Deep Dive
Chip selloff reflects profit-taking, not fundamentals
Josh Lipton hosts Market Domination on July 17, 2026 with roughly one hour until market close. The semiconductor index has entered bear market territory, with massive losses across the sector: SanDisk down 30% for the week, Global Foundries down 16%, and DRAM stocks down 16%. Jake Connelly attributes this partly to profit-taking after an extraordinary run where stocks like Micron surged 200-300%, but also to legitimate concerns about capital expenditure sustainability and whether AI spending will deliver promised returns. Bank of America analysts counter that semiconductors remain positioned for a strong year-end finish due to robust order backlogs and improving AI return on investment, with free cash flow expected to return in 2028. The market is priced for perfection in chips, meaning even modest disappointments trigger selloffs. However, the broader concern is whether companies will continue capex spending at current levels if ROI doesn't materialize as hoped.
Iran escalation threatens oil chokepoint and civilian targets
The Strait of Hormuz is entering its sixth consecutive day of conflict, with both sides expanding beyond military targets to civilian infrastructure. The US struck several bridges and Bucher city in Iran, which houses the country's only nuclear power plant—a significant escalation threatening environmental catastrophe. Iran responded by attacking desalination plants in Kuwait, critical facilities for a region dependent on these installations to produce freshwater. Saudi Arabia ships approximately 5 million barrels per day of oil through pipelines from the Persian Gulf to the Red Sea, making this chokepoint vital to global energy supply. India has warned shippers the route is too dangerous to traverse, and insurance costs are climbing. While shipping traffic has slowed significantly, it has not stopped entirely. The humanitarian risks from desalination plant damage and the geopolitical uncertainty around the nuclear plant create a volatile backdrop for energy markets, with oil prices rising and inflation concerns mounting.
Software losing pricing power as AI reshapes budget priorities
Rishy Galoria from RBC Capital Markets clarifies how IBM's warning about customer spending shifts signals fundamental disruption in the software industry. The key misconception is that AI will simply replace enterprise software; instead, AI and software are increasingly intertwined, and companies cannot achieve cost savings by wholesale switching to open-source AI without integrating them into existing software stacks. The evidence is stark: 77% of CIO contract negotiations now require renegotiation, indicating software vendors are losing pricing power. Tech layoffs this year have nearly matched the entirety of last year, forcing companies to reduce headcount rather than add staff—meaning software licensing models tied to per-user pricing no longer work. Pricing is shifting from annual per-user models to consumption-based and outcome-based pricing, where software vendors absorb efficiency gains from AI. Companies that have historically relied on price increases are suffering more than those previously underpriced. Microsoft emerges as the strongest positioned because its enterprise software stack is broad enough to integrate multiple frontier AI models, avoiding dependence on proprietary LLM advantage.
Chinese AI models narrow US technological lead
Dan Howie discusses China's rapid advancement in AI development, specifically Moonshot's new K3 model released as an open-weights architecture. K3 achieves frontier-level capabilities comparable to Anthropic's Claude 5 and OpenAI's GPT 5.6, yet operates far more cost-efficiently because it's open-weights—meaning users can download the model for free but pay for server access. This raises uncomfortable questions about whether American tech companies' massive capex investments are justified. DeepSeek previously demonstrated that Chinese engineers can train competitive models on lower-powered GPUs, suggesting a more efficient approach than Western labs. The gap between US and Chinese AI capabilities is shrinking, and the cost differential calls into question the ROI of current US spending levels. While data center infrastructure likely remains valuable, the question is whether $500 billion in VC funding flowing into AI (H1 2026 alone exceeded all of 2025) is necessary, or if Chinese approaches prove more efficient.
Anthropic positioned for IPO, OpenAI faces management red flags
Aman compares IPO prospects between Anthropic and OpenAI, revealing why Anthropic appears better positioned despite OpenAI's higher valuation. Anthropic's annual recurring revenue reached approximately $60 billion, growing at 10x year-over-year—growth unprecedented in public or private companies anywhere globally. More importantly, Anthropic is likely to achieve profitability in Q2, which was unexpected at year's start. OpenAI faces a critical impediment: significant management departures including CEO Sam Altman, founder Ilya Sutskever, founder John Schulman, super aligning co-lead Yan LeCun, COO Brad Lightcap, Mura Murati, Andrej Karpathy, Fuji Simo, and CMO Kate Roser. As Aman notes, when a team hasn't gelled and hasn't spent time together, taking a company public becomes very difficult. This instability makes public investors anxious about execution and vision. SpaceX's post-IPO stock performance offers context: priced at $135, it peaked at $200 but trades at $123.99, yet analysts like Louie Dealama from William Blair view this as normal volatility, noting that historically strong companies often trade below IPO price before recovering.
Gen Z savings crisis and upcoming earnings catalysts
Carrie Hannon highlights a concerning retirement savings trend: nearly half of Gen Z workers prioritize vacation savings over retirement, while 3 in 10 workers over 30 do the same. The issue is compounded by the fact that retirement seems like a distant concept to younger workers, causing them to miss critical years of compound investment growth. She recommends aiming for 12-15% of income toward retirement savings, combining personal contributions with employer matching (average 4% with some reaching 7%). Starting early with even small amounts leverages compound returns over decades. The broadcast concludes by previewing key earnings next week including Intel, Texas Instruments, GM, American Express, and Verizon, with Tesla and Alphabet reporting after Wednesday's close. Notably, a slowdown in capex could be a red flag for Wall Street, signaling weakening confidence in AI spending justification. Economic data week is light, with initial jobless claims and mortgage rates due Thursday and new home sales for June on Friday.