The other side of the AI boom: from SaaS light assets to AI heavy assets. Financial challenges for American giants are just beginning.

A new report from the Artificial Intelligence column of Heard on The Street shows that despite the strong profitability of the largest tech companies in the United States, their aggressive investment in AI computing infrastructure is beginning to weaken their traditional financial fundamentals that have put them ahead of global enterprises. Financial statements may enter a period where massive AI expenses make it difficult to see strong AI profits. Since 2023, Microsoft Corporation, Alphabet Inc. Class C (parent company of Alphabet), and Amazon.com, Inc. have collectively invested over $600 billion in infrastructure spending related to AI. This massive expenditure is starting to reshape their long-standing strong balance sheets. Data shows that the seven major tech giants in the United States, known as the Magnificent Seven, had very little debt and substantial cash reserves in the early stages of the AI computing infrastructure trend. These reserves came from strong revenue unrelated to AI. Meanwhile, their combined net profits from 2023 to 2025 are forecasted to exceed $750 billion, providing a strong and reasonable financial basis for the surge in AI infrastructure spending. The so-called "Magnificent Seven", which account for a high weight in the S&P 500 Index and the Nasdaq 100 Index (approximately 35%), includes Apple Inc., Microsoft Corporation, Alphabet Inc. Class C, Tesla, Inc., NVIDIA Corporation, Amazon.com, Inc., and Meta Platforms (parent company of Facebook). They are the core drivers of the record highs in the S&P 500 Index and are considered by top Wall Street investment firms to be the combination most capable of bringing huge returns to investors in the biggest technological transformation since the internet era. However, the increasing AI investments are diluting liquidity. Even Microsoft Corporation, with a market capitalization close to $4 trillion, currently has cash and short-term investments accounting for only about 16% of total assets, significantly lower than the strong level of about 43% in 2020. Alphabet and Amazon.com, Inc. are also experiencing similar declines in cash reserves as their spending rapidly expands with the significant increase in data center construction and AI hardware procurement. The tech giants' long-standing pride in free cash flow is also under pressure. Alphabet and Amazon.com, Inc.'s free cash flow is expected to be significantly lower than last year, and Microsoft Corporation's free cash flow is expected to decrease further after long-term data center leasing is included in capital expenditures. The latest Wall Street forecast data shows that these three tech giants will collectively spend about $1 trillion on AI infrastructure from 2023 to next year. Other tech giants are also increasing their borrowing to keep pace with the AI computing infrastructure spending in the AI trend. Meta (parent company of Facebook) recently doubled its borrowing by selling $30 billion in bonds, and Oracle Corporation raised $18 billion after reaching a large-scale cloud computing service agreement with OpenAI. Analysts from the Artificial Intelligence column of Heard on the Street pointed out on Sunday that these major shifts (including reduced cash buffers, significantly increased capital requirements, and higher leverage) are driving large tech companies towards a more capital-intensive "AI computing power business model". They are no longer just seen as high-profit pure SaaS software or cloud computing platform-type asset-light companies, and investors may need to develop new valuation systems for them, similar to evaluating hardware and asset-driven industries. The ultimate success will depend on whether large and long-term investment bets can be transformed into strong profit margins like Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR and Foxconn. Analysts expect that the new evaluation criteria will receive more attention, such as the specific adoption of AI computing power customers, contract backlog, and monetization of generative AI application software services. Recent market reactions indicate that investors have significantly less tolerance for tech companies whose AI spending has not yet translated into substantial profits: for example, Amazon.com, Inc. dropped by about 5% last week, and Meta's stock price has dropped by over 20% since hitting its peak in August. The column analysts point out that the increasing investments by tech giants also increase the risk of strategic errors, whether through excessive construction of AI computing power infrastructure capacity or supporting cutting-edge innovative technologies that fail to gain market acceptance. Even how AI resources are allocated has become a strategic challenge. The CFO of Microsoft Corporation, for example, mentioned that the company is shifting more AI computing power infrastructure resources to its application software division, leading to a decrease in computing power resources available for Azure traditional cloud computing services. The column analysts conclude that large tech companies are confident that their unprecedented ambitions in AI will provide optimistic long-term prospects for long-term investment returns, but they also bring new and potentially increasing growth uncertainties. With the overall extension of the trend towards capital-intensive investment cycles and the emerging threat of underutilized AI computing power infrastructure, the financial fundamentals of the large tech giants are just beginning to trend in a new direction.