Market Concentration and Pricing in the Tech Sector
In the contemporary world, a handful of large technology companies have come to dominate the stock indices. This dominance is particularly noticeable within the S&P 500 and global indices, where these technology platforms hold an unusually high share. Bloomberg provides data supporting this observation.
Interestingly, AI narratives have been found to account for the majority of stock market gains since the end of 2022. This is indicative of AI’s growing influence and importance in the global economy. However, it also highlights the potential volatility of the market. Even a mild shock, such as the emergence of a surprising competitor or regulatory action, can cause significant shifts in market value.
An example of this was the DeepSeek episode in early 2025. A cheaper model from China briefly wiped out substantial amounts of market capitalization, demonstrating how quickly the market mood can change when the narrative shifts.
Expenditure That Exceeds Current Returns
Investment in AI infrastructure has skyrocketed, reaching historic levels. Large technology companies are collectively spending hundreds of billions of dollars per year on data centers, GPUs, and power. Some forecasts, such as those published by Gartner, suggest that AI-related investments will exceed $500 billion annually for several years.
However, there seems to be a mismatch between expenditure and revenue. Revenue from direct AI services is still much smaller, measured in the tens rather than hundreds of billions in some segments. Despite the high level of investment, most companies experimenting with generative AI are not yet seeing a significant impact on their bottom line.
Extensive studies, such as those conducted by McKinsey & Company, show that most AI initiatives have so far shown little or no measurable ROI. Many projects improve individual productivity, but not overall margins or revenue growth. In many cases, AI is often still stuck in pilot mode and is not deeply embedded in operations. Without a clear winning story, it may be challenging to justify large early investments in the long run.
Circular and Aggressive Financing
Some AI contracts and investments appear to be structured to keep the sector’s momentum going. Providers are purchasing large blocks of cloud capacity from each other in advance, and AI labs are committing to spending huge sums on specific infrastructure providers. These commitments then appear as future revenue growth on the supply side, even if the buyer does not yet have an easy way to recoup that money.
While this isn’t necessarily unethical, it does create a feedback loop in which optimistic assumptions on both sides reinforce each other. If one part fails, the loop can quickly unravel, leading to potential instability in the market.
Physical Limitations: Energy, Cooling, and Land
AI is more than just software; it’s a significant consumer of concrete, copper, and megawatts. Modern AI data centers can consume as much electricity as a large city, putting strain on local networks, water supplies, and approval processes. Governments and regulators are beginning to question whether unlimited AI expansion is compatible with climate goals and local infrastructure.
If power or cooling becomes a hard limit in key regions, some current investment plans may have to be scaled back. Such a hard stop is a classic trigger for asset revaluation, and could lead to significant changes in the AI market.
For more in-depth analysis on this topic, click Here.
