$10,000 in NVIDIA in 2014 is $5.4 million. The interesting number is the $523,236 you had to watch disappear.
The return is the least useful fact about this trade. What decided whether you collected it was a ten-month stretch in 2022 when your position fell by 63% — and you were still up thirty-fold while it happened.
- What happened
- $10,000 invested in NVIDIA at the first close of January 2014 was worth $5,410,701 by September 2026 — a 54,007% total return, or 64.4% a year compounded over 12.7 years.
- Why
- Almost nobody collected it. Between November 2021 and September 2022 the same position fell 62.9%, from $832,510 to $309,274. Ten months, $523,236 gone, with no way of knowing at the time that it would come back.
- The numbers that matter
- $5,410,701NVIDIA, from $10,000$42,792S&P 500, same window−62.9%Worst fall, Nov 2021–Sep 2022$523,236Paper value lost in that fall
- Why investors care
- The lesson that transfers is not 'buy NVIDIA'. It is that the return you read about and the return you receive are separated by whether you can sit still, and that a position can be up thirty-fold and still feel like a disaster.
- What happens nextas we saw it on 14 Sep 2026
- Run the same arithmetic on any company, amount and year with our calculator — including the ones where picking the company lost to simply buying the index.
A 62.9% fall. Both numbers are enormous multiples of the $10,000 that started it — which is exactly why it was so hard to hold.
| Nov 2021 | $832,510 |
|---|---|
| Sep 2022 | $309,274 |
This series asks the same question every time, and it is never the one the headline asks. Not what did it return, but could you have held it? Those are different questions, and only the second one decides what ends up in your account.
NVIDIA is the clearest case anyone has produced this decade.
The number everyone quotes
The number almost nobody quotes
Between November 2021 and September 2022, that position fell 62.9%.
Nothing about that period announced itself as temporary. The demand that had lifted the shares was visibly reversing, the sector was working through an inventory glut, and the recovery that now looks inevitable was, at the time, a claim about the future with no evidence attached to it.
The part that makes this hard, and it is not what you think
Against simply buying the market
This is the asymmetry the survivor story hides. The calculator that produced these figures will run the same arithmetic on Intel, on Ford, on Pfizer. The results there are not inspiring, and they are drawn from exactly the same list of large, well-known American companies a reader in 2014 would have been choosing from.
What actually transfers
- The return you read about and the return you receive are separated by one thing: whether you were still holding.
- A position can be up thirtyfold and still feel like a loss, because people measure from the peak rather than from the entry.
- A 63% fall is not a signal that you were wrong, and it is not a signal that you were right. It is what owning a single company costs.
- Choosing NVIDIA in 2014 — from a field that also contained Intel — is the entire difficulty, and no backward-looking calculator performs it.
You can run this on any company, amount and starting year yourself — the calculator is here, and it shows the drawdowns by default rather than on request.
Sources
- What if you had invested? — NVIDIA, $10,000, 2014Wealth Signal · Sep 14, 2026 · Primary source
- Form 8-K: ten-for-one forward stock split, filed May 22, 2024NVIDIA via SEC EDGAR · May 22, 2024 · Primary source
- NVIDIA Announces Financial Results for First Quarter Fiscal 2027NVIDIA · May 20, 2026 · Primary source
$10,000 a year ago: bitcoin, gold, and the part return charts never show
One of these turned $10,000 into $13,115. The other turned it into $5,310. The interesting question is not which — it's what you would have had to endure to still be holding.
What a $5.45 trillion price tag actually assumes about NVIDIA
The most common argument about NVIDIA is whether it is expensive. That is the wrong question. Here is the arithmetic of what the market is currently paying for — with every assumption disclosed.
Market Autopsy: how Silicon Valley Bank died in 44 hours
A bank with $200bn in assets failed in less than two days. The cause was not fraud, and it was not a bad loan book. It was a duration mismatch, a badly sequenced announcement, and a group chat.