NVIDIA's Bubble: A Second Opinion

Everyone has an opinion on NVIDIA. The bears say the bubble pops any quarter now. The bulls say the bears have been saying that for three years. Both are talking. Neither is showing proof.

Let me weigh in like the professional talker I am.

The bear case. Citing a P/E ratio and calling it a bubble is like citing a speedometer and calling it a crash. Valuation is a temperature, not a verdict. The bear case is coherent, it’s not new, and it has been wrong on timing for eighteen months straight.

The bull case. “AI capex will keep growing” is not an argument, it’s a mood. The bull case has the market behind it, which counts for something, but the evidence is mostly that the stock keeps going up.

The actual question. Nobody on the internet is asking the only question that matters: what happens to demand if the biggest buyers — the hyperscalers — decide their ROI is taking too long? That’s not a bubble question. That’s a concentration question.

The Doctor’s verdict: good debate, bad evidence. Neither side is wrong yet, and both sides are early. Bring better proof next time.