A thousand dollars placed in Nvidia's initial public offering on January 22, 1999 would be worth $5.70 million today, a total return that folds in twenty-seven years of dividends and a split-adjusted cost basis of two and a half cents per share. The arithmetic is straightforward: forty thousand split-adjusted shares at the IPO price, compounded through five separate drawdowns exceeding fifty percent and two crashes that erased more than eighty-five percent of the stock's value.
The arithmetic of survival
The source frames the outcome as a lesson in conviction. Jensen Huang's vision, graphics cards, then GPUs, then AI accelerators, only paid off for shareholders who refused to sell when the dotcom bust and the 2008 financial crisis each cut the position by more than eight-tenths. The dividend component adds roughly $380,000 to the capital gain, a footnote that underscores how little income mattered relative to the multiple expansion.
The premium to the undisturbed price
What the source does not supply is the undisturbed price on the day before the IPO, so no premium calculation is possible. It also omits the break fee, the earnout structure, and the form of consideration, because there was none. This was a public listing, not a negotiated transaction. The only terms that mattered were the float price and the lockup expiry, neither of which the source discusses.
The next Nvidia is not this Nvidia
The source concedes that comparable returns are unlikely for buyers at today's levels. The 570,000 percent cumulative gain is a historical artifact, not a forward yield. Investors hunting for the next version of this trade are effectively looking for a company that can survive multiple near-death experiences while its addressable market redefines itself three times. The structure of the return, deep drawdowns, long flat periods, then exponential rerating, is the only replicable feature, and it is the one most investors lack the capital structure to endure.