Few investors would have predicted that 2026 would bring a new conflict with Iran or significant U.S. action involving Venezuela. Yet here we are, more than halfway through the year, with markets continuing to push higher despite ongoing geopolitical tensions, including the Russia-Ukraine war and an accelerating artificial intelligence competition between the United States and China.
At a time when information is more abundant than ever, the challenge for investors is no longer accessing news—it is determining what is meaningful and what is simply noise. Headlines, opinions, and market commentary move faster than ever, making it increasingly difficult to separate lasting trends from short-term distractions.
Markets are often driven by narratives, and today is no exception. Strong stories can create powerful momentum, but they can also lead to speculation, concentrated market trends, and increased volatility.
The key question for investors is, What truly matters, and what is simply noise?
What is Important for Markets
Currently, there is one theme influencing markets more than almost any other, artificial intelligence. AI represents a significant technological advancement, combining breakthroughs in computing power, data, and human ingenuity. Its potential applications span nearly every industry, from automating business processes to improving productivity, accelerating research, and transforming how companies operate. Whether AI ultimately becomes one of the most important technological shifts in history remains to be seen, but markets are clearly placing significant value on its future potential.
AI-related companies have become an increasingly important part of equity markets. While there is no official classification for an “AI company,” investors typically include companies involved in semiconductors, hyperscale cloud computing, enterprise software, networking, data centers, and other critical AI infrastructure.
Based on various market analyses, including research from J.P. Morgan Private Bank, Michael Cembalest, and Bianco Research, AI-related companies are estimated to represent roughly 45% of the S&P 500 by market capitalization, depending on the definition used. The exact percentage can be debated, but the broader point is clear: investor expectations around AI have become a major driver of market performance.
The market is not simply investing in AI companies—it is investing in the belief that AI will reshape the global economy. Nearly every major company is now exploring how to incorporate AI into their business models, making it one of the most important investment themes of this cycle.
The AI Expansion
The world’s largest technology companies, Amazon, Microsoft, Alphabet, Meta, and Oracle, are leading the global AI infrastructure buildout by integrating generative AI into their cloud platforms and investing heavily in the systems required to support it.
This investment extends far beyond software. Companies are spending billions on AI data centers, advanced GPUs, custom-designed AI chips, high-speed networking, and the power and cooling infrastructure needed to operate increasingly complex AI models. Over the past several years, AI has moved from an emerging technology to a central component of cloud computing, enterprise software, and the broader digital economy.
The scale of this investment is significant. Goldman Sachs estimates that the largest technology companies will invest approximately $5.3 trillion between 2025 and 2030 in AI infrastructure and data center expansion.1 This represents a substantial increase from previous estimates and highlights the rapidly growing demand for AI computing capacity.
To put the size of this investment into perspective, U.S. Gross Domestic Product (GDP)—the total value of goods and services produced within the economy was approximately $30.8 trillion in 2025. If annual AI infrastructure investment were to approach $1 trillion per year, it would represent more than 3% of U.S. GDP.
While historical comparisons are imperfect, previous periods of major infrastructure investment provide useful context. The expansion of the U.S. railroad network in the late 1800s and the buildout of the electrical grid in the early 1900s were transformative periods that required enormous amounts of capital and reshaped the economy for generations. These however led to major bust cycles due to overbuilding and slower adoption that initially projected.
The Fears
AI has become one of the most significant drivers of equity markets. However, one area that receives less attention is the growing exposure within fixed income markets. The AI investment cycle is not being funded solely through company cash flows and equity markets—large technology companies are also increasingly accessing debt markets to finance infrastructure expansion.
Over the past year, major technology companies have issued significant amounts of investment-grade bonds to support capital-intensive investments, including data centers, AI infrastructure, and computing capacity. This means the AI trade extends beyond stocks and into the broader financial system, with increasing exposure across both equity and fixed income markets.
At the moment, there are no immediate signs of stress. However, markets may become more cautious if AI-related revenue growth fails to meet expectations, if companies invest ahead of demand, or if competing AI models from other regions, including China, begin to challenge the technological advantage of U.S. companies.
