For investors who have grown accustomed to Nvidia acting as the primary engine of market momentum, the morning of November 9, 2026, brought a jarring shift in sentiment. Shares of the semiconductor giant, which has long commanded the attention of Wall Street, plummeted in pre-market trading, falling 5.19% to reach $176.96. On that same day, the Nvidia board of directors initiated an investigation into whether CEO Jensen Huang had been targeted by foreign intelligence services. Compounding the gravity of the situation, Huang issued an admission to company employees, confirming that he had been unfaithful to his wife.
In the high-stakes world of modern finance, the sheer size of a company often dictates the ripple effects of its failures. Nvidia has become a colossus of market capitalization, and when such a massive entity stumbles, the immediate instinct of the market is to brace for a broader contagion. The financial impact was immediate and severe for the company itself, with $270 billion in market value wiped off Nvidia’s valuation in the wake of the news. For many observers, the question was not whether Nvidia would suffer, but how much of the rest of the market would be dragged down into the abyss along with it.
UK analyst Julien Garran just published the most alarming assessment of AI investment to date. His firm MacroStrategy Partnership, which advises 220 institutional clients, claims we're in the biggest bubble in history. The AI bubble is 17 times larger than the dot-com crash and four times bigger than the 2008 housing crisis. The math behind it: Garran uses the Wicksellian differential to calculate capital misallocation. When corporate debt costs run 2 percentage points above nominal GDP, capital allocates efficiently. For a decade, quantitative easing pushed rates below that threshold. The pre-2008 crash misallocation was around 18 percent of GDP. Garran estimates the current figure at 65 percent, including AI, real estate, venture capital, and yes, even NFTs. The fundamental problem: no killer app exists. Ten AI startups with zero profits gained nearly $1 trillion in market value over the past year. Everyone except NVIDIA is bleeding money. His most controversial claim: LLMs can only handle jobs where nobody checks if you got it right. At one software company, AI task completion ranged from 1.5 to 34 percent, and even the higher rate wasn't consistent. The cost curve is breaking. ChatGPT-3 cost $50 million to train. GPT-4 cost $500 million. GPT-5 reportedly cost $5 billion and didn't deliver substantial improvements over its predecessor. When does it pop? Garran admits he can't call the top. Markets hit all-time highs last week. But VC funding for AI startups is drying up because valuations are absurdly high. That leaves SoftBank, foreign governments, and NVIDIA as the only players with capital to keep the party going. Two scenarios if he's wrong: The bubble inflates longer, wasting more capital on economically unproductive projects and dragging down future GDP. Or someone achieves superintelligence and reshapes society completely. He's betting on the first scenario. Whether you agree or not, the tension is real: massive investment, limited profitability, and breakthroughs that haven't materialized.
Yet, as the trading session unfolded, a peculiar phenomenon emerged: the broader indices did not collapse. While the Nasdaq 100 saw a minor decline, US stock futures for the Dow Jones and the S&P 500 actually rose. This divergence highlights a critical, often misunderstood aspect of how modern stock indices are constructed. The market is not a monolith; it is a collection of different mathematical frameworks, each designed to weigh the influence of individual companies in distinct ways. The Dow Jones Industrial Average, for instance, is price-weighted, meaning that a company’s influence is tied to its share price rather than its total market value. In this index, Nvidia holds only a 2% weight, effectively insulating the Dow from the full force of the company’s volatility.
The S&P 500 and the Nasdaq 100 operate on a different logic, utilizing market-cap weighting to determine the influence of their constituents.
The S&P 500 and the Nasdaq 100 operate on a different logic, utilizing market-cap weighting to determine the influence of their constituents. In the S&P 500, Nvidia accounts for 6.3% of the index, while its presence in the Nasdaq 100 is significantly more pronounced, at 11%. At that 11% threshold, a slide in Nvidia’s share price would typically be expected to drag on the benchmark by over half a percentage point. However, the fact that these indices remained relatively stable suggests that investors are performing a sophisticated triage, isolating the Nvidia crisis as a company-specific issue rather than a systemic threat to the semiconductor sector or the broader economy.
13 years ago, I joined NVIDIA Research working on mobile computing (aka Tegra). The NVIDIA might be seen as the most valuable company now, but it was in survival mode back then. We were trying to create a new SoC market to compete with MediaTek and Qualcomm who enjoyed more than half of the market, when the rest was dominant by Apple A-chip. As gaming moving from the desktop to mobile, the desktop gaming GPU market was shrinking. Expanding Tegra business at that time was the only way to survive. Unfortunately, the biggest mobile computing market was the phone. To enter the phone market you need a modem. Getting the modem to certify by Verizon or AT&T was insanely hard and taking a long time. With Qualcomm bundling their modem with extremely competitive pricing, it was very hard to compete. Tegra did not go anywhere. Then, we were trying to explore other markets where we could sell Tegra at higher margin without a modem. Automotive seems to be a good fit. As car OEMs began to showcase their concepts at CES, “Smart Cars” sounded like a buzz word. In reality, most car OEMs were quite conservative. Most of them did not want to put any graphics chips into their production cars. For one, they did not have a big display. European designers did not like touch screen. Metal and leather, they said. Ironically, this was the time when Nokia sold to Microsoft after being defeated by Apple because it was dragging its feet to make a good touch display phone. Fortunately, not far from Santa Clara, there was an up and coming car company eager to try something new. The moment they decided to put two Tegra chips in a car, Jensen breathed sign of relief. The showcase of CES 2011 was a success, and that helped Tegra chip expanded to other OEMs. Jensen even invited Elon to give a talk at our “small” auditorium (well, we were less than 6000 people back then). I was there without knowing I might work for him one day. Eventually our mobile computation team developed a set of computer vision toolkits on Tegra for various kinds of OEMs. Whoever we could sell chip to, we made demo for them. Then AI began to change everything. The classic computer vision like Sobel Filter or global optimization no longer sounds sexy or useful. Deep convolutions became the king. 9 years later, I picked up the problem again that I first met. What a strange fate.
Perhaps the most telling indicator of this investor sentiment is the performance of Nvidia’s peers. While Nvidia shares were in freefall, other major semiconductor players, including AMD and Broadcom, were trading higher. This decoupling suggests that the market is viewing the current crisis through a narrow lens. Investors are not abandoning the semiconductor industry; they are simply recalibrating their exposure to Nvidia specifically. By separating the company’s internal leadership crisis from the underlying demand for its technology, the market is demonstrating a level of discipline that prevents a singular corporate scandal from evolving into a sector-wide panic.
As the situation continues to unfold, the events of November 9 serve as a case study in market mechanics. The resilience of the Dow and the S&P 500 in the face of such a significant decline in a major tech leader underscores the importance of index diversification. While the personal and professional challenges facing Jensen Huang are undoubtedly significant for the future of Nvidia, the broader financial ecosystem has proven itself remarkably adept at absorbing the shock. For now, the market remains focused on the fundamentals, treating the Nvidia investigation as a contained event rather than a harbinger of a wider downturn.
Singapore, Monday. Jensen Huang admitted an affair and NVIDIA's board opened a probe; Nasdaq futures -0.6%. Friday: SOX +4.2%, but the Nasdaq 100 fell 4.9% on the week. Our read: AI demand is real, yet the rebound still runs on short covering. $NVDA https://t.co/UULxdhuYp3