[DigitalToday reporter Jinju Hong (홍진주)] As the weighting of artificial intelligence (AI)-related stocks in the S&P 500 expands to about 45 percent of market capitalisation, concentration in a small number of megacap technology stocks is emerging as a key risk for U.S. equities.

On April 22 local time, blockchain media outlet Cryptopolitan reported that AI-related stocks, including data centres, semiconductor makers and power companies, already account for more than 40 percent of the S&P 500’s total value.

Megacap technology stocks led by Nvidia, along with increased AI infrastructure investment, have been at the centre of the index’s gains. Nvidia had the largest weighting in the index at 7 percent as of March 30, topping Apple at 6.3 percent, Microsoft at 4.6 percent and Amazon at 3.7 percent. The combined weighting of the top 5 AI companies is about 30 percent, the highest level of concentration in 50 years.

The weighting of the top 20 AI-related stocks also neared half of the index. That is higher than at the peak of the dotcom bubble in 2000. As market money concentrated early this year on AI infrastructure and semiconductors, other sectors such as cybersecurity and enterprise software were relatively sidelined.

As concentration deepens, warnings about structural risk are also growing. Critics say “index deleveraging” could occur if a pullback in some large AI stocks cannot be offset by the rest of the market. Some in the market also assess that the S&P 500 is effectively moving like a fund centred on megacap technology stocks.

Rising AI investment is also having a major impact on corporate profit structures. Goldman Sachs estimated that AI infrastructure investment in 2026 will account for about 40 percent of the S&P 500’s total profit growth. Data centre construction and AI capital spending have already entered a structural scale and are forecast to reach about 2 percent of U.S. gross domestic product by the end of 2026.

Expanded investment by megacap technology companies is also accelerating shifts in market structure. AI spending by hyperscalers such as Alphabet and Microsoft is surging, and this year’s AI infrastructure investment by the so-called “Big 4” of Amazon, Alphabet, Meta and Microsoft is expected to reach as much as 700 billion dollars. That is a 50 to 60 percent increase from a year earlier.

Performance gaps are also clear. Since the launch of ChatGPT in 2022, AI-related companies have risen about 200 percent, while the average gain among the remaining hundreds of companies has been less than 30 percent. That is why the market is raising the possibility of a broader re-rating if the AI investment cycle slows.

The problem is that the diversification effect is weakening. As industrials, energy and technology sectors are all linked to investment in data centres and AI infrastructure, portfolio diversification is becoming increasingly difficult, analysts say. That is adding to concerns that index volatility could widen even on small negative news.

The need to change investment strategies is also being raised. Morgan Stanley and Goldman Sachs said it is necessary to look beyond simple exposure to technology stocks and focus on AI-adopting companies with pricing power, along with real-economy infrastructure-related stocks such as manufacturing and energy.

Industries benefiting from the trend also appear to be gradually expanding. GE Vernova, Seagate Technology, Palantir Technologies and Super Micro Computer were cited as early beneficiaries, and interest has recently spread to companies building physical AI infrastructure such as Lumentum, Vertiv Holdings and Coherent.

The market points to whether large-scale AI investment can translate into actual revenue and profit growth as the key issue ahead. With AI established as a core driver of the stock market’s rise, whether results can support expectations is expected to shape the market’s direction.