Because markets are heavily influenced by narratives, a shift in investor expectations could create significant volatility in AI-related companies. History has shown that even transformative technologies can experience sharp market corrections when expectations move faster than fundamentals.
The semiconductor industry provides a useful example. Demand for AI infrastructure has created significant demand for components such as high-bandwidth memory (HBM), leading to supply constraints and strong pricing power for manufacturers. Memory companies have benefited from this environment as investors anticipate sustained AI-driven demand.
However, semiconductors have historically been a cyclical industry. Periods of strong demand often lead companies to expand capacity, which can eventually create oversupply and pressure pricing. The challenge for investors is determining whether AI represents a structural, multi-decade shift in computing demand or whether parts of the industry are experiencing a traditional semiconductor cycle amplified by a powerful new theme.
This is the central debate facing investors today: Are AI-related semiconductors, memory, and infrastructure companies entering a new era of sustained growth, or are we witnessing another cycle that will eventually normalize?
Is this time different?
The answer will likely only become clear with time. As history has shown, major technological transformations often appear obvious in hindsight—but navigating them in real time requires balancing opportunity with disciplined risk management.
Markets
While markets have been heavily focused on artificial intelligence and the potential opportunities it creates, investors are also facing a number of broader concerns. One of the most important is valuation.
Several long-term valuation indicators suggest that U.S. equities are trading at historically elevated levels. These measures do not predict short-term market movements, and expensive markets can remain expensive for extended periods of time. However, they can provide insight into the potential long-term return environment investors may face.
One widely followed valuation measure is the Buffett Indicator, which compares the total market capitalization of U.S. equities to U.S. GDP.
Buffett Indicator = Total U.S. Stock Market Value ÷ U.S. GDP
The logic behind the indicator is straightforward: over time, the value of companies should generally be connected to the size and growth of the economy that supports them. When the ratio becomes significantly elevated compared with historical averages, it suggests that investors are paying a higher price for future earnings.
Today, the Buffett Indicator is near historically high levels, reflecting the strong performance of U.S. equities and the increasing market value of large technology companies. However, it is important to recognize that the measure has limitations. Many large U.S. companies now generate substantial revenue outside the United States, meaning their market value is not solely tied to domestic economic output.
Another valuation measure receiving significant attention is the Cyclically Adjusted Price-to-Earnings Ratio (CAPE), developed by economist Robert Shiller. Unlike a traditional price-to-earnings ratio, which uses only one year of earnings, CAPE adjusts for economic cycles by using 10 years of inflation-adjusted earnings.
CAPE = S&P 500 Price ÷ 10-Year Average Inflation-Adjusted Earnings
Historically, higher CAPE ratios have been associated with lower expected long-term returns, while lower CAPE ratios have generally occurred during periods when future returns were stronger. Importantly, CAPE is not a timing tool—it does not indicate when a market decline may occur—but rather provides context around expected returns over longer periods.
Research from StarCapital examined thousands of historical periods to study the relationship between starting CAPE valuations and subsequent 15-year real returns. Their analysis found a consistent pattern: as starting valuations increased, average future returns tended to decline.
StarCapital, Predicting Stock Market Returns Using CAPE.
Their research showed examples of historical outcomes:
- CAPE of approximately 8: Average 15-year real returns of roughly 13% annually
- CAPE of approximately 24–25: Average 15-year real returns in the low-to-mid single digits
- CAPE of approximately 32: Historically associated with much lower or negative real returns over the following 15 years
*The current CAPE ratio is around 40 in the US.2
One important historical example is Japan in the late 1980s. Japanese equity valuations became extremely stretched, with CAPE ratios reaching levels far above historical norms. Investors who purchased at those peak valuations experienced poor long-term returns as valuations eventually normalized.
However, the comparison should be viewed carefully. The U.S. market today is not identical to Japan in 1989. The U.S. economy is more diversified, American companies have stronger global earnings exposure, and many of today’s largest companies are highly profitable businesses with significant cash flows.
