Disclaimer
This is not substantive investment research or a research recommendation, as it does not constitute substantive research or analysis. This material should be considered as general market commentary.
‘Tang ping’ is a Chinese phrase that some readers may have heard of. The literal translation is ‘lying flat’ and it refers to a rejection of the punishingly long hours that workers in the world’s second largest economy are often faced with.
A key part of the tang ping concept is the idea that, beyond a certain point, working more hours ceases to increase productivity or even has the opposite effect. This may sound like an impossibility to the average British slacker, but given it’s common for Chinese workers to be asked to work the so-called ‘996’ model - 9 am to 9 pm, 6 days per week - it may not be so far fetched.
The concept came to mind this week when reading a report in the Wall Street Journal on AI. Tucked away at the end of the piece was a claim that firms have spent $50bn on Nvidia chips in 2023 but only generated $3bn in revenues. Just as you can’t expect people to work more and more hours and produce increasing levels of productivity, it seems dumping huge sums of money into technology investments does not lead to automatic revenue growth and profitability.
You could argue that the topline figure is simply initial capex and that the sector will soon be booming. However, using IRS standards, the book value of an Nvidia chip would be zero after five years and that’s not accounting for the fact that chips may cease functioning long before then or be made redundant by new and better models. This would suggest further capex will be required, as opposed to this being a large up front investment that won’t have to be repeated.
The broader point is that there appears to be no simple relationship between capex and profitability in the nascent AI space. Instead you have a lot of companies funnelling money into projects that, at least so far, aren’t generating a commensurate level of sales. And for readers that argue we already have a winner in the form of OpenAI, it’s worth keeping in mind that Netscape was once the leading web browser - remember them?
Given that valuations in the US stock market seem to be so driven by hype around the sector, this raises an interesting dilemma. On the one hand you could see Nvidia as being almost like a venture capital proxy. Money flows from VCs into AI startups, who then proceed to dump a huge proportion of those funds into Nvidia chips.
At the same time, it is hard to see how such a scenario can continue ad finitum. As the figures above suggest, a lot of companies will churn out as they ultimately fail to produce a viable, profitable set of products or services. Nvidia would remain a beneficiary in the AI rush while it lasts but what happens after that?
Press continue to read the full article...
Kepler Trust Intelligence provides research and information for professional and private investors. In order to ensure that we provide you with the right kind of content, and to ensure that the content we provide is compliant, you need to tell us what type of investor you are.
Continue
