Decoding the Big Tech Earnings "Super Bowl": A Framework for Identifying Winners and Losers in the $725 Billion CapEx Era
On April 29–30, 2026, Alphabet, Amazon, Microsoft, Meta, and Apple released their quarterly earnings one after another. These 48 hours, described by Bloomberg Tech as the "Super Bowl of Tech Earnings," marked a turning point where the "returns" on AI investment were seriously questioned for the first time.
The total capital expenditure plans for the four companies (Alphabet, Amazon, Microsoft, and Meta) for the full year 2026 will reach a maximum of $725 billion, with the majority allocated to AI data center-related projects. The biggest concern for investors is whether this massive investment is yielding concrete revenue.
In this article, we will analyze the earnings of each company across the board and organize the structural points of contention regarding winners and losers that investors and business professionals should grasp.
1. Alphabet: The Structure of a "Clear Winner"
1-1. Google Cloud: 63% growth, $460 billion in backlog
Alphabet exceeded analyst expectations for both EPS and revenue in the first quarter, with Google Cloud revenue recording 63% growth year-over-year. The cloud backlog nearly doubled from the previous quarter, exceeding $460 billion.
This 63% growth rate is the fastest in Google Cloud history, and it is worth noting that it was achieved on an annual run-rate basis of over $70 billion.
Even more noteworthy is the improvement in margins. The operating margin for the cloud business expanded rapidly from 9.4% in the same period last year to 32.9%.
1-2. Why is Alphabet the only "winner"?
The reason the market valued Alphabet is that the causal relationship between AI investment and revenue was the clearest. CFO Anat Ashkenazi stated, "We are seeing unprecedented internal and external demand for AI computing resources."
Revenue from products based on GenAI models grew by approximately 800% year-over-year. Additionally, CEO Sundar Pichai stated that paid monthly active users for Gemini Enterprise grew 40% quarter-over-quarter and that they have secured contracts with major brands.
Pichai said, "The number of contracts in the $100 million to $1 billion range has doubled year-over-year, and we have signed multiple contracts worth over $1 billion."
2. Amazon: AWS fastest growth in 15 quarters, but FCF drops sharply
2-1. 28% growth in AWS and the rise of in-house chips
Amazon reported first-quarter EPS of $2.78 and revenue of $181.5 billion (up 17% year-over-year), both exceeding analyst expectations.
AWS revenue was $37.6 billion, growing 28% year-over-year and exceeding the consensus estimate of $36.64 billion. This is the fastest growth rate in 15 quarters.
What is noteworthy is the rapid growth of their in-house silicon. Amazon's in-house chip business (Graviton, Trainium, Nitro) has an annual revenue run-rate exceeding $20 billion and is recording triple-digit growth rates year-over-year.
2-2. Sharp decline in free cash flow is a cause for concern
Capital expenditure reached $44.2 billion for the quarter, a sharp increase from $25 billion in the same period last year. Free cash flow for the past 12 months dropped sharply from $25.9 billion to $1.2 billion. How long this massive investment will continue in the long term is the biggest concern for investors.
After the earnings announcement, the stock price fell more than 3% in after-hours trading.
3. Microsoft — AI ARR at $37 Billion, but with "Modest Acceleration"
In Microsoft's third quarter (calendar Q1 2026), the annual revenue run rate for its AI business reached $37 billion, marking 123% growth year-over-year.
Paid seats for M365 Copilot surpassed 20 million. The ARR for LinkedIn's agentic AI products also exceeded $450 million, indicating rapid enterprise adoption.
However, the stock fell 2.5% as the description of Azure's growth acceleration as "modest" for the second half of the year failed to meet market expectations.
4. Meta — Largest Increase in CapEx, but Lacking "Proof" of AI Revenue
4-1. Revenue growth, but missing "that one extra number"
Meta's earnings were received coldly by the market, with the stock falling over 5% after the company provided guidance suggesting flat revenue growth in Q2.
As a Bloomberg Tech anchor aptly pointed out, the problem is the lack of "one more concrete metric showing that AI investment is paying off." While ad pricing rose 12% and revenue was in line with consensus, the company failed to present a clear growth engine like Alphabet's Google Cloud.
