Yahoo Finance
Yahoo Finance4d ago
Finance

Yahoo Finance Live: Daily Market Coverage - July 9, 2026 3PM - 5PM (ET)

120 min video5 key momentsWatch original
TL;DR

Markets rallied on AI optimism despite Iran tensions, with NASDAQ up 1.3% and semiconductors surging as companies rein in AI spending while infrastructure demands drive long-term growth.

Key Insights

1

Token pricing model brokenToken-based AI pricing collapsed. Coinbase cut its AI bill 50% in months by routing simple queries to cheaper models, proving the January 30 agentic AI boom is forcing companies to manage costs ruthlessly.

2

Behind-the-meter powerData centers cannot wait 5-7 years for grid connections, so Meta, Oracle, and Amazon are building on-site power plants using natural gas turbines—which are completely sold out through 2030.

3

100 gigawatt shortfallBank of America projects a 100-gigawatt US electricity shortfall by 2030 driven by AI data center demand growing 30% annually, forcing utilities and ratepayers to foot infrastructure bills when contracts get cancelled.

4

Meta first user declineMeta's sequential decline in users for the first time ever is structural red flag despite lower AI valuations versus Microsoft, which has no leading large language model of its own.

5

China memory market shiftChina's DRAM share sits at 5% today but could explode to 30% as Apple sources Chinese memory and Iranian oil flows east, reshaping semiconductor and geopolitical leverage.

6

Retirees spend less over timeRetirees' spending actually decreases year-over-year in retirement, contradicting traditional 4% withdrawal rate models, and AI tools gave conflicting emergency fund advice ($21k to $38k) based on demographic prompts.

Deep Dive

Market rally on AI cost discipline, not euphoria

Markets closed higher on July 9, 2026, with the NASDAQ up 1.3%, S&P 500 up 0.8%, and the Russell 2000 jumping 1.3%. Tech led with a 2% gain, driven by semiconductor strength—Micron up 6%, ARM up 10%, Lumentum up 13%. Meta jumped nearly 5% and Tesla rose 3%. The driver wasn't fresh AI optimism but rather pragmatic cost control. Dan Niles from Niles Investment Management explained that starting in January when agentic AI launched, companies burned through budgets fast—Uber blew its yearly token budget in four months. By June, companies snapped into action, routing simple queries to cheaper open-source models while reserving expensive models for complex tasks. Coinbase CEO cut its AI bill nearly 50% while actually increasing token production. This rotation is healthy, not a correction, Niles argued. The real growth engine—agentic AI formalizing on January 30—remains intact and will force massive memory upgrades with Nvidia's Vera Rubin platform requiring 3.5x to 10x more memory than current systems.

Data center boom collides with power grid reality

Bank of America projects a catastrophic 100-gigawatt US electricity shortfall by 2030, driven by AI data center demand growing 30% annually. The problem is brutal: natural gas turbines are completely sold out through 2030, and data centers take 5-7 years to connect to the grid with stable loads. So Meta, Oracle, Amazon, and others are building on-site power plants—behind-the-meter solutions—to avoid waiting. Communities are pushing back hard. One host's sister in Loudon County, Virginia's data center alley, saw her electricity bills triple. Utilities and ratepayers bear the cost when contracts get cancelled, not the tech companies. There are real constraints: power, supply chain bottlenecks, and labor. NTG Global Data Centers, the third-largest globally with 160 facilities, hasn't used water in designs since 2016. Data centers contributed 90% of US net new GDP in 2025 and generated 162 billion in tax revenue from 2017-2023. But the overbuilding risk is real—Meta's already selling excess compute capacity and construction is slowing. If this continues, ratepayers get stuck with the bill.

Meta vs. Microsoft: valuations mislead on AI competitive reality

Niles makes a contrarian call: Meta is a better AI play than Microsoft despite being downgraded on weak AI agent adoption. Meta's Llama open-source model is priced 75% less than Anthropic or OpenAI, giving it pricing power. Microsoft has 78 billion in cash but no leading large language model—it owns 27% of OpenAI and relies on that relationship. Anthropic dominates enterprise, Google owns consumer. The structural risk for Microsoft is severe: Starbucks just cut 400 million annually in software spending by deploying AI models. That's the real threat to Microsoft's core business. Meta's valuation, while pressured year-to-date, reflects a company regaining open-source leadership after DeepSeek's temporary dominance. The gut punch is Meta's first-ever sequential decline in users across its entire family of apps. That's not noise—there are only two ways to make money in advertising: user count and revenue per user. Both are now at risk. Niles still sees upside, but the user problem is structural, not cyclical.

