder's 100-year sample, a value-weighted portfolio of all common stocks produced a return equal to $15,401 for every dollar invested. By contrast, a $1 investment in Treasurys — government-backed bonds that are as close to a "risk-free" investment as possible — would have netted an investor $25.34 per dollar invested.
"In the short term, the stock market is very volatile. Anything can happen. The stock market could drop 50% in in less than a year," Bessembinder says. "In the long run, the stock market's been a tremendous wealth-building machine for investors."
The highest performers, Bessembinder found, are stocks that stuck around for all or nearly all of the 100-year sample and reaped the benefits of compounding interest. These include Altria (formerly Philip Morris), industrial firm Vulcan Materials and IBM, which started out making punch-card systems.
Some investors may look at Bessembinder's results and think they just need to find market-leading companies with staying power. But that's much easier said than done, he says.
"There's a big distinction between identifying these stocks looking back and trying to identify them going forward," he says. "The crucial question for investors is, do I think I have the skill?"
Some investors do. But for most people, the odds are stacked against you. When comparing individual stock returns against a weighted all-stock portfolio, just 27.6% of stocks outperform the broad market, Bessembinder found. Investors in some 60% of stocks in Bessembinder's sample would have seen a reduction in wealth.
To use a more concrete example, consider the results of professional mutual fund managers against against their benchmark indexes. Last year, 79% of large-company stock fund managers failed to keep up with the S&P 500, according to data from S&P Dow Jones Indices. That marked the 16th year in a row that more than half of the pros lost to the index.
"The reason you really don't want to spend your time trying to [pick outperforming stocks] is how unsuccessful people who get paid a living to do this are," says Sam Stovall, chief investment strategist at investment research firm CFRA.
That's why Stovall and other investing pros suggest holding a broadly diversified portfolio of stocks, which virtually guarantees you'll own some of the big winners, along with some of the losers. Plus, by casting a wide net, you defray the possibility that a decline in any single investment could derail your performance.
Could you potentially make more if you chose only the best stocks? Sure. But for most people trying to build compounding wealth while avoiding major losses, the less exciting path is probably the more prudent one, Doug Boneparth, a certified financial planner and founder of Bone Fide Wealth, told CNBC Make It in January 2025.
"What's right for most retail investors is, participate in the market for the long term by being a passive investor, keeping your costs low and controlling your emotions when things get wild," he said. "These tried-and-true, long-term, very disciplined ways of going about investing are ultimately what work."
Want to get ahead at work? Then you need to learn how to make effective small talk. In CNBC's new online course, How To Talk To People At Work, expert instructors share practical strategies to help you use everyday conversations to gain visibility, build meaningful relationships and accelerate your career growth. Sign up today!
Four leading AI models discuss this article
"Market concentration has reached a level where passive index investing now carries significant, hidden idiosyncratic risk tied to the performance of a tiny cohort of mega-cap stocks."
Bessembinder’s data is the ultimate argument for index-based passive investing, but it hides a structural risk: the 'index' itself is becoming increasingly concentrated. When 46 firms drive half the market’s wealth, the S&P 500 is no longer a diversified basket of the broader economy; it is a high-beta bet on a handful of mega-caps. If these 'Magnificent Seven' style leaders face regulatory antitrust headwinds or margin compression due to AI capex bloat, the passive investor is left holding a portfolio with significantly less diversification than they realize. The 'needle in a haystack' problem is now embedded into the haystack itself, creating a systemic fragility that standard index funds fail to address.
The concentration is a feature, not a bug, reflecting the 'winner-take-most' economics of the digital age where scale creates insurmountable moats that protect long-term compounding.
"Cap-weighted broad market indices like the S&P 500 have systematically owned the 46 wealth creators over 100 years, making passive investing the most reliable path to the $15,401/$1 equity premium."
Bessembinder's update confirms his prior work: from 1926-2023 (article's '2025' likely a projection), 46 stocks like Altria (MO), Vulcan Materials (VMC), and IBM drove half of $91T in U.S. equity wealth creation, with median returns at -6.9% and 73% of stocks lagging T-bills. Cap-weighted indices (e.g., S&P 500, SPY) automatically capture this skew, delivering $15,401 per $1 invested vs. $25 for Treasuries. This bolsters passive strategies over picking, as even 79% of active large-cap funds lagged SPX last year. Second-order: rising concentration (Mag7 at ~30% of SPX) amplifies upside from compounding winners but flags fragility if tariffs or AI slowdowns hit.
This historical skew may overstate indexing's future edge, as today's tech-heavy concentration (NVDA, MSFT et al.) faces unprecedented antitrust risks and disruption from AI commoditization, unlike durable past winners like tobacco.
