The AGM focused on AI investment returns, product ramp-up, the company’s three long-term growth engines, chip smuggling, and China export issues. Jensen Huang said AI can generate profits as long as its output is valuable (e.g., automated code generation), highlighting Nvidia’s systems’ advantages in both cost efficiency and revenue potential. The company also reaffirmed its commitment to return 50% of free cash flow to shareholders. On products, Huang said Vera Rubin is now in full-scale production, with its GPU-CPU architecture addressing computing bottlenecks and expanding into the AI agent computing market.
U.S. major indexes were mixed, with the Nasdaq and Philadelphia Semiconductor Index underperforming as AI and technology stocks faced profit-taking after strong gains. Elevated valuations and rate sensitivity continued to pressure growth stocks. The S&P 500 declined modestly, while the Dow was more resilient, supported by stable cash flow and lower-valuation constituents, highlighting a clear rotation toward defensive sectors.
U.S. May PCE inflation rose about 4.1% YoY, with core PCE at roughly 3.4%, broadly in line with market expectations. Concerns over runaway inflation eased somewhat, but expectations for rapid Fed rate cuts have declined sharply. The 10-year Treasury yield initially rose before retreating, supporting a modest gain in the bond market. Investment-grade bonds remained relatively stable, with credit spreads largely unchanged. Meanwhile, the U.S. dollar index stayed firm, reflecting continued market pricing for higher interest rates over a longer period.
Micron Beats Expectations; 16 SCAs Reshape Business Model
Micron reported strong FY3Q26 results, with revenue of US$41.46bn (+74% QoQ, +346% YoY), surpassing the previous quarter’s record high and sending the stock sharply higher. Non-GAAP EPS reached US$25.11, about 20% above consensus. For FY4Q26, the company guided revenue to US$50bn (+21% QoQ, +342% YoY), with management attributing the outlook to AI-driven structural demand growth.
Goldman Sachs raised its forecast for U.S. electricity demand growth, driven by data center expansion and increased deployment of grid infrastructure and behind-the-meter (BTM) power generation (Note 1), much of which relies on natural gas. Lengthy permitting processes and supply chain constraints mean hyperscalers and data center operators must address both near-term (five-year) and long-term (10–15 year) power supply challenges. While nuclear power offers superior generation capacity, it requires substantial upfront investment and years before becoming operational. For hyperscalers, speed to deployment matters more than minimizing electricity costs, making energy planning increasingly complex.
JPMorgan and Goldman Sachs both identify copper as a critical material for power generation, grid infrastructure, and renewable energy. AI-driven investment in grid upgrades, power infrastructure, and energy storage is expected to accelerate demand, raising the risk of a structural copper supply shortage. JPMorgan estimates the mining industry will need to invest about US$350bn by 2040 to meet growing copper demand. However, limited new mine supply and constraints on expanding existing operations could make copper-focused miners and potential M&A activity key market themes.
