Warren Buffett’s investment advice is putting retirement portfolios at risk

Few investors command more credibility than Warren Buffett. During the six decades he ran Berkshire Hathaway (NYSE:BRK-A | NYSE:BRK-B), shareholders enjoyed a cumulative return of roughly 6,099,294%, a 19.9% compound annual growth rate. The S&P 500 returned roughly 39,054% over the same period, or about 10.4% annually. When Buffett stepped down as CEO at the end of 2025, handing the reins to Greg Abel, he left behind the most remarkable compounding record in modern financial history.

Yet throughout that tenure, Buffett argued that most investors should avoid stock-picking altogether. At Berkshire's annual meetings and in shareholder letters, he repeatedly recommended a low-cost S&P 500 index fund as the best vehicle for most retirement savers. The irony is striking: the man who outperformed the market for six decades spent those same decades telling ordinary investors not to try.

Investors have listened, and the flows prove it. Vanguard's S&P 500 ETF (NYSEARCA:VOO) crossed a historic threshold on June 2, 2026, becoming the first exchange-traded fund in history to surpass $1 trillion in assets under management. But the very popularity of that advice is creating a risk that many retirement savers have not yet priced in.

Buffett's Advice Helped Create a Trillion-Dollar ETF

According to Reuters, VOO attracted $69 billion of net inflows in the first part of 2026, following $118 billion in 2024 and $138 billion in 2025. No ETF has pulled in more investor money over that stretch. The fund's low expense ratio of 0.03% has been central to its appeal, helping it overtake State Street's SPDR S&P 500 ETF (NYSEARCA:SPY) as the world's largest ETF.

"Buying the market" has also looked attractive on pure performance grounds. The S&P 500 completed nine consecutive weeks of gains through late May 2026, climbing to fresh all-time highs. The index gave back some of those gains as artificial intelligence stocks sold off following Broadcom's (NASDAQ:AVGO) disappointing earnings report, though the market remained up sharply for the year. Staying the course, in other words, has been rewarded.

That track record is precisely what makes the underlying risk so easy to miss. Investors believe they are getting broad diversification. The numbers tell a different story.

The Diversification Investors Think They Own

Investors buying VOO, SPY, or iShares Core S&P 500 ETF (NYSEARCA:IVV) typically believe they are spreading risk across 500 companies. In practice, the index is far more concentrated than that framing suggests. According to MacroMicro data, the 10 largest stocks in the S&P 500 represented 37.5% of the index's total market capitalization at the end of May 2026. That figure had pulled back from a record 43% reached in March, but it remains one of the highest readings in market history. By comparison, RBC Wealth Management data shows the top 10 accounted for roughly 41% of the index's weight at the end of 2025, more than double their share a decade earlier.

A decade ago, those same 10 stocks represented just 15.3% of the index's market cap. When Buffett told shareholders at Berkshire's annual meeting five years later that they would be better off in index funds than picking individual stocks, the figure had risen to 27.2%. The concentration has since climbed by nearly 38% beyond that point. In practical terms, as RBC Wealth Management put it, more than $40 of every $100 invested flows into just 10 companies, creating a feedback loop where passive inflows disproportionately support the largest stocks and reinforce their performance leadership regardless of underlying fundamentals.

Of those 10 stocks, seven are directly tied to the AI boom:

• Nvidia (NASDAQ:NVDA)
• Alphabet (NASDAQ:GOOG) and (NASDAQ:GOOGL)
• Microsoft (NASDAQ:MSFT)
• Amazon (NASDAQ:AMZN)
• Taiwan Semiconductor Manufacturing (NYSE:TSM)
• Broadcom

Investors buying index funds to reduce concentration risk are, in effect, making a large, correlated bet on the same companies driving AI enthusiasm and pushing valuations higher. Unlike past periods when the top 10 spanned unrelated industries, today's leaders are tightly linked by a shared theme: AI adoption and monetization.

What Happens If the AI Trade Slows?

Retirement investing is about protecting purchasing power across decades, not maximizing returns in any single year. That is where concentration becomes a genuine concern rather than an abstract one.

If AI spending continues expanding at its current pace, index investors will likely benefit alongside the megacap leaders. But if corporate AI budgets slow, data center spending moderates, or earnings growth misses expectations as occurred with Broadcom, the same stocks that powered the market higher would weigh heavily on index performance. J.P. Morgan Asset Management notes that the top 10 stocks currently trade at roughly 26x earnings, about 25% above their long-term averages, and that when valuations are priced for perfection, corrections tend to be felt more acutely in concentrated portfolios.

An index where 10 stocks account for more than one-third of its value carries a materially different risk profile than one where those same stocks represented 15% a decade ago. The Herfindahl-Hirschman Index, a standard measure of market concentration, stood 30% above its five-year average as of mid-2026 according to AhaSignals data based on SSGA holdings, meaning the index effectively behaves as though it contained only about 54 equally weighted stocks rather than 503.

Buying the market used to mean buying broad diversification. Today, it increasingly means making an outsized bet on a handful of technology leaders whose fortunes are tied to a single investment cycle.

Key Takeaway

Buffett's advice remains sound in principle. Index funds are low-cost, tax-efficient, and have outperformed the majority of active managers over long time horizons. But investors should recognize how much the market has changed since he first championed them.

VOO, SPY, and IVV remain useful vehicles. They no longer provide the same level of diversification they once did, however, because a small group of AI-driven giants now dominates the index in a way that has no modern precedent. The historical average concentration for the S&P 500's top 10 stocks hovered between 18% and 23% for most of market history from 1990 through 2015. Today's figure is roughly double that range.

Retirement investors should not assume that buying an S&P 500 fund automatically eliminates concentration risk. The data shows the opposite. Understanding what you actually own, not what the fund name implies you own, may be the most consequential retirement planning insight of this market cycle.

Editor's note: This article has been updated to reflect Warren Buffett's retirement as Berkshire Hathaway CEO at the end of 2025 and Greg Abel's appointment as his successor, to revise the Berkshire cumulative CAGR to 19.9% and the S&P 500 comparison return to roughly 39,054% per Berkshire's annual report data, and to add J.P. Morgan Asset Management and AhaSignals concentration data showing the top 10 S&P 500 stocks trade at roughly 26x earnings and the index now behaves statistically like a 54-stock portfolio.

Contact [email protected] for any questions or corrections.