AI and the Markets: Progress, Promise, and Prudence

Over the past three years, the stock market has been dominated by one powerful theme: artificial intelligence. Since the release of ChatGPT in late 2022, a small group of large technology companies—often called the “Magnificent Seven”—has powered the majority of the market’s advance.

While the S&P 500 has risen sharply since that time, roughly three-quarters of the total gain has come from those seven companies alone. They now represent about one-third of the index’s total market value, a concentration level that investors have rarely seen in modern history.

Echoes of the 1990s—But on firmer ground

The parallels to the late-1990s dot-com era are hard to miss. Then, the internet promised to reshape the economy, valuations soared, and investors raced to participate. The story ended with a painful reset when many of those early internet companies failed to deliver.

Today’s situation feels familiar—but not identical. The leaders of the AI boom are not speculative startups; they are some of the most profitable and well-capitalized firms in the world. They are generating real earnings, building real infrastructure, and reshaping real industries. Still, even the strongest businesses can see their stocks overextend when enthusiasm runs ahead of fundamentals.

The scale of investment is extraordinary

Unlike the dot-com era, the AI transformation requires enormous physical investment. Data centers, chip fabrication plants, and energy infrastructure are being built at a record pace to meet surging demand for computing power.

• Major technology companies (Amazon, Google, Meta, Microsoft, Oracle, and Coreweave) are projected to spend $443 billion in data-center investment in 2025, up from $133 billion in 2023.
• OpenAI has committed to $1.4 trillion in spend on data-center capacity.
• Broad estimates suggest that $6–7 trillion in cumulative data-center investment could be needed globally by 2030 to support AI growth.
• On the most optimistic side of things, Nvidia’s CEO, Jensen Huang, has indicated that data-center spend could approach $3-4 trillion per year by 2030.

This level of commitment underscores that the AI revolution is not simply speculative—it is becoming embedded in the real economy.

Navigating concentration and opportunity

At NPF, we view AI as a genuine technological and economic force, but also as an area where discipline and diversification are essential. The dominance of a handful of names means portfolios can become concentrated without investors realizing it. Our approach is to participate thoughtfully—owning high-quality companies that benefit from long-term innovation while ensuring the portfolio remains balanced across sectors and styles.

The reality is technological transformations don’t happen overnight and speedbumps are sure to occur. In the case of AI, think of things like the massive power consumption the data centers require, thus driving up energy prices until power generation can catch-up. Sure, further innovation can solve these bottlenecks, but it might take longer than hoped for the economics of AI to bear fruit.

Markets rarely reward extremes. In the late 1990s, investors who stayed diversified and valuation-aware ultimately fared far better than those who chased momentum. The same lesson applies today.

The bottom line

AI will likely define the next decade of technological progress, just as the internet defined the last. But markets tend to overestimate what is possible in the short-term, (which drives stocks much higher) and underestimate what is possible over the long term. As stewards of client capital, our job is to engage with opportunity—without losing sight of prudence.

We remain committed to participating in innovation, maintaining balance, and protecting capital through every phase of the cycle.