On July 22, 2026, the first wave of U.S. stocks Mag7 earnings season landed in an extremely divided manner:
Same theme (AI), same batch of institutional investors, same week, two completely opposite stories.
Alphabet Q2 revenue $119.8B (+24% YoY), Google Cloud growth +82% to $24.8B, hitting a record high; but full-year capital expenditure guidance rose to $195–205B, free cash flow turned negative at -$5.9B, leading to a sharp drop after hours. Tesla Q2 revenue $28.2B, deliveries of 480,126 vehicles both hit records, but EPS was only $0.33 (expected $0.53), profits plummeted 57% YoY, gross margin 16.8%.
Meanwhile, tensions in the Middle East continued to rise, crude oil surged strongly this week, and energy stocks broke out first. Flash PMI Manufacturing 53.9 / Services 51.2, the job market remained robust, and the macroeconomic foundation stayed solid.
Four main themes resonated intensely within the same week; any single one alone would have been sufficient to drive the weekly market. This article deconstructs the core investment logic of US stocks for the week from four dimensions.
Alphabet delivered a "beautiful-looking" earnings report this quarter:
Total Revenue $119.8B, +24% year-over-year, beating analyst expectations
Google Cloud Revenue $24.8B, +82% year-over-year, beating expectations and hitting a record high—cloud operating margin reached 35.6%, triple that of the same period last year
Search Ads grew steadily, and the launch of AI Overview did not erode core advertising revenue
But there is a hidden "landmine" in the earnings report:
Metric | Value | Market Interpretation |
Full-year capex guidance | $195–205B | Significantly higher than market expectations |
Free Cash Flow | -$5.9B (turned negative) | First time turning negative in a quarter in history |
Management Commentary | Spending in 2027 "will increase significantly" | Market worries about no ceiling |
Cloud backlog | $514B | The most important support for bulls |
Both interpretations have support—this is the core divergence behind GOOGL's sharp after-hours drop this week: Bulls believe that the 35.6% operating margin and $514B backlog prove returns are being realized; bears worry that if AI demand slows, astronomical capex will become sunk costs. The Mag7 earnings discussion has evolved from "whether to grow" to "whether the cash burn is worth it."
Tesla's Q2 represents another form of divergence:
Revenue $28.2B, hitting a record high and exceeding analyst expectations
Deliveries 480,126 units, beating expectations by ~18% and setting a new single-quarter record
But EPS $0.33, only 62% of the market expectation ($0.53)
Profits down 57% YoY, with gross margin at 16.8%, under continued pressure
Metric | Q2 Data | Interpretation |
Revenue | $28.2B (Record High) | Surge in deliveries; demand is not the issue |
EPS | $0.33 (vs. Expected $0.53) | Profit quality significantly below expectations |
Gross Margin | 16.8% | Cost of trading price for volume continues to manifest |
Profit Change | -57% YoY | Volume up, profits down; profitability model under pressure |
This earnings report conveys one message to the market: Tesla's current competitive advantage lies in "being able to sell," but whether it can "sell for profit" is the most critical variable for the next phase.
Robotaxi is the only path capable of rewriting this narrative—Miami expansion, Cybercab mass production launch. If autonomous driving software service revenue grows proportionally, the profit model will shift from "automaker" to "AI service platform," at which point the gross margin logic will be entirely different.
GOOGL has set its full-year capex guidance at $195–205B; previously, Microsoft also disclosed AI infrastructure investment plans of a similar scale. What does this figure imply?
The combined AI capital expenditure of the top three global cloud providers is expected to exceed **$600B** by 2026
Every dollar requires compute chips, power infrastructure, cooling systems, and storage memory—these four categories of companies are the most certain beneficiaries
Beneficiary Chain | Representative Tickers | Logic |
Compute Chips | NVDA | GPU procurement is the largest single item in cloud providers' AI investments |
AI Power | GEV (GE Vernova) | Surging electricity demand from data centers, tight supply of power equipment |
Storage/HBM | MU (Micron, SK Hynix ADR pending listing) | AI chips like Blackwell must be equipped with HBM memory |
Cooling/Networking | SMCI, ANET | Increased data center density drives simultaneous growth in cooling and networking investments |
GE Vernova saw a significant surge in orders this Q2, with power demand from AI data centers becoming the largest incremental source of its orders.
Cloud providers are spending over $195B to build data centers, each requiring tens to hundreds of megawatts of power supply. It takes time to expand supply-side production capacity for power equipment, while demand has risen at a nearly vertical slope—this supply-demand gap has created a high-prosperity cycle for GEV's order backlog.
Unlike computing chips, there is no controversy over "whether AI demand is real" for power infrastructure: data centers consume electricity once built, and the order cycle for power equipment is longer with higher visibility. This is the core logic behind including GEV in the AI sector this week.
Tensions in the Middle East are the core driver behind this week's oil price rally: US-Iran negotiations have stalled, reigniting market fears of supply disruptions, while Brent crude prices remain at high levels.
As the largest non-AI catalyst in the US stock market this week, crude oil has allowed energy stocks to benefit first amid broader market divergence.
XOM (Exxon Mobil), the world's largest integrated oil company, directly benefited from rising oil prices this week:
Perspective | Analysis |
Geopolitical Risk Hedge | If Middle East tensions persist, oil prices will remain supported at high levels; XOM is the most direct beneficiary |
Cash Flow Stability | In a high oil price environment, XOM's free cash flow is substantial, supporting shareholder returns |
Portfolio Diversification | With AI themes dominating the market, energy stocks provide portfolio buffering through uncorrelated assets |
Valuation Perspective | Compared to the high valuations of the AI sector, energy stocks' P/E ratios offer more defensive value |
Adding XOM to this week's allocation follows a two-layer logic: first, it purely benefits from the short-term catalyst of rising oil prices; second, it serves as a defensive position for portfolio diversification against the backdrop of diverging valuations among the Mag7.
