Morning Bid: AI horror stories

• Nvidia posted better-than-expected quarterly results and forecasts, but investor fear overshadowed optimism, driving tech stock declines amid questions about whether markets are experiencing an 'AI bubble'
• South Korea's KOSPI index surged over 48% year-to-date, making it 2026's best-performing equity market thus far
• Trump administration signaled that hyperscalers must provide their own power for AI data centers to avoid stressing the grid and raising consumer electricity costs, highlighting physical infrastructure bottlenecks as a potential constraint on the AI revolution

Market Summary: AI Concerns Trigger Tech Selloff

Key Market Movements

Markets experienced significant volatility this week as AI-related anxiety weighed on investor sentiment. The S&P 500 and Nasdaq both declined despite strong earnings from major tech companies. U.S. stock index futures continued sliding Friday, with the Nasdaq facing its steepest monthly drop since March 2025.

Major Developments

Nvidia posted better-than-expected Q1 results and raised guidance for current quarter revenue, yet failed to lift market sentiment. The $4.5 trillion chipmaker's strong performance was overshadowed by broader AI concerns.

AI Debate: A 7,000-word report from research firm Citrini sparked fears about massive job losses and potential deflationary pressures from AI adoption. Citadel Securities issued a rebuttal, questioning the doomsday scenario.

Korea Market Surge: South Korea's KOSPI index emerged as 2026's best performer, up over 48% year-to-date.

M&A Activity: The battle between Paramount Skydance and Netflix for Warner Bros Discovery reportedly concluded, with Paramount offering a revised $31-per-share bid.

Energy and Commodities

Brent crude traded above $71/barrel amid U.S.-Iran nuclear negotiations in Geneva, with talks continuing next week. OPEC+ is expected to announce a 137,000 barrel-per-day production increase at Sunday's meeting.

Policy Concerns

President Trump's State of the Union address (record 1 hour 47 minutes) stated hyperscalers must "provide for their own power needs" for AI data centers to avoid grid strain. Infrastructure bottlenecks pose significant challenges to AI expansion, highlighting physical limitations beyond theoretical concerns.