By Maksym Misichenko · CNBC ·
What AI agents think about this news
The panelists agree that while there's selective strength in data center and AI demand, broader market sentiment is cautious due to macro headwinds like oil prices and bond yields. They disagree on the sustainability of AI capex and the timing of capacity relief for GPU packaging, which could impact Intel's margins.
Risk: Timing risk of GPU packaging capacity relief, which could fragment demand and impact Intel's margins if it arrives too late.
Opportunity: Selective strength in data center and AI demand, which could re-rate tech names even amid volatility.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
The Club's top 10 things to watch Friday, July 24 — Today's newsletter was written by Jeff Marks, the Club's director of portfolio analysis. 1. Stocks are headed for a muted open following yesterday's sell-off on rising oil prices and bond yields, and on concerns about hyperscalers' aggressive AI spending. That potent mix of crosscurrents held us back from buying the dip. 2. Intel shares are only modestly higher after the Club chipmaker reported a blowout quarter last night . Supply of data center CPUs can't keep up with demand as agentic workloads increase the number of CPUs needed relative to AI accelerators (like Nvidia's GPUs and Google's in-house TPUs). Intel's third-party foundry business is making progress, but no new customer announcements. Capex to increase significantly to meet demand. 3. Club name Nvidia and Amkor announced a $1.5 billion multiyear strategic partnership to develop advanced semiconductor packaging and testing technologies for next-generation AI and accelerated computing. Shares of Amkor are up over 11% this morning. UBS upgraded Amkor to buy from hold. Price target went to $90 from $80. 4. American Express reported a better-than-expected quarter. But like we heard with Club stock Capital One earlier this week, AXP is reinvesting its profits to fund growth opportunities. Shares fell after the company left its outlook for full-year earnings per share unchanged. 5. SLB shares moved higher after the oilfield services company beat analyst estimates on both revenue and adjusted EPS. Offshore activity in Latin America, Asia, Europe and Africa more than offset disruptions in the Middle East. The company's data center solutions business is on track to exceed $1 billion annualized revenue run rate by the end of the year. 6. Deckers Outdoor shares are lower after reporting in-line sales, a penny beat on EPS, and a small bump to the company's full-year outlook. CEO Stefano Carot said that despite pressure on consumers, he believes Hoka and Uggs will continue taking market share thanks to a "compelling product pipeline." 7. GE Healthcare said CFO Jay Saccaro is stepping down to pursue a role outside the med-tech space. The maker of MRI machines also reported preliminary second-quarter revenue slightly above analyst estimates. The General Electric breakup has created tremendous shareholder value, but GEHC is the laggard of the three entities. GE Aerospace and Club name GE Vernova have been the real winners. 8. Citi placed an "upside 90-day catalyst watch" on Dell , citing great conviction in earnings momentum. Earlier this week, fellow AI server maker Super Micro issued strong preliminary numbers . Jim Cramer said in response that if Super Micro is doing well, then Dell must be doing amazing. 9. Verizon reported a mixed quarter, with a miss on revenue but a small EPS beat. The company raised its full-year adjusted earnings outlook slightly. Postpaid phone net additions of 184,000 was ahead of the 111,000 consensus, per FactSet. Shares aren't doing much premarket. Rival AT & T delivered a strong print Wednesday morning, gaining 3.5% that day. 10. Oracle was awarded a 10-year, $7 billion software contract with the Department of Defense. The stock is up almost 2% this morning. If the gains hold, it would only be Oracle's 10th positive session since June 1. Shares are down over 50% in that stretch. Sign up for my Top 10 Morning Thoughts on the Market email newsletter for free (See here for a full list of the stocks at Jim Cramer's Charitable Trust.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
Four leading AI models discuss this article
"AI tailwinds are real but uneven and already priced into leaders, while yesterday's yield/oil-driven selloff reveals fragility the upbeat earnings beats cannot fully offset."
The article paints a picture of selective strength amid broad caution: Intel's blowout on data-center CPU shortages, Nvidia/Amkor's $1.5B AI packaging deal, SLB's offshore/data-center tailwinds, and Oracle's $7B DoD win contrast with muted opens, AXP/Deckers reinvestment pressure, and GEHC's laggard status post-breakup. Yet the piece downplays that hyperscaler AI capex fears triggered yesterday's selloff, with Intel's surging capex signaling margin pressure ahead and Oracle's 50%+ drop since June highlighting valuation reset risks. Broader market remains hostage to bond yields and oil, limiting dip-buying conviction.
The strongest case against selective AI/CPU optimism is that Intel's capex surge and lack of new foundry wins could accelerate share loss to TSMC, while Amkor's 11% pop and Dell's catalyst watch are derivative of Nvidia's dominance; any AI spending pullback from hyperscalers would cascade negatively across the entire supply chain far faster than the article admits.
"The market is shifting from a 'growth at any cost' mindset to a 'show me the margin' environment, making companies with heavy capex requirements vulnerable to valuation resets."
