Nvidia reported $81.6 billion in revenue for the first quarter of 2026, beating analyst expectations by about $2.4 billion. Earnings per share came in at $1.87, also above consensus. The company guided for $89 to $93 billion next quarter. By every conventional measure, this was a strong earnings report. The stock fell 5%.
This pattern, where a company delivers better results than expected and the stock still drops, is one of the clearest signals that a price already reflects years of optimism. It came up today across two of our portfolios. Performance holds the Nasdaq-100 (an index fund heavily weighted toward technology companies like Nvidia), and Classical holds the S&P 500, where Nvidia is one of the largest components. Both felt the drag.
The framework for understanding this comes from Bravos Research, a macro analysis firm that has been tracking the semiconductor sector closely. Their observation, published in May 2026, is that the semiconductor index has risen 150% in two years and outperformed the broader stock market by 120%. This meets what the National Bureau of Economic Research, the organization that officially dates recessions in the US, considers the formal definition of a bubble: a sector that approximately doubles relative to the broad market in under two years.
What makes a bubble dangerous is not that the underlying business is bad. Nvidia's business is genuinely strong. The danger is that the stock price has already absorbed years of future growth into today's valuation. When that happens, even excellent earnings produce a shrug, because the market was already pricing in excellent earnings. The bar shifts from "will they grow?" to "will they grow faster than the fastest growth already priced in?" That is an almost impossible bar to clear quarter after quarter.
The practical lesson here is about how price and fundamentals can diverge. A great company can be a terrible investment if you buy it after the price has already run far ahead of the business. Bravos has been rotating their exposure out of semiconductors and into energy producers, energy infrastructure, and base metals like copper and lithium. The logic is straightforward: the same AI buildout that is making Nvidia's chips sell so well is also consuming enormous amounts of electricity, cooling water, and raw materials. Those inputs have not yet seen the same price appreciation as the chips themselves.
For our portfolios, this does not trigger any trade today. Performance holds the Nasdaq-100, where Nvidia is one piece of a broader basket. A single earnings report, even a disappointing stock reaction, does not change the mandate. But the pattern is worth watching. If the semiconductor sector starts declining while the broader market holds up, that would be a rotation signal consistent with the late-stage bubble framework that Bravos describes.
What would change this view: if semiconductor valuations correct meaningfully (30% or more from peak) without a corresponding earnings collapse, that would reset the bubble clock and make the sector attractive again. The thesis is about price relative to fundamentals, not about the fundamentals themselves. Nvidia could keep growing revenue at 50% a year and still be overpriced if the stock already expects 80%.