Iran sanctions and oil: world has adapted, markets yawn

US-Iran exchanges continued for a second day, with ships in the Strait of Hormuz dropping from 50-100 per day pre-war to teens per day now. Yet Brent crude rose only 6% over a week and stayed well below 100 per barrel. Why the muted reaction? The world has adapted. China is deploying massive strategic petroleum reserves, and most Iranian oil heads to China anyway—which doesn't care about US sanctions. The US reversed its plan to lift sanctions, putting them back on, but market impact is minimal. The risk assessment hinges on conflict duration and Iran's willingness to toll the strait. If a full-scale war erupts and the strait closes, that's a different story. For now, it's manageable. Alpine Macro and Oxford Economics say the world's rerouting mitigates impact, and China's been reducing oil demand for over four months without visible shortages.

Housing market shows resilience; young adults stuck with parents

Housing market data revealed mixed signals. Median US home price hit nearly 441,000 dollars at a 6.5% 30-year mortgage rate. The policy tailwind is real: a bipartisan housing affordability bill set to become law Friday, July 10, represents the first federal attempt in 30 years to address the crisis. Nearly 20% of Americans ages 25-34 live with parents or grandparents—a record. Yet inventory is much higher than during COVID, and wages are outpacing home prices for the first time in years, signaling slow affordability improvement. New home sales are elevated versus existing home sales. The catch: housing starts remain near early COVID recession levels, and builders lack financial incentives to build more. Regional variation matters—Florida showed unexpected demand strength with year-over-year inventory declines after mortgage rates moved lower, while Washington state lagged due to lower AI investment presence. Analysts were skeptical the affordability bill would meaningfully increase supply.

Luxury hotels and consumer brands defy macro headwinds

Nell Hotels announced global expansion, launching properties in New York (134 rooms) and Aspen (92 rooms) under the Nell Hotels banner while keeping the iconic Little Nell Aspen as its flagship. CEO Jeff Tusano, who started at 16 working at Howard Johnson's, emphasized the owner-operator model ensures quality over scale and cuts bureaucracy that hurts customer experience. The global luxury hotel market is projected to reach 350 billion by 2027, growing at 6.2% from 2022-2027. Affluent travelers are recession-resistant. Levi's delivered seven consecutive quarters of top and bottom line beats with guidance raised again. Women's business surged 11%, white denim jumped 70%, and the company expanded addressable market 15 times through denim lifestyle positioning. One-third of growth came from that expanded market. Levi's now greets customers with women's assortment in 70% of US full-price stores and operates across price tiers: value, Red Tab mid-tier, and Blue Tab premium at 200 to 250 dollars per jacket. The brands winning are those meeting actual consumer needs with authentic positioning.

Takeaways

  • AI spending is normalizing fast—companies routing cheap queries to open-source models while reserving expensive models for complex tasks, cutting costs while maintaining productivity gains.
  • Data center on-site power plants are becoming standard because grid connections take 5-7 years; expect utilities and ratepayers to absorb infrastructure costs if demand disappoints.
  • Meta's structural user decline is the real risk, not valuations; Levi's and other brands winning by expanding addressable markets and meeting actual consumer needs.
  • Iran tensions are priced in with manageable oil impact due to China's reserves and rerouting; the real constraint is electricity shortfall by 2030, not geopolitical supply disruption.

Key moments

5:00Token-based AI pricing breaks under cost pressure

Coinbase CEO reported cutting AI spending bill by almost 50% despite producing more tokens

15:00100-gigawatt electricity shortfall projected by 2030

Bank of America predicts US could be short 100 gigawatts of electricity by 2030

20:00Meta's first sequential user decline in company history

The number of users across their entire family of apps was actually down sequentially for the first time ever in the company's history

30:00Strait of Hormuz shipping collapses amid Iran tensions

We're seeing about ships passing each day in the teens. That's down from 50 just a few days ago, not to mention over a hundred before the war happened

40:00Levi's expands addressable market 15 times with women's pivot

We've expanded our addressable market by 15 times

Get AI-powered video digests

Follow your favorite creators and get concise summaries delivered to your dashboard. Save hours every week.

Start for free