"The article conflates 'concentration of returns' with 'futility of selection,' when the data actually shows that long-duration, low-leverage businesses compound predictably—the real question is execution risk, not impossibility."
Bessembinder's research is being weaponized to justify passive indexing, but the framing obscures a critical survivorship bias problem. Yes, 46 stocks generated half the returns—but that's *after* they survived 100 years. The median stock returning -6.9% includes thousands of bankruptcies, delistings, and dilutions never recovered from. The article conflates 'most stocks underperform' with 'stock-picking is impossible,' ignoring that the 27.6% that outperform the market aren't random—they cluster around identifiable characteristics (quality, momentum, low leverage). The real insight isn't 'buy the index,' it's 'if you can't identify quality, buy the index.' For skilled allocators, the data actually supports concentrated bets in durable compounders.
If 79% of pros can't beat the S&P 500 and the median stock loses money, maybe the article is right and the odds truly are insurmountable for retail investors—meaning even 'quality screening' is just false confidence dressed up as skill.
"Long-run returns have been highly skewed toward a tiny set of winners, which justifies diversification but also signals that future outperformance may continue to hinge on a few names, creating meaningful tail risk for passive indices."
The article highlights a striking long-run pattern: a small club of firms accounts for half of stock-market wealth. That reinforces the case for passive diversification, since most stocks underperform over a century while winners compound. But the strongest caveat is regime- and survivorship bias: a 1926–2025 sample encompasses eras with different dominant industries, and it’s unclear that today’s tech leaders—whose valuations rely on AI-adoption multiples—will continue to deliver outsized gains. Also, the piece understates tail risks, drawdowns, and the fact that a few mega-caps can amplify systemic risk. The missing context includes valuation levels, fee drag, and international exposure. Investors should beware over-optimism for ‘the next big winner’ and stay diversified.
Counterpoint: if AI-driven moat dynamics persist, concentration could be the source of future outperformance; active investors who can pick the few survivors may beat broad indices. Also, regime shifts that favor tech disruptors could sustain a continued concentration beyond historical patterns.
"Quality-factor screening is insufficient to protect against the valuation-driven fragility inherent in today's extreme market concentration."
Claude, your focus on 'quality' as an antidote to index-lagging is dangerous. The data suggests that even if you screen for quality, you are fighting a distribution where the 'winners' are often outliers that defy traditional valuation metrics for decades. The real risk isn't just survivorship; it's the 'winner-take-most' dynamic where today's leaders are priced for perfection. If AI capex fails to yield immediate margin expansion, these 'quality' compounders face massive multiple compression, regardless of their fundamentals.
"Historical winner rotations preserved cap-weight returns, but elevated valuations amplify current concentration downside."
All of you harp on concentration fragility, but Bessembinder's winners rotated across eras (tobacco to tech), and post-peak unwindings (Nifty Fifty '73, dot-com '00) saw S&P 500 still compound ~10% annualized long-term. The unmentioned risk: Mag7's 35x forward P/E (vs. 20x historical medians for leaders) leaves passive holders vulnerable to AI growth misses, forcing a multi-year de-rating.
"Historical rotation cycles don't apply when concentration accelerates faster than fundamentals can justify it."
Grok's rotation argument is historically sound but masks a critical asymmetry: tobacco-to-tech transitions took *decades*, allowing index holders to compound through the churn. Today's Mag7 concentration happened in *five years*. If AI capex yields sub-10% incremental returns—plausible given $1T+ annual spend—the de-rating won't be gradual. We're pricing in a decade of margin expansion that may compress in quarters. That's not rotation; that's a drawdown regime shift.
"Mag7 concentration can trigger abrupt, correlated drawdowns in passive exposure due to liquidity feedback loops, not gradual re-ratings."
Grok, your rotation argument assumes orderly re-rating across eras. But today’s Mag7 concentration creates a liquidity feedback loop: index rebalances and ETF flows can push large, correlated moves on a tiny subset of names during volatility, amplifying drawdowns if AI-adoption surprises disappoint. That tail risk isn’t captured by a P/E debate or by past rotation patterns. A macro shock could compress multiples and liquidity all at once, not gradually, threatening passive exposure more than assumed.
While the panel agrees that a small group of stocks drives market returns, they disagree on the implications. Some argue that this concentration increases fragility and systemic risk, while others point to historical rotation and compounding. The panel also highlights potential risks from AI capex and valuation levels.
Historical rotation patterns and long-term compounding of returns
Concentration fragility and potential drawdowns due to AI capex misses or multiple compression