To The Moon

On November 19, 2025, Nvidia $NVIDIA(NVDA)$ delivered yet another blockbuster quarterly report that not only shattered Wall Street expectations but also sent its stock soaring in after-hours trading. Revenue clocked in at a record $57.01 billion — up 62% year-over-year and 22% sequentially — while adjusted earnings per share hit $0.81 (or $1.30 on a reported basis in some metrics), comfortably beating consensus forecasts. More importantly, the company guided for Q4 revenue around $65 billion, well above the $62 billion analysts had anticipated. Blackwell GPU sales were described as "off the charts," cloud providers' capacity is fully utilised and sold out for quarters ahead. CFO Colette Kress reaffirmed visibility into roughly half a trillion dollars in AI infrastructure orders for calendar 2025-2026 alone.

The immediate market reaction was unambiguous: Nvidia shares jumped 3-5% after hours, dragging the broader semiconductor complex higher (AMD +4%, TSMC +4% in Asia, Micron +3%, etc.) and lifting mega-cap tech names like Microsoft, Amazon, and Meta. One of the week's two major uncertainty "boots" — Nvidia's earnings — has landed firmly on the bullish side. The question now: Does this erase the fear of an AI bubble? Is Jensen Huang correct in stating that "AI has no bubble"? Are we standing at the foothills of a new semiconductor supercycle?

Short Answer: Yes, for Now — and Probably for Longer Than Sceptics Admit

Jensen Huang opened the earnings call by directly confronting the chatter: "There's been a lot of talk about an AI bubble. From our vantage point, we see something very different." He framed the current moment not as speculative excess but as the convergence of three simultaneous platform shifts:

1. From general-purpose to accelerated computing

2. From traditional AI to generative AI

3. From generative to agentic/physical AI (reasoning agents, robotics, autonomous systems)

Nvidia, Huang argued, sits at the centre of all three because its full-stack architecture (CUDA software + hardware) is the only one that scales seamlessly across them. This isn't marketing fluff — it's backed by the numbers. Demand isn't softening; it's accelerating and compounding. Inference workloads are now growing exponentially alongside training, installed base utilisation is at 100%, and every major hyperscaler, plus a growing list of sovereign AI nations and enterprises, are scrambling for capacity.

The bubble thesis has rested on three pillars, all of which took hits this week:

In short, this wasn't a "good but expected" quarter — growth re-accelerated sequentially for the first time in several quarters, and the guide implies continued acceleration into fiscal 2027.

Can the Semiconductor Sector Enter a New Bull Run?

The short answer is yes, but the character of this rally will differ from the explosion seen in 2023-2024.

• Valuation Compression: Strong earnings help "compress" the P/E (Price-to-Earnings) ratio. Even if the stock price goes up, if earnings go up faster, the stock actually becomes "cheaper" fundamentally. This invites institutional money that was previously sitting on the sidelines due to valuation fears.

• From Training to Inference: The 2025 narrative is shifting. While the initial boom was about training models (which requires massive raw power), the next phase is inference (running the AI for users). NVIDIA’s forward guidance likely hints that inference demand is exploding. This suggests a sustainable, long-term revenue tail, rather than a one-off spike.

• Broadening the Rally: A "rising tide lifts all boats." NVIDIA’s success signals health for the entire supply chain—from memory manufacturers (like Micron/SK Hynix) to foundries (TSMC) and equipment makers (ASML). We should expect a sector-wide rotation where laggards in the semi space begin to catch up to NVIDIA.

The Macro Factor: The "Second Boot"—Thursday’s employment data.

This context is critical.

Paradoxically, the market is currently in a "Goldilocks" zone for tech.

• Scenario A: If the employment data is weak (signalling a cooling economy), investors will flock to "Growth" stocks like NVIDIA because they are the only companies capable of growing earnings in a recession.

• Scenario B: If the data is strong (soft landing), the broader market rallies, and risk appetite increases.

The only danger to the semiconductor rally is Stagflation (high inflation + low growth), which seems unlikely in the current print. Therefore, NVIDIA’s earnings provide a safety net. Even if the macro data is shaky, the sheer force of AI spending provides a floor for the tech sector.

Conclusion: A "Show Me" Market

NVIDIA’s report has successfully kicked the can down the road. They have proven that the build-out phase of the AI revolution is far from over.

• Does it disprove the bubble? It disproves an imminent burst. It proves the hardware demand is organic and robust.

• Will chips rise again? Yes. The "Fear of Missing Out" (FOMO) has now transitioned into "Fear of Underweighting." Fund managers who cut exposure to chips, expecting a pullback, are now forced to chase the rally.

Strategic Verdict: The green light is back on for the semiconductor sector. However, smart capital should remain vigilant. Watch the margins of the software companies in the next two quarters. NVIDIA is the engine, but the software companies are the fuel. As long as the engine is running this hot, the direction is up.

@TigerWire