The question is whether NVDA can grow into its valuation long enough to justify the price. The market cap is now over $3.16tn. For that sum would you prefer to buy NVDA or the entire economy of Italy or France and still have $158bn left over?

To justify its price NVDA would need to ride the tech capex into AI boom (or bubble) for how many years to reach $120bn in earnings (implying a p/e of 25)? Or for a terminal p/e of 30, $$105.3bn? So the question about whether NVDA can grow into its valuation is a fundamentally a question about when the tech capex bubble will burst, as all bubbles eventually do, and how much that bursting bubble will shrink NVDA's earnings as the capex into AI growth flattens out.

If you think $125 is fair value and want to buy NVDA at a 20% discount, sell a cash-supported 08/2025 expiration $100 strike put on it for $12.30 and get paid $1230 to wait to get your price at a net cost of $87.70 if it crashes, or keep the option premium if it doesn't. Heads you get your price, tails you keep the money by selling expensive hedging to NVDA's shareholders. 54.4% implied volatility on the put tells you you're being well paid to sell overpriced insurance to calm their nerves. When fear is expensive, sell it. When cheap, like the VIX at 12, buy it.

Go out to 12/2026 to sell the same thing for $2099. But you have to want to own NVDA in two years with far higher probability that Mr. Market will at some point panic about the tech capex into AI growth flatlining. For the same option premium money as the first example, you can drop the strike price down to $79 if you go out to 12/2026 expiration. For that one you need only $7900 in spare cash to support shorting that put. If exercised in a bursting bubble, you'll own NVDA at little more than half today's price. It's like reserving a hotel room on Priceline with a rock bottom bid, but you don't care when in the next two years you get the room.