Comment by JumpCrisscross | Hacker News Reader

Comment by JumpCrisscross | Hacker News Reader

Parent Full thread

For a growing company, the PEG ratio [1] is more appropriate than P/E. What it asks is how many years of sustained growth brings the P/E ratio in line with population. (As is customary in finance, it's inverted from its intuitive form.)

OpenAI reportedly "doubled its annualized revenue to $3.4 billion in the past six months" to June from "$1.6 billion in late 2023" [2]. That implies a CAGR of 350% [a].

$100bn is a 29x revenue multiple to that ARR. S&P 500's average is 2.9x [3]. For OpenAI to match that multiple its sales would need to grow at 350% for 3 years. That would put OpenAI near the top 100 companies in America by sales [4]. Huge. But not unprecedented.

[1] https://www.investopedia.com/terms/p/pegratio.asp

[2] https://www.theinformation.com/articles/openais-annualized-r...

[a] (3.4 / 1.6) ^ (1 / 0.5) - 1

[3] https://www.multpl.com/s-p-500-price-to-sales

[4] https://en.wikipedia.org/wiki/List_of_largest_companies_in_t...

Information on openai changes quickly and is very unreliable, but:

https://www.maginative.com/article/openais-annualized-revenu...

This site says 3.2B of that comes from subscriptions. Omitting usage via API and assuming 1 user = 240$ annually that makes 133.3 million subscriptions.

https://backlinko.com/chatgpt-stats (says Jun. 04, 2024 but this might be outdated a bit more)

Here it says openai has roughly 3.9 million subs in USA. On that basis I would assume roughly 8 million in Europe. Even if we add India and China(but they don't operate in China currently from what I gather) with numbers raised linearly along their population this doesn't add up. Maybe they rake in a lot via their enterprise plans or am I dumb and missing something ?

> Information on openai changes quickly and is very unreliable

Agree. My napkin analysis is dependent on the reliability of those data.

> Omitting usage via API

That's a big assumption. I'd assume larger purchasers have negotiated a minimum buy in exchange for discounts.

> sales would need to grow at 350% for 3 years

I assume you mean 350% per year over 3 years. That works out to sales growing (1+3.5)^3 = 91-fold over three years.

Sure, it could happen. What's more, if OpenAI achieves AGI, growth could be even faster. But what are the odds?

For investors counting on those crazy insane growth rates to justify a present valuation "above" $100B, the potential payoff is vertigo-inducing uncertain.

I did my math wrong. It’s grow 350% for about 19 months. 10x.

Thanks. That's still a pretty high hurdle, and it justifies only the current valuation. To justify a greater future valuation (say, 5x the current valuation, or "above" $0.5T), even faster growth would be required. That's a bold gamble.

Reply on news.ycombinator.com