As we celebrate the four-year anniversary of the current bull market, the strength of the broader market masks a more complicated picture that has become much more pronounced over the past year. While major indices have so far delivered another year of above-average returns, significant shifts in leadership have created a much more uneven experience for investors. As capital has flowed toward artificial intelligence (AI) winners and investors have reassessed the companies most vulnerable to disruption, four out of every ten stocks in the S&P 500 actually posted negative returns over the 12 months ending September 30, 2026.

This has been a particularly challenging environment for stock-pickers. Investors with concentrated positions have likely underperformed the S&P 500 this year, too. Only about one-third of S&P 500 companies outperformed the index, reflecting unusually narrow market leadership. Perhaps more surprising, several members of the much-celebrated “Magnificent Seven,” including Microsoft and Meta, and even beloved Berkshire Hathaway, ended the past 12 months with negative returns.

Looking more closely at this year’s winners, investors have favored cyclical companies delivering outsized earnings growth, much of it driven by a supply-and-demand imbalance across the semiconductor supply chain. Tight supply has given these companies significant pricing power, pushing prices and profitability to historically high levels. However, history suggests that these conditions may not last. As new manufacturing capacity comes online and supply begins to catch up with demand, prices can often fall sharply. This, in turn, will put pressure on profit margins. The timing of this likely outcome is unknown and remains a significant source of debate given the ongoing shortage of computing capacity and the substantial financial resources of the largest companies investing in AI infrastructure.

As the AI cycle matures, though, we believe market leadership should broaden. The next phase of the opportunity may shift towards companies that successfully use AI to improve productivity, reduce costs, and increase revenue. As these benefits become more visible in corporate earnings, leadership could expand beyond today’s infrastructure winners to other businesses across a much wider range of industries. Not every company investing in AI will generate attractive returns, but this evolution could create new opportunities for investors willing to look beyond the market’s current leaders.

These potential shifts also underscore the importance of diversifying concentrated stock positions over time. Investors who have benefited from owning a handful of exceptional performers can find themselves increasingly exposed to the fortunes of a single company, industry, or investment theme. While reducing a concentrated position can be difficult, particularly when the tax consequences are significant, a thoughtful, gradual approach to diversification can help manage risk while preserving participation in future market gains. The goal is not to predict when today’s leaders will falter, but to build portfolios positioned to participate in tomorrow’s opportunities while limiting risk.

Recently, bond yields have risen in response to a strong economy, sticky inflation, and a new Federal Reserve Chair whose communication style is quite different from that of his predecessor. While this has contributed to volatility in stocks and bonds, higher bond yields provide investors with more investment options than they have had in much of the past decade. With yields now offering more attractive and predictable sources of income, bonds can play a meaningful role in balancing portfolio risk, particularly after a period of strong equity market gains.

The main question that most market participants are trying to answer now is how much more room the equity market has to move higher. History offers a useful perspective. Looking at bull markets dating back to the 1930s, the average cycle has lasted 4.9 years and delivered a cumulative return of 177.6%. The current cycle has already produced substantial gains, with the S&P 500 returning approximately 125%, including dividends. But history shows that bull markets can continue well beyond their early years. The bull market of the 1990s lasted 9.4 years and generated a cumulative return of 417%. Similarly, the bull market that followed the 2008 Global Financial Crisis lasted nearly 11 years and produced a return of approximately 400%.

Technology was an important driver of both of these past cycles. The internet boom of the 1990s and the growth of digital businesses following the Global Financial Crisis demonstrate how innovation can fuel years of earnings growth and rising stock prices. AI could prove similarly transformative. But even powerful technological shifts do not guarantee strong investment returns at any price. Starting valuations, along with the pace at which such innovation translates into profits and the expectations already reflected in stock prices, all matter.

Despite the market’s strong gains, we believe overall valuations remain reasonable relative to historical levels, and today’s market does not exhibit the widespread speculative excess typically associated with a classic bubble. The fact that so many stocks declined over the past year reinforces how uneven this advance has been. Looking ahead, we expect solid earnings growth to remain an important driver of equity returns, supported in part by continued investment in, and adoption of, AI.

That said, the path forward is unlikely to be smooth. More recently, investors have begun questioning whether some of the AI infrastructure companies with the highest leverage relative to their cash flow will be able to generate positive returns on the investments they are making. As these concerns have grown, borrowing costs for some of these companies have risen. Even though these same considerations have already sent some shivers through the equity market, we view this as a healthy development. Moreover, the bond market appears to be imposing greater discipline on companies taking on significant debt, potentially helping to curb excesses before they possibly evolve into a broader financial bubble.

Ultimately, we believe the current environment calls for neither complacency nor excessive caution. The combination of reasonable valuations, solid earnings growth, and the potential for AI to drive productivity across the economy provides an ongoing, generally constructive backdrop for equities. At the same time, shifting market leadership, the risk of normalizing profit margins, and questions about the returns on AI infrastructure investments warrant a disciplined approach. We believe the greatest opportunities may emerge as the benefits of AI spread beyond today’s market leaders, reinforcing the importance of both diversification and looking beyond recent winners. Our focus remains on identifying attractive long-term investments, rebalancing when opportunities arise, and helping clients participate in future growth without taking excessive risks.