
“A financial bubble occurs when optimism about future growth drives asset prices to levels that become increasingly difficult to justify by reasonable expectations of future economic returns”.
Almost every article about tech in the mainstream media these days starts with the premise of an “AI bubble” — or at least contains a reference to one — with surprisingly little concern about actually proving or disproving the statement.
Well, if AI is a bubble, Nvidia should be exhibit number one. It is the epitome of the AI boom.
And Nvidia just reported an impeccable quarter.
In Q2 FY27, revenue reached $96.2B (over $1.0B per day), up 106% YoY — more revenue than all Bovespa-listed companies combined. Growth accelerated sharply despite a revenue base that is now roughly three times larger than it was two years ago.
For Q3, Nvidia guided to $108B in revenue, up 89% YoY, which would make it the company’s first $100B+ quarter.
But the most remarkable number was not Q2 or Q3.
For the first time, Nvidia provided preliminary guidance a full year ahead: it expects FY28 revenue to grow approximately 70% — versus Wall Street expectations of 44%.
And there is another important detail: Nvidia says it is supply-constrained. According to the CFO, customers’ forecasts would point to growth closer to 100% if Nvidia had enough supply.
Net income grew even faster, up 126% to $59.7B, for an extraordinary 62% net margin. As a result, Nvidia’s forward P/E is now 23.93x (8/26/26). For perspective, after becoming the epicenter of the AI revolution, Nvidia is trading at less than half the P/E at which it ended nine of the past ten years — above 47x earnings.
In addition, nobody says that the Ibovespa is in bubble territory, but Totvs — Brazil’s largest listed software company — currently trades at a forward P/E of around 18–19x, not far from Nvidia’s 24x. Yet Totvs faces a much higher cost of capital — remember that valuation is ultimately NPV — while growing in reais at roughly one-third of Nvidia’s rate.
So, where exactly is the bubble?
Conclusion
Mainstream journalists are not immersed in Silicon Valley. More importantly, many see AI as a threat to their own livelihoods, creating an obvious potential conflict of interest. They are hardly disinterested observers.
The “AI bubble” narrative tells us as much about those writing about it as it does about AI itself.
But anxiety and feeling cornered in their careers by AI do not define a financial bubble.
Economics does. And the economics are telling a very different sto