The Lingering Suspense of Tech Giants' Earnings Season: As Both "CapEx" and "AI Revenue" Grow, Who Will Win the Narrative?

Tech giants are undergoing an unprecedented capital expansion, but market patience for this high-stakes gamble is diverging.

This earnings season reveals a core contradiction: the scale and speed of AI investment far outpace the rhythm of revenue monetization. While robust growth in cloud business has bought some companies breathing room, the collective deterioration of free cash flow has put pressure on the stock prices of others. The winner of this narrative battle will gradually emerge over the next few years.

In this quarter's reports, Amazon's capital expenditure surged 69% year-over-year to approximately $54 billion, Meta's capital expenditure jumped 83% year-over-year to $31.1 billion, Microsoft's capital expenditure also grew by 69%, and Google announced it was raising its full-year spending target by $15 billion to a maximum of $205 billion. According to Wall Street estimates compiled by FactSet, the combined capital expenditure of the four companies is expected to reach approximately $1.6 trillion over the next two years. Since the AI race began in 2023 until June of this year, the cumulative capital investment by the four companies has exceeded $1.1 trillion.

Market reactions were sharply divided. Amazon's cloud business recorded its fastest growth rate since early 2022, with its after-hours stock price rising nearly 10%; Microsoft Azure's quarterly sales hit a record high, and its stock price rose more than 15% the following day. Meta's stock fell more than 7% due to revenue prospects missing expectations coupled with continued spending expansion. Google experienced its first-ever quarterly negative free cash flow since its listing, with its stock price dropping more than 6% the next day. The combined free cash flow of the four companies fell to a decade-low of approximately $7 billion, marking a historic low.

Scale of Spending: An Arms Race Without a Finish Line

From single-quarter data, the combined capital expenditure of the four companies reached $170 billion in the second quarter of this year, a year-over-year increase of approximately 78%.

This figure is being refreshed every quarter, and upward revisions have become routine—Amazon raised its full-year capital expenditure forecast by $20 billion to approximately $220 billion this quarter, mainly due to a significant rise in memory chip prices; Google increased its full-year target by $15 billion; Meta raised the lower bound of its full-year capital expenditure from $125 billion to $130 billion.

The continuous expansion of spending scale is partly due to supply chain pressures.

Reportedly, shortages of memory chips have driven up procurement costs. Microsoft disclosed in April this year that soaring chip prices would add an extra $25 billion to its spending this year. This cost pressure has even affected Apple, which is not directly involved in the AI race—the company warned that rising costs would depress sales and profit margins, causing its stock price to fall by 6.3%.

Rishi Jaluria, an analyst at RBC Capital, stated, "There is basically no end in sight for the growth in capital expenditure. Investors need these companies to walk a tightrope between AI investment and maintaining competitiveness in their existing businesses."

Cloud Computing: The Core Channel for AI Monetization

Currently, the accelerated growth of cloud computing business is what best supports the narrative that "investment is effective."

Amazon's cloud business, AWS, saw sales grow 37% in the second quarter to $42.2 billion, with operating profits rising simultaneously; Microsoft Azure's annual sales exceeded $100 billion for the first time, with its quarterly growth rate being the fastest since 2022; Google's cloud business added approximately $11 billion in revenue year-over-year.

The cloud business customers of the three companies range from OpenAI and Anthropic to various enterprise users. Contract backlogs driven by AI computing power demand have become a key indicator supporting growth expectations. The total value of signed contract backlogs for Amazon, Google, and Microsoft combined is close to $1.7 trillion, more than double that of a year ago.

Amazon CEO Andy Jassy stated during the conference call that cloud services are "very likely to become a trillion-dollar annual revenue business for us in the future." Microsoft CEO Satya Nadella said that the company opened 31 data centers across five continents last quarter, and overall computing capacity will double within two years.

AI Revenue: From Investment Narrative to Monetization Verification

In this quarter's financial reports, the four companies responded to the market's most core question—"Money has been spent, so how much has been earned back?"—in a more quantified manner for the first time.

Amazon provided the most specific figures first: the annualized revenue run rate of AWS's AI business has exceeded $25 billion, and its self-developed chips (Trainium and Graviton) also surpassed $25 billion, both achieving triple-digit percentage year-over-year growth. Jassy highlighted this as a core disclosure point of the earnings report during the conference call, marking the beginning of Amazon separating AI revenue from overall cloud business data for market scrutiny.

