Summary of Key Points
On the evening of July 22nd, the financial reports of Google and Tesla simultaneously revealed a critical issue: both companies experienced a significant decline in free cash flow. For Google, it was the first time in decades that its free cash flow turned negative for a single quarter, at -$5.855 billion; for Tesla, it was the first time in two years. The reason behind this is that both companies are investing heavily in AI infrastructure. Google's capital expenditures doubled year-over-year to $44.9 billion, while Tesla's increased by 142% to $5.79 billion. Even worse, this is not an isolated phenomenon; other tech giants such as Amazon, Meta, and Microsoft are also facing the crisis of depleted free cash flow, collectively engaging in a "race to burn money" for AI development. This shift marks a turning point in the narrative around AI investment: the market's focus has shifted from wondering how much growth AI can bring to questioning who will bear the cost of that growth.
1. Free Cash Flow Decline: How Did the Giants Run Out of Money?
Let's break it down simply: Free cash flow is the money a company earns after deducting large investments necessary for operations (such as building data centers and purchasing equipment). This remaining funds can be used for dividends, stock repurchases, or to address crises.
Currently, the free cash flows of these giants have hit rock bottom:
- Google: Free cash flow in Q2 was -$5.855 billion, the first negative figure in decades. Capital expenditures increased from $22.4 billion last year to $44.9 billion, and the company raised its annual capital expenditure forecast from $180-190 billion to $195-205 billion, with the CFO stating that it will continue to grow significantly in 2027.
- Tesla: Free cash flow in Q2 was -$1.09 billion, the first negative figure in two years. The company invested $5.79 billion in AI capabilities, autonomous driving, and robotics production lines.
- Other giants are in even worse shape: Amazon had operating cash flow of $26 billion in Q1 but capital expenditures of $44 billion, leaving a substantial gap; Meta's free cash flow halved in the first quarter, and it suspended stock repurchases while issuing $550 billion in debt; Microsoft's free cash flow is volatile, with analysts predicting a negative figure for Q4.
Bank of America previously said Google was the "only exception," but that view has been overturned. Now all four major cloud providers (Amazon, Microsoft, Alphabet, Meta) are experiencing significant free cash flow declines.
2. Where Did All the Money Go? The Bottomless Hole of AI Infrastructure
The answer is straightforward: The demand for AI infrastructure—computing power, data centers, and chips—is consuming enormous amounts of capital.
- Google: $44.9 billion in Q2 was invested in AI capabilities, data centers, and TPU chips. The CFO noted that despite capacity expansion over the past three years, demand still exceeds supply, to the point where they had to rent additional computing power from third parties (such as CoreWeave).
- Tesla: $57.9 billion was invested in AI capabilities (autonomous driving models) and Optimus robotics production lines. Elon Musk emphasized that robotics are among the most challenging areas for scaling, but investment is necessary since all future businesses will rely on AI.
Industry projections are even more alarming: Morgan Stanley predicts that the five major cloud providers (Amazon, Google, Meta, Microsoft, Oracle) will spend $805 billion on AI infrastructure in 2026 and $1.1 trillion in 2027—this is roughly equivalent to the annual GDP of the Netherlands!
In simple terms, AI is like an insatiable appetite that requires continuous funding; if companies don't invest, they will fall behind their competitors.
3. Where Did the Shareholders' Money Go? From Dividends to Borrowing
Previously, when these giants made profits, their favorite activity was to repurchase shares (boosting stock prices and benefiting shareholders). But now?
- Google: Suspended share repurchases for the first time since 2015.
- Meta: Suspended share repurchases for the longest period since 2017 and issued $550 billion in debt within six months.
- Google itself also issued $310 billion in new debt, along with additional bonds in euros and Canadian dollars.
Why? Because all the money is being invested in AI. Franklin Templeton analysts commented, "It would be perplexing if Tesla reduced its AI investments, given that all its future businesses depend on AI."
This means that while companies once used profits to reward shareholders, they now use them for capital expenditures, leaving them with no cash for share repurchases and forced to borrow.
4. The Imbalance in the Supply Chain: Chip Manufacturers Profit, While Cloud Providers Suffer
The benefits of AI investment are largely going to chip manufacturers:
- On the evening of July 22nd, chip stocks surged: Supermicro Computers rose 23%, Dell increased by 9%, and CoreWeave (a company that rents out computing power) rose 3.7%.
- Samsung's profits in the second quarter increased by 18 times compared to the same period last year, surpassing NVIDIA; the three leading memory companies (Samsung, SK Hynix, Micron) saw their stock prices rise by an average of 639%, with market values increasing by $1.07 trillion.
In contrast, the market value of the seven major U.S. cloud providers shrank by $2.2 trillion in June. The Financial Times pointed out that investors are no longer favoring companies that burn money but are betting on chip manufacturers, as they are the direct beneficiaries of AI infrastructure development.
5. The Prisoner's Dilemma: A Race to Burn Money That No One Wants to Stop
Professor Leuz of the University of Chicago described this situation as a "prisoner's dilemma": If you don't invest in AI, your competitors will gain an advantage, so you must follow suit, even if it means continuing to spend excessively.
- Sundar Pichai (Google CEO) said, "We are at the very early stages of an AI transformation, and the future looks more promising." For example, Google's cloud business has $514 billion in pending contracts, and search volume increased by 10% after adding AI summaries—these are signs of positive outcomes.
- However, Leuz warns that this pattern is similar to over-investment in heavy industries like telecommunications and chemicals: excess capacity leads to declining profit margins. Some companies even use special purpose vehicles (SPVs) to fund data center projects, allowing external investors to bear the debt without it appearing on their financial statements (such as Oracle's and OpenAI's $300 billion contracts), masking the true financial pressures.
In other words, everyone recognizes the potential for overcapacity, but no one wants to be the first to stop investing, as doing so would mean losing out.
6. Why Is This Night Being Mentioned Again and Again?
Previously, the market focused on how much growth AI could bring. Now, the question is: "Who will pay for that growth?" This year, companies are planning to invest $725-805 billion in AI; next year, the figure is likely to exceed one trillion. This money could either lead to increased AI revenue and profits or create the largest technology bubble in history.
At the next earnings call, investors' questions may no longer be about AI's potential growth but about when they will recoup their investments. This night marks a watershed moment as AI moves from being a promising story to a real-world challenge.
(The entire analysis is written in plain language, avoiding technical jargon, making it easy for non-financial professionals to understand.)