Fed Holds Off on Changes, Maintains Outlook for Two Rate Cuts in 2025 — For Now The Federal Reserve kept interest rates steady on Wednesday and reaffirmed its forecast for two rate cuts in 2025, despite growing economic uncertainty fueled by President Donald Trump’s new tariffs. “Uncertainty today is unusually elevated,” Fed Chair Jerome Powell said at his press conference. “We need to see how things actually play out.” While the Fed’s official projection still shows two cuts next year, Powell emphasized that policymakers are in no rush to move until they have a clearer picture of how the economy responds. He noted the economy remains stable enough to allow for patience. “We are well-positioned to wait for greater clarity,” he said. Markets welcomed the Fed’s cautious approach, with the Dow Jones Industrial Average climbing 0.92% and the S&P 500 up 1.08%. Still, signs of division within the Fed are emerging. Eight officials now expect only one or zero cuts this year — double the number from December — hinting at a more hawkish tone. Matthew Luzzetti, chief U.S. economist at Deutsche Bank, said markets had been bracing for exactly that shift due to persistent inflation concerns. But others pointed out Powell’s calm demeanor. “He seemed very relaxed about inflation,” said Padhraic Garvey, ING’s head of research for the Americas. At the same time, the Fed lowered its 2025 GDP growth forecast from 2.1% to 1.7% and predicted inflation would rise to 2.7% by year-end — well above its 2% target. Powell attributed part of the rise to tariffs but suggested it might be temporary, with inflation expected to ease to 2.2% in 2026. The Fed’s mixed messaging left some economists puzzled. “Uncertainty is high, inflation is climbing, officials are split on rate moves, but they still forecast two cuts. That seems wildly unlikely,” economist Robert Frick commented on social media. Luzzetti noted that any cuts this year would likely be reactionary — what he called “bad-news cuts” — in response to slowing growth. In a separate move, the Fed announced it will slow its balance sheet runoff, allowing just $5 billion in Treasurys to mature without replacement each month, down from $25 billion, while continuing to let $35 billion in mortgage-backed securities roll off. In a rare dissent, Fed Governor Christopher Waller voted against the decision, preferring no change to the runoff pace. The economy has clearly cooled since the start of the year, with uncertainty around trade and immigration policies adding pressure. Meanwhile, inflation remains stubbornly above target, complicating the Fed’s ability to respond proactively. With the White House set to announce new tariffs on April 2, economists warn the Fed could soon face even more challenging conditions. “I think by the time Powell holds his next press conference in May, the data will look quite a bit uglier,” said Scott Anderson, chief U.S. economist at BMO Capital Markets.

Nvidia’s Keynote Delivered Energy — but Not the Surprise Investors Wanted Jensen Huang took the stage with trademark enthusiasm, speaking for nearly two hours — reportedly without a teleprompter — about Nvidia’s bold future. But for investors hoping for a game-changing announcement, the presentation fell short. Nvidia shares (NVDA) closed down 3.4% on the day, deepening losses that had already begun before Huang’s keynote. While Nvidia’s roadmap is undeniably impressive — highlighting advances in AI, robotics, autonomous driving, and quantum computing — much of what Huang presented felt familiar. “The things Huang said had already been said,” observed Maribel Lopez, founder of Lopez Research. Investors had been hoping for an unexpected catalyst: a surprise revenue stream or a groundbreaking product reveal. Instead, Huang pointed to areas that are still in development and years from delivering meaningful financial returns. Quantum computing remains a long-term prospect, and while robotics is growing, it won’t scale as quickly as Nvidia’s core chip business, according to Lopez. Jefferies analyst Blayne Curtis also expressed mild disappointment, saying he had hoped for “more proof points” demonstrating how Nvidia could expand its total addressable market and drive greater cost advantages. Still, he acknowledged Nvidia’s unmatched strength across hardware, software, and vertical markets. Huang did reveal Nvidia’s product cadence moving forward: a new platform every 12 to 18 months. Next up is Vera Rubin, arriving in late 2026 with 3.3x the performance of today’s Grace Blackwell system and 144 GPUs. Vera Rubin Ultra will follow in late 2027, delivering a massive 14.4x performance increase with 576 GPUs. The company also teased its next-generation platform, “Feynman,” expected in 2028. Curtis noted that Vera Rubin appears to be an incremental upgrade, with Rubin Ultra marking the more significant leap. Even so, Nvidia’s ability to innovate rapidly continues to give it an edge. “Even when they stumble, they recover fast,” Lopez said, pointing to the company’s unwavering ability to stick to its ambitious product cycle. Huang ended by reminding the audience that Nvidia’s technology isn’t an impulse buy. “This isn’t like buying a laptop,” he said. Enterprises need time to plan budgets, infrastructure, and power demands — part of a much larger vision that remains firmly in place, even if short-term surprises are in short supply.