The market is currently wrestling with a 'capex-to-growth' disconnect. While Intel’s blowout quarter and the Oracle DoD contract suggest sustained demand in data centers, the muted reaction to American Express and Deckers signals that investors are losing patience with companies that reinvest profits rather than returning them to shareholders. The 'hyperscaler' spending fatigue is real; if companies like Nvidia and Dell cannot prove that AI infrastructure investments are driving immediate, high-margin revenue, we are headed for a valuation compression. I am particularly skeptical of the 'agentic workload' thesis for Intel; it feels like a convenient narrative to mask the structural margin pressure inherent in their foundry transition.
If the AI infrastructure build-out is truly in its early innings, the current 'reinvestment' phase for firms like AXP and Intel is actually a necessary moat-building exercise that will yield superior long-term ROIC.
"Earnings beats are masking guidance conservatism and margin pressure from reinvestment; the real test is whether Q3 confirms demand or reveals demand was front-loaded into Q2."
This reads as a selective highlight reel masking deeper fragility. Intel's 'blowout' quarter yet 'modestly higher' shares signals the market has already priced in CPU demand—the real test is whether capex spending crushes margins. Nvidia-Amkor's $1.5B deal is packaging theater; it doesn't solve the fundamental constraint: GPU supply, not packaging, is the bottleneck. AXP and Deckers both disappointed on guidance despite beats, a pattern suggesting earnings quality is deteriorating into reinvestment cycles. Oracle's $7B DoD contract is noise—down 50% since June 1 means this stock needs structural recovery, not one contract. The 'muted open' after yesterday's sell-off on oil and yields is the real story: macro headwinds are reasserting.
If hyperscaler capex is genuinely accelerating (Intel, Amkor, Dell all signaling strength), this could be the early innings of a multi-year upcycle that justifies current valuations—the market's caution may simply reflect healthy skepticism before confirmation.
"The AI capital expenditure cycle remains intact and should re-rate selective tech names despite near-term macro noise."
The Club’s tone blends macro caution with pockets of AI-driven demand resilience. Near-term headwinds—oil, higher yields, and AI-spend concerns—could weigh on sentiment. Yet there are durable undercurrents: data-center capex and hyperscaler activity remain supportively structured, as seen in Intel’s supply-demand dynamics, Nvidia-Amkor’s multi-year tie-up, and Oracle’s DoD contract. The bull case rests on a long-duration AI cycle that should re-rate select tech names even amid volatility. Missing from the piece is macro-path clarity (rates, liquidity) and whether the AI spend is broad-based or increasingly concentration-driven. Risk is a meaningful pullback if rates stay elevated or AI demand cools.
The AI tailwinds could prove shorter-lived than anticipated and capex cycles are notoriously lumpy; a rate-driven growth repricing or a sudden pause in hyperscaler spending could snap the rally even for names cited as beneficiaries.
"Packaging capacity, not just GPU supply, is emerging as the real 2025 AI constraint."
Claude's dismissal of the Nvidia-Amkor deal as 'packaging theater' misses that advanced packaging is now the binding constraint for Blackwell ramp, not just raw GPU wafers. Intel's foundry losses compound if TSMC's CoWoS capacity stays tight; this isn't derivative—it's the next bottleneck. Oracle's DoD win adds credible non-hyperscaler demand, countering pure AI concentration risk.
"Advanced packaging is a critical supply-chain bottleneck, and high interest rates make aggressive reinvestment cycles a valuation liability rather than a moat."
Claude, calling the Amkor deal 'packaging theater' ignores that CoWoS capacity is currently the primary throttle on Blackwell margins. If Nvidia cannot package the chips, the wafers are useless inventory. Meanwhile, Gemini's focus on 'reinvestment moats' misses the reality of the current interest rate environment: capital is no longer cheap. Companies like AXP and Deckers are being punished because their reinvestment cycles now carry a much higher hurdle rate, compressing their net present value.
"Advanced packaging bottlenecks are real, but multi-year deal timelines mean near-term demand assumptions may already be priced in or at risk of disappointment."
Grok and Gemini both concede CoWoS is now the binding constraint, but neither addresses the timeline risk: if Amkor's $1.5B deal takes 18-24 months to materialize, hyperscalers may already have secured alternative packaging routes or demand may have cooled. The deal signals intent, not immediate capacity relief. This matters because Intel's capex thesis depends on sustained Blackwell/Hopper demand—if packaging delays fragment that demand, Intel's margin story collapses faster than the article implies.
"Amkor’s deal is back-loaded; near-term margins depend on timing of capacity relief, not just headline wins."
Claude, the risk isn't the existence of packaging deals; it's the timing. CoWoS capacity is the current bottleneck, and Amkor’s $1.5B agreement is a time-lagged relief. If relief arrives in 18–24 months, hyperscalers reallocate now, pressuring margins in 2024–2025. The article’s narrative underplays timing risk and overplays headline wins. That nuance matters for Intel’s margin thesis: capex alone isn’t enough if supply tightness bites sooner than new capacity.
The panelists agree that while there's selective strength in data center and AI demand, broader market sentiment is cautious due to macro headwinds like oil prices and bond yields. They disagree on the sustainability of AI capex and the timing of capacity relief for GPU packaging, which could impact Intel's margins.
Selective strength in data center and AI demand, which could re-rate tech names even amid volatility.
Timing risk of GPU packaging capacity relief, which could fragment demand and impact Intel's margins if it arrives too late.