The takeaway is not that a market decline is imminent. Rather, valuation indicators suggest that investors should have realistic expectations for future returns and recognize that periods of strong market performance are often followed by periods of consolidation.
Markets are currently pricing in significant optimism around AI, economic growth, and corporate earnings. If those expectations are met, markets can continue to perform well. However, when valuations are elevated, the margin for disappointment becomes smaller.
Final Thoughts: Opportunity, Optimism and Risk Management
When looking at markets today, there are reasons for both optimism and caution.
In the short term, the opportunities surrounding artificial intelligence are significant. AI is driving innovation, attracting massive investment, and creating opportunities across industries. Market momentum remains strong, and it is entirely possible that AI-related companies and broader markets continue to perform well as businesses invest in this next generation of technology.
However, longer-term investors should also recognize the risks that come with periods of strong market enthusiasm.
U.S. equities currently represent an unusually large share of global markets. Many global indexes now have significant exposure to U.S. stocks, with some allocations exceeding 70%. This concentration reflects the strength of American companies, particularly large technology firms, but it also creates a potential vulnerability if investor preferences begin to shift toward other regions, sectors, or asset classes.
Market leadership has historically rotated over time. During the late 1990s technology boom, technology companies grew to represent a substantial portion of the S&P 500 before declining sharply following the dot-com bubble. Today, technology-related companies once again represent a historically large share of the market. Could this continue? Absolutely. Strong companies can continue growing and valuations can remain elevated longer than many investors expect.
However, history suggests that periods of extreme concentration eventually tend to normalize.
AI-related sectors, particularly semiconductors and memory companies, have experienced extraordinary momentum as demand for AI infrastructure continues to accelerate. Companies are investing aggressively because securing technological leadership has become a strategic priority. The competition among businesses and countries to build AI capabilities is real, and many companies are willing to spend heavily today to gain an advantage tomorrow.
The challenge for investors is distinguishing between a transformational technology and the expectations already embedded in stock prices. AI may fundamentally change the global economy, but even transformative technologies can experience periods of excessive optimism, volatility, and valuation resets.
One of the most important lessons from previous market cycles is that recognizing a great technology does not always mean avoiding losses. Legendary investor Stanley Druckenmiller reflected on this after losing billions during the technology bubble despite understanding the risks:
“So, I’ll never forget it. January of 2000 I go into Soros’s office and I say I’m selling all the tech stocks, selling everything. This is crazy at 104 times earnings. This is nuts. Just kind of as I explained earlier, we’re going to step aside, wait for the net fat pitch. I didn’t fire the two gun slingers. They didn’t have enough money to really hurt the fund, but they started making 3 percent a day and I’m out. It is driving me nuts. I mean their little account is like up 50 percent on the year. I think Quantum was up seven. It’s just sitting there.
So like around March I could feel it coming. I just – I had to play. I couldn’t help myself. And three times during the same week I pick up a phone, don’t do it. Don’t do it. Anyway, I pick up the phone finally. I think I missed the top by an hour. I bought $6 billion worth of tech stocks and in six weeks I had left Soros and I had lost $3 billion in that one play. You asked me what I learned. I didn’t learn anything. I already knew that I wasn’t supposed to do that. I was just an emotional basket case and couldn’t help myself. So, maybe I learned not to do it again, but I already knew that.”
The lesson is not that investors should avoid innovation. The lesson is that even experienced investors can struggle when excitement, momentum, and expectations become overwhelming.
Markets are difficult because the future is uncertain. Artificial intelligence may become one of the most important technological developments in history, but the path forward will likely include periods of volatility and adjustment.
The coming years may bring tremendous opportunities, but they will also require discipline. Managing risk, maintaining diversification, and protecting capital will remain essential.
We are living through a period of significant technological change. The opportunities are substantial, but history reminds us that every new era brings both breakthroughs and challenges.
1 Goldman Sachs Research, Private Markets Are Expected to Have a Growing Role in Data Center Financing.
2https://ycharts.com/indicators/cyclically_adjusted_pe_ratio