Reassessing Data Center Investment in the AI Infrastructure Race
Two years since ChatGPT changed the world. The democratization of generative AI has unleashed an unprecedented investment frenzy in the technology industry. In particular, big tech companies such as Google (GOOGL), Microsoft (MSFT), and Meta (META) are pouring trillions of dollars into building massive data center infrastructure required for AI training and inference. However, recent market reactions suggest growing questions about how much of this massive investment will actually translate into real revenue.
Unprecedented Investment Scale, Uncertain Returns
Market data as of October 27, 2026 reveals the basis for this concern. Microsoft rose 36.31% over three months, recording a market cap of $3.91T, but its P/E ratio of 29.4 is significantly higher than other companies (GOOGL 16.3, META 20.9). This suggests that the current strong stock price reflects high expectations for future earnings. Conversely, Google declined 6.47% over one month, and Meta plunged 13.34% over one month. Both companies are investing tens of billions of dollars in AI data center expansion, yet the market is sending skeptical signals about the impact.
In Google’s case, the net profit margin is very high at 54.8% and revenue growth is healthy at 24.2% (YoY), yet the stock price remains weak. Similarly, Meta has the steepest revenue growth among the three companies at 26.1% (YoY), but its net profit margin is only 31.1% and the stock has recorded significant declines on a one-month basis. This suggests that investors are concerned about deteriorating profitability in the future rather than current earnings.
Data Center Investment: “How Much Will It Return?”
The core issue is paradoxical. Demand for computing power needed to train and operate AI models is growing exponentially. Yet the revenue model to support it remains unclear. Cloud service margins fall short of the advertising business of the past, and the enterprise AI market has yet to reach a stage of generating substantial revenue. More critically, all major companies are making similar scale AI investments. As a result, each company faces inevitable price competition while bearing massive capital expenditures (CapEx). Microsoft’s P/E ratio of 29.4 appears to reflect market confidence in its technological leadership and profitability despite these concerns, but Google’s high net profit margin (54.8%) and Meta’s strong revenue growth (26.1%) have proven insufficient to prevent current stock declines.
Divergent Approaches to Balance
The financial metrics of the three companies suggest different stories. Microsoft’s high P/E reflects market confidence in revenue diversification through the OpenAI partnership and the Copilot ecosystem. In contrast, Google and Meta remain in the phase of large-scale data center investment while generating insufficient immediate returns, making market valuations stricter. Particularly for Meta, which has experience pouring tens of billions of dollars into the “metaverse,” investor skepticism may be stronger. Microsoft’s relatively strong stock performance reflects its strategy of linking AI investment to existing businesses—office productivity and enterprise cloud.
Future Outlook: Convergence or Divergence?
The trend of reassessing data center investment seems likely to continue. If returns from AI fail to keep pace with investment scale, a correction phase is unavoidable. Conversely, if new applications grow rapidly, current investment levels could be justified. For investors, this is the time to closely examine the revenue growth rates of the three companies (GOOGL 24.2%, MSFT 17.7%, META 26.1%), trends in profit margins, and the specific AI revenue pathways presented in upcoming quarterly results.
This article is for information only and is not investment advice.