Next week (Jul 28–Aug 1) will be one of the most event-dense weeks of Q3 2026:
Date | Event | Key Highlights |
Jul 28–29 | FOMC Interest Rate Decision | CPI has set a dovish tone; focus shifts to whether Powell's wording officially opens the window for a September rate cut |
Jul 29 | Mag7 Wave 2: MSFT + META | Azure AI growth and Copilot commercialization; whether Meta's ad revenue can maintain high earnings visibility |
Jul 30 | Mag7 Wave 3: AMZN + AAPL | AWS growth and profitability; Apple Intelligence penetration during the iPhone 17 upgrade cycle |
Based on current composite signals, the July FOMC is most likely to keep interest rates unchanged. The real market focus lies in whether Powell's wording at the press conference is significantly dovish, and whether it clearly provides forward guidance for a rate cut in September.
If the rate cut path becomes clearer after the FOMC, the benefit to discount rates for high-valuation tech stocks will once again trigger a valuation repair rally for growth stocks.
Microsoft: Whether Azure AI growth continues its accelerating trend; whether Copilot commercial revenue can become a significant new growth driver
META: Whether ad revenue can exceed expectations again driven by AI-powered precise targeting; whether the commercialization path of the Llama open-source strategy is clear
If both exceed expectations, the divergence on "earnings quality" within the AI theme will narrow, and the market focus may shift from "whether AI investment is worth it" back to "AI monetization capability."
Holding | Sector | Weekly Change | YTD Change | Core Logic |
NVDA NVIDIA | AI Computing / Chips | +3.2% | +12% | GPU procurement is the largest single item in cloud providers' $195B+ spending; Blackwell shipments continue to accelerate |
GEV GE Vernova | AI Power | -2.8% | +56% | Significant growth in Q2 orders; surging power demand from AI data centers, high prosperity in power infrastructure |
GOOGL Alphabet | Internet / Cloud Computing | -10% | +1.7% | Cloud +82% exceeded expectations; but capex $195–205B, free cash flow turned negative, sharp drop after hours |
TSLA Tesla | EV / Consumer Tech | -16% | -28.9% | Record revenue, profit -57%; Robotaxi expansion is the only path to valuation restructuring |
XOM ExxonMobil | Energy | +5.2% | +30.1% | Tensions in the Middle East, oil prices at highs; biggest beneficiary of non-AI catalysts this week |
Q1: GOOGL dropped significantly in after-hours trading. Is it oversold or fairly priced?
This depends on how you view the return cycle of the $195–205B capex. Bears argue that negative free cash flow signifies "trading profits for growth," with the risk that these sunk costs cannot be recovered if AI demand slows down. Bulls point to the $514B cloud order backlog and a 35.6% operating margin—returns are already being realized, merely obscured by short-term expenses. The core question is not "will it rise," but "when will the market reprice this return."
Q2: TSLA's profit dropped by 57%. Is it still worth holding?
This depends on whether you are buying "TSLA as an automaker" or "TSLA as an AI service platform." Based on current earnings quality, TSLA's valuation as an automaker is clearly overpriced; however, if the proportion of Robotaxi software service revenue grows, the earnings model will shift completely, making the current loss period a strategic entry point. The progress of Cybercab mass production and commercialization data from the Miami pilot are the most critical indicators to watch in the coming quarters.
Q3: AI is burning through $195B+, who will ultimately win this "arms race"?
The winners on the supply side are often not end users, but infrastructure providers—this is the origin of the "shovels and picks" logic. Every dollar spent by cloud providers flows into computing chips (NVDA), power equipment (GEV), and storage memory (MU/SK Hynix). These companies have high demand certainty and relatively reasonable valuations, making them a better risk-reward allocation direction within the AI theme.
Q4: Is the surge in oil prices a sustained trend or a short-term disruption?
Whether geopolitical risks in the Middle East persist is the core variable for judging oil price trends. If US-Iran negotiations remain stalled, support for high oil prices may continue; if negotiations show positive progress, oil prices face the risk of a rapid pullback. For energy stock allocations, it is recommended to treat them as "hedging positions" rather than "thematic positions"—the proportion should not be too heavy, but against the backdrop of diverging profitability in the AI theme, energy stocks serve as an effective portfolio buffer.
Q5: MSFT and META earnings next week, which is more critical?
Each has its own focus. Microsoft's key highlight is the growth rate of Azure AI—if it can maintain or accelerate, it will directly validate the logic of "AI service monetization," boosting confidence in the cloud sector, which has been suppressed by capex concerns; META's key highlight is the quality of advertising revenue growth. If AI-driven precise targeting continues to boost advertising ARPU, META's valuation logic will be further strengthened. If both exceed expectations, the market may refocus on AI monetization capabilities rather than the speed of AI cash burn.
Q6: How can ordinary users participate in U.S. stocks RWA with the lowest threshold?
MEXC U.S. Stocks RWA is now officially live, with a minimum investment of just 5 USDT—flexible allocation via fractional shares, no need to open a traditional U.S. stock account, zero fees with direct connection to traditional brokers, access to 7,000+ U.S. stocks and ETFs anytime, smooth liquidity, and a seamless trading experience.
Disclaimer: This content is based on public market information as of July 23, 2026, and has been independently compiled by the MEXC US Stock Spot Team. It is for reference only and does not constitute investment advice. The market involves risks, so please invest with caution.