A little under two years ago I proposed a possible "backwards" way to figure out whether Nvidia was a reasonable investment at a given price.

The assumptions were that you wanted a nominal 9%/year annualized return over the next 5-10 years, that earnings growth rates would gradually slow, and that terminal multiples would be in the teens. With those assumptions you could calculate the EPS growth rates required. You could then see whether they were plausible, which would tell you whether the current price was reasonable.

At the time the price was $130.78, and the ten years of EPS growth rates required were:
29%
27%
25%
23%
22%
20%
19%
18%
17%
15%
This list assumed that each year's EPS growth rate was 93% of the previous year's growth rate: very modest rate of slowing.
These growth figures seemed a little optimistic to me, so it wasn't demonstrably clear that the entry price was a good one.

It has been a little under two years since then. The stock price is up 19%/year since then, and earnings have continued to knock it out of the park.

What does the same approach tell us now?
Excluding one time items, last year's EPS came in at $4.77, only 12% above what the old table suggested. Analyst consensus for 2026 is $8.29, which isn't a completely wild guess because of their large known order book. I'm going to pencil in $7.50 as a starting figure for a bit more conservatism.

To get 9%/year annualized 5-10 years out, you need the following EPS growth rates needed to make 9%/year nominal at the average ending date 5-10 years out, with each year's growth only 80% of the prior year's:
25%
20%
16%
13%
10%
8%
7%
5%
4%

These seem to be very reasonable nominal EPS growth rates. At the very least, it's far more modest than what they have managed lately--first quarter EPS were up 97% over the same period last year.

Using that table the average nominal EPS in years 5-10 (2031-2026) would then be $18, which on a multiple of 19 would give you an average market price in that window of $341, which would get you 9.0%/year annualized (as 7.5 years) from today's $178.68.

So, despite being valued in the trillions, it might actually be a defensible investment based on valuation at current prices. Any further drops would add a margin of safety.

Obviously this is entirely speculative, but it shows you what kind of things would have to happen for today's price to be a good one. Seems like a useful trick.

Jim