How do these AI companies turn profitable on a short enough timescale for that to make sense? Suppose a step change AI model comes out tomorrow with good enough reasoning to basically replace an employee. Businesses need to retool workflows and processes to accommodate, even if it's better. This is years away, not months. There's not enough market for the compute. And that all assumes science fiction level results, which there is absolutely no indication they will achieve.
I don't think the big companies will fail, but their stock prices could dramatically drop. I think newer smaller companies like OpenAI and Anthropic could easily fail, as well as a bunch of other AI start ups. Altogether it could make for some difficult financial times like in 2000 and 2008
The NVIDIA case is such a strange example to use to make the argument the author is trying to make.
NVIDIA raised $25 billion and had $85 billion in orders. Because of the demand, it was able to upsize its offering and issue bonds in maturities ranging from 2 to 30 years at quite favorable interest rates. The amount raised is a quarter of a year's free cash flow and the spread tightened during the book building process, so bond investors obviously aren't on the same page as the author.
You really can't make a bearish argument about the amounts being raised without putting the numbers in perspective. Yes, the issuances are big, but the equity and cashflows are also big, so the amounts being raised in the bond market don't really align to the author's skepticism when it comes to NVIDIA, Google, Meta.
The author would have a stronger case with Oracle but that alone wouldn't support the "Big Tech" story line.
Edit: $25 billion is a quarter's worth of free cash flow for NVIDIA, not half a year's as I originally stated.
13 comments
[ 3.9 ms ] story [ 37.6 ms ] threadI also notice the article is "tagged" as if its title had been split on lowercase 's'es:
> Big tech i
> borrowing like never before and the fed ju
> T made that a lot more expen
This level of attention to detail does not exactly inspire confidence.
https://pivotal.substack.com/p/minsky-moments-in-venture-cap...
It's still not super huge compared to the amount of debt in other industries, but I guess the thought is it's riskier?
NVIDIA raised $25 billion and had $85 billion in orders. Because of the demand, it was able to upsize its offering and issue bonds in maturities ranging from 2 to 30 years at quite favorable interest rates. The amount raised is a quarter of a year's free cash flow and the spread tightened during the book building process, so bond investors obviously aren't on the same page as the author.
You really can't make a bearish argument about the amounts being raised without putting the numbers in perspective. Yes, the issuances are big, but the equity and cashflows are also big, so the amounts being raised in the bond market don't really align to the author's skepticism when it comes to NVIDIA, Google, Meta.
The author would have a stronger case with Oracle but that alone wouldn't support the "Big Tech" story line.
Edit: $25 billion is a quarter's worth of free cash flow for NVIDIA, not half a year's as I originally stated.
https://www.reuters.com/business/media-telecom/spacex-banker...