4-2. "Organic" increase in CapEx stands out
Meta raised its full-year 2026 capital expenditure guidance to $125–$145 billion, an upward revision from the previous $115–$135 billion. The company cited rising component costs and the addition of data center capacity for AI workloads as the reasons.
One investor noted on the program that "Meta does not yet have a cloud business." While cloud providers can visualize AI sales as revenue, the fact that Meta's massive investments only return indirectly through improvements to its own advertising business is the fundamental reason for the stock price decline.
Meta CEO Mark Zuckerberg stated, "We are increasing infrastructure CapEx. Most of that is due to component costs, particularly rising memory prices," while emphasizing that his conviction in AI investment remains unshaken.
5. Why NVIDIA fell — The rise of custom silicon
Despite the increase in capital expenditures, NVIDIA's stock price fell. The reason lies in the fact that companies aggressively highlighted the importance of their own in-house designed chips (custom silicon).
An analyst on the program analyzed that "as Google mentioned custom silicon and the shift toward inference progresses, the view has spread that companies like Broadcom will benefit more."
Google: Expanding availability of TPUs to external customers
Amazon: In-house chip business ARR over $20 billion, Trainium 3 shipments started
Microsoft: Maia chip under development
Meta: MTIA chip under development
However, NVIDIA GPUs remain indispensable for LLM training, and there is a strong view that NVIDIA will continue to be the "base player" in the medium to long term.
6. Anthropic: $900 Billion Valuation, Potential to Surpass OpenAI
On the same day as the earnings rush, another major piece of news broke.
Anthropic has begun considering a new funding round at a valuation exceeding $900 billion. If realized, it would surpass rival OpenAI to become the world's highest-valued AI startup.
The size of the round is expected to be between $40 billion and $50 billion. The company's annual revenue run rate has reached $30 billion, compared to approximately $10 billion in revenue last year.
Behind this are massive investments from strategic partners. Amazon has pledged up to $25 billion in investment, and Google is planning an investment of up to $40 billion. Bloomberg reports that Anthropic is considering an IPO as early as October.
7. Samsung: Profits Up Over 8x on AI Memory, Yet a "Duality" Emerges
7-1. Operating Profit Up 756%, Exceeding Full-Year 2025 in a Single Quarter
Drawing attention alongside the Big Tech earnings were the results from Samsung, which sits "upstream" in the AI infrastructure chain.
Samsung Electronics reported revenue of 133.9 trillion won (approximately $90 billion) and an operating profit of 57.2 trillion won for the first quarter, both setting new all-time highs. Operating profit increased by over 750% compared to the same period last year.
The operating profit for the first quarter alone exceeds the 43.6 trillion won for the full year of 2025. This figure speaks to how explosive the demand for AI memory has become.
7-2. A Structure Where the Chip Division Accounts for 94% of Profits
The chip division accounted for 94% of operating profit, while profits in the mobile and home appliance divisions fell by approximately 40% year-over-year. Semiconductor operating profit reached 53.7 trillion won, a staggering 48-fold increase from 1.1 trillion won in the same period last year.
Samsung has begun shipping HBM4 and SOCAMM2 memory chips for NVIDIA's new AI accelerator platform, "Vera Rubin."
7-3. Memory Shortage Expected to Continue Through 2027
Samsung executives stated during the earnings call that available memory supply is significantly below customer demand. They noted, "The demand fulfillment rate is at an all-time low, and customers concerned about supply shortages are pulling forward demand for 2027," suggesting the gap between supply and demand is expected to widen further in 2027.
An analyst at Counterpoint Research pointed out, "Memory is establishing itself as an independent factor that determines the success or failure of AI infrastructure."CNBC
7-4. Implications for Investors
Samsung's earnings clearly illustrate how the surge in Big Tech capital expenditure is rippling through the upstream supply chain. Just as Meta and Microsoft cited "rising memory prices" as a reason for increasing their CapEx, the rising costs of building AI infrastructure are squeezing the profit margins of downstream companies. Meanwhile, Samsung faces a duality: it is both the greatest beneficiary of upstream price hikes and the greatest victim of downstream cost increases.