Microsoft's annualized AI revenue run rate reached $37 billion. Coupled with Microsoft 365 Copilot paid seats exceeding 30 million (approximately 20 million three months ago, a sequential increase of about 50%), it has formed a dual-wheel monetization structure of "Cloud AI + Office AI." Nadella emphasized, "We are pushing the cost-efficiency curve across new frontiers to ensure that every customer can convert tokens into actual business outcomes."

Although Google did not separately break down AI revenue, management confirmed that Vertex AI, Gemini API, and TPU computing power leasing are the core drivers of the 82% growth in cloud business—Gemini API's token processing volume per minute soared from 16 billion in the previous quarter to 22 billion, and Gemini Enterprise has been adopted by nearly 90% of Fortune 100 companies. Google Cloud's backlog orders surged from approximately $460 billion in the previous quarter to $514 billion, an increment of over $50 billion, mainly driven by enterprise AI demand. The company expects that more than 50% of this will be converted into revenue within the next 24 months.

Meta's AI monetization path is entirely embedded within its advertising engine.

The annualized revenue run rate of AI-driven Advantage+ advertising tools this quarter is approximately $75 billion**, and the annualized revenue of AI value optimization tools exceeds $20 billion (doubling within a year). AI's enhancement of recommendation systems and ad conversion rates led to a 12% year-over-year increase in ad unit price and a 14% growth in ad impressions, driving a 27% year-over-year increase in Family of Apps advertising revenue to $59.4 billion. Zuckerberg stated during the conference call, "AI is accelerating our core business today"—but without an independent cloud business as a direct monetization channel, Meta's return on AI investment is harder for the market to price compared to the other three companies.

Free Cash Flow: Pressure is Consensus, Divergence Lies in Expectation Management

Despite revenue growth, the deterioration of free cash flow is an unavoidable common theme in this quarter's earnings reports.

Amazon's operating cash flow for the past 12 months was $161.4 billion, a 33% year-over-year increase, but after deducting infrastructure investments such as data centers, free cash flow plummeted to negative $7.6 billion. Google experienced its first negative free cash flow in over 20 years since its listing in the second quarter, with a single-quarter deficit of approximately $6 billion. The combined free cash flow of the four companies fell to approximately $7 billion, the lowest in ten years, with only Microsoft and Meta achieving positive values.

Jassy admitted that while building multiple data centers simultaneously, and with a lag of about two years from project initiation to production, "we will spend a large amount of capital expenditure in the short term and face free cash flow pressure until these data centers come online sequentially."

Dec Mullarkey, Managing Director at SLC Management, stated, "For investors, the era of growth at any cost is over. They want to see spending effectively translate into performance, as seen with Google, Microsoft, and Amazon." This also partly explains why Meta's stock price faced greater pressure after the earnings report.

Concentration Risk: The Butterfly Effect of OpenAI and Anthropic

Another hidden danger of this capital expenditure frenzy lies in the fact that part of the demand is highly concentrated among a few AI startups.

The significant growth in contract backlogs largely stems from multi-year computing power purchase commitments from OpenAI and Anthropic. This means that the return on investment for tech giants partially depends on whether these two companies can continue to raise funds and fulfill contracts, as well as whether their respective IPO plans can be realized.

Moody's warned that this structure forms "a more cyclical system that may mask real demand," constituting a concentration risk that cannot be ignored.

Meanwhile, disclosures regarding future financial commitments by each company reveal deeper balance sheet locking risks.

According to the UK's Financial Times, in the second quarter alone, Meta, Google, and Microsoft signed nearly $900 billion in new AI-related obligations, covering data center leases, infrastructure procurement commitments, and new debt. Meta additionally added $68 billion in data center leases in July. Amazon has not disclosed similar detailed data.

Divergence in Narratives: Who Is Winning, Who Is Waiting

After this earnings season, the market's criteria for evaluating tech giants' AI investments are becoming more specific—not just looking at "how much was spent," but more importantly, "how much was earned."

Amazon and Microsoft have temporarily won market tolerance for high capital expenditure thanks to the accelerated growth of their cloud businesses; Google has raised investor concerns due to cash flow issues, despite impressive cloud business revenue; Meta's situation is relatively complex—AI's empowerment of ad targeting drove total revenue to grow 28% year-over-year to $61 billion, but without a cloud business as a direct monetization channel, and with CEO Mark Zuckerberg's statements about leasing data center computing power considered lacking clear planning, the market reacted negatively.

Alphabet CFO Anat Ashkenazi stated during last week's analyst conference call, "As long as we see these attractive investment opportunities, we will continue to invest." This sentence represents the common stance of the management teams of the four companies. But as RBC's Jaluria said, "Investors are being forced to re-examine their timelines." The answer to this race between capital expenditure and AI revenue remains undecided.