AI Won the Quarter. Big Tech Lost June.
GOOD MORNING, Wall Street closed out its strongest quarter in years, but the rally is getting harder to read. AI-linked stocks still powered the market higher, even as June exposed a sharp split between the companies spending on AI and the suppliers getting paid to build it. Meanwhile, jobs data showed a labor market that remains stable but increasingly difficult to move through. MARKETS | TLDR
AI TRADEThe AI Rally Has a Spending TestThe S&P 500 and Nasdaq finished the second quarter with their strongest percentage gains since 2020, powered in large part by the continued AI infrastructure buildout. But June complicated the story. The major indexes pulled back during the month as investors questioned how long hyperscalers can keep accelerating spending — and whether the returns will justify the capital going into chips, servers, power and data centers. That distinction matters because the AI trade is no longer one trade. Suppliers tied directly to infrastructure demand have continued to benefit, while several of the biggest AI spenders have been punished for the size of their commitments. The Magnificent Seven lost more than $2 trillion in market value during June, even as semiconductor names remained among the strongest parts of the market. The market is starting to separate companies collecting AI capex from companies funding it. The bull case is still intact: earnings remain solid, AI demand is real and broader market participation has improved. But the second half now carries a higher burden of proof. Reuters notes that investors will be watching whether AI spending can continue supporting profit growth while valuations remain elevated and interest rates stay restrictive. Strong demand alone may no longer be enough; the next phase is about economics. CHIP CAPEXSouth Korea Just Made a $520 Billion Memory BetSamsung Electronics and SK Hynix are planning one of the largest semiconductor capacity expansions ever, with a South Korean initiative centered on roughly $520 billion of investment in new chip plants. The plan includes four new fabrication facilities and is designed to sharply expand the country's memory-chip capacity as AI workloads drive demand for high-bandwidth memory and other advanced components. For markets, the scale cuts both ways. In the near term, it reinforces the idea that AI infrastructure demand remains strong enough to justify extraordinary capital spending. Memory has become a critical bottleneck in AI systems, and suppliers such as SK Hynix have benefited from tight capacity and pricing power. More supply could help unlock future AI deployment — while also creating a large new revenue pool for semiconductor equipment, construction, power and materials companies. The caveat is the memory industry's history. Samsung and SK Hynix have spent decades managing boom-and-bust cycles caused by capacity arriving after demand peaks. Reuters notes that the new buildout is already raising questions about what happens if hyperscaler spending cools before the fabs are fully utilized. The investment is a vote of confidence in AI demand, but also a reminder that today's shortage can become tomorrow's oversupply. HEADLINES
UPCOMING
DEEP INSIGTHSAI Spending, Earnings Hopes and the Fed Will Decide the Second HalfReuters' mid-year market framework is useful because it connects the three variables now doing most of the work in U.S. equities: whether AI capex can keep converting into earnings, whether corporate profits can clear a higher bar, and whether rates remain restrictive under the Fed. May JOLTS: More of the SameIndeed Hiring Lab adds context missing from the headline job-openings number: openings are still elevated, but quits remain low and workers are less confident about finding something better. That helps explain why the labor market can look healthy in aggregate while feeling increasingly stagnant to households. |