If the DRAM price growth rate falls below 20% and HBM growth slows to under 30% over the next two quarters, the profit peak may be near. On the other hand, analysts point out that if losses in the mobile division are kept within 1 trillion won and chip profits remain above 50 trillion won, the forward P/E ratio of approximately 4.9x remains attractive.
8. Apple: March Quarter Revenue Record Broken, China Recovers Sharply by +28%
8-1. Double-digit growth across all segments and regions
Apple reported EPS of $2.01 (vs. $1.95 expected) and revenue of $111.8 billion (vs. $109.66 billion expected) for the second quarter of fiscal year 2026 (calendar Q1 2026), beating analyst estimates on both counts. Revenue was up 17% year-over-year, marking a record for the March quarter.
Gross margin reached 49.3%, an improvement from 47.1% in the same period last year.
By segment:
iPhone: $56.99 billion (vs. $46.84 billion last year, +22%) — March quarter record
Mac: $8.4 billion (vs. $7.95 billion last year) — Strong performance from MacBook Neo
iPad: $6.9 billion (vs. $5.56 billion last year) — Launch of M4-powered iPad Air
Services: $30.9 billion (vs. $26.6 billion last year, +16%) — All-time high
8-2. China's rapid recovery is the highlight of the earnings
Geographically, the company achieved year-over-year revenue growth in all regions. Of particular note is Greater China, where revenue increased 28% to $20.5 billion from $16 billion in the same period last year.
Tim Cook stated, "I am proud to report a record-breaking March quarter, driven by strong demand for the iPhone 17 series—the most popular lineup in Apple history—and double-digit growth across all regions."
8-3. Q3 guidance is a surprise
Apple guided for revenue growth of 14-17% year-over-year for the June quarter (Q3). The analyst consensus was 9.5% growth (approximately $103 billion), making this a significant positive surprise.
The board of directors approved a new $100 billion share buyback program and increased the quarterly dividend by 4% to $0.27.
8-4. AI Strategy: Gemini partnership with Google, Siri overhaul
Apple announced a multi-year partnership with Google at the beginning of the year, adopting the Gemini AI model as the foundation for the next-generation Siri. Cook stated during the earnings call, "Our collaboration with Google is progressing well, and we are also satisfied with the initiatives we are pursuing independently."
Google Cloud Chief Thomas Kurian said, "As Apple's preferred cloud provider, we are co-developing next-generation Apple Foundation Models based on Gemini technology. These models will power future Apple Intelligence features, including a more personalized Siri arriving later this year."
However, like other consumer electronics companies, Apple is facing supply chain constraints due to a global memory shortage caused by AI demand. While Cook noted that they "exceeded guidance despite supply constraints," the potential for rising memory costs to impact future margins remains a risk factor.
8-5. CEO transition: John Ternus to take office on September 1
Apple announced Tim Cook's retirement on April 20. Cook will become Executive Chairman effective September 1, and John Ternus, SVP of Hardware Engineering, will become the new CEO.
This earnings report served as the first major sentiment test for investors ahead of the leadership transition, but by beating expectations on revenue, EPS, and gross margins, and providing strong Q3 guidance, it fostered a sense of positive confidence.
Furthermore, R&D spending reached $11.4 billion, a 33% increase year-over-year, indicating a steady expansion of investment in AI-related research and development.
9. Notable Side Stories — Stripe × Gemini, 137 Ventures, SpaceX IPO
9-1. Stripe and Google Partnership
Stripe announced a partnership with Google to provide payment features powered by Gemini AI. Co-founder John Collison predicted that AI agents would gradually climb the "trust curve," stating that AI commerce would begin with small decisions like "purchasing a domain name."
9-2. 137 Ventures — 100x Returns on SpaceX
137 Ventures raised $700 million across two new funds, bringing its assets under management to $15 billion. The firm first invested in SpaceX in 2010 and has written over 20 checks since then. Its position in SpaceX has reached over $10 billion, with early investments reportedly yielding roughly 100x returns.
Regarding reports that SpaceX might hold an IPO in late June, the founder expressed a long-term holding stance, stating, "I am looking forward to the next 20 years of SpaceX's business." On the other hand, he also touched on the risks of fraudulent transactions via SPVs in the private market, warning, "A year after the IPO, some people will find out that the shares they thought they owned didn't actually exist."
