Successful investor Jim Mellon is always a voice worth listening to. Recently, he appeared on The Master Investor Podcast, where he made some interesting points about the stock market. Let’s take a look at a couple of them.
AI bubble
The first thing worth mentioning is that Mellon isn’t buying hype around artificial intelligence (AI) stocks. He thinks the AI boom is essentially a bubble that’s destined to pop, saying that the “great bust…will inevitably come in AI in the relatively near future. We don’t know when, but sure enough, there will be a bust“.
He’s also bearish on Magnificent 7 stocks, pointing out that basically every single financial institution and many retail investors already hold them. They make up a big chunk of the S&P 500. Who, he asks, are “the marginal extra buyers” when everyone already owns the stocks?
To my mind though, this was true six months ago. Yet shares of Nvidia and Microsoft are up 47% and 30%, respectively, while Alphabet popped 9% yesterday (3 September) to hit a record high. Clearly, there are still enough buyers around to keep cash flowing into these names.
However, I feel he makes a good point when he says that most cloud giants are essentially doing the same thing. They’re all building AI data centres, packed mainly with Nvidia chips, to pump out similar AI models. Mellon likens this to railroads in the 1850s, where most shareholders in rail companies didn’t do very well.
I do think there’s a risk of ‘commoditisation’ for AI start-ups, meaning they’re all producing very similar products. And that’s why I think the latest valuations of OpenAI and Anthropic — $500bn and $183bn, respectively — look crazy. This part of the AI market is a bubble waiting to pop, in my opinion.
However, I don’t think the likes of Amazon (NASDAQ:AMZN) and Alphabet are at insane levels. They already have very large profits to back up their valuations.
Robotics revolution
In the podcast episode, Mellon said he’s uber-bullish on humanoid robotics: “We will have more robots on the planet by 2050 than there are human beings, many more, and they will be doing everything.”
At first glance, this world in 25 years would appear to suit Nvidia. Humanoids need huge computing power for vision, movement, and decision-making. Billions of robots would mean surging demand for Nvidia’s AI chips/robotics platforms, unless Chinese competition intensifies.
However, I also think Amazon stands to gain massively from this revolution. With over 1m robots deployed, Amazon’s robotic workforce is nearly matching its human staff of roughly 1.5m. Millions more advanced bots would mean faster picking, packing and shipping, with lower labour costs.
Meanwhile, autonomous delivery vans and last-mile robots – both of which Amazon is heavily investing in – could cut costs further. The end result may be noticeably higher profit margins.
Because, as Mellon says, robots “are able to work 24 hours a day, don’t pay National Insurance, not yet anyway, although they may do in the future, don’t complain and are non-unionised.”
Of course, there’s more to Amazon than just robots. It’s facing near-term uncertainty with tariffs, which could lead to higher prices and a slowdown in its core e-commerce operation.
But trading on a reasonable forward price-to-earnings ratio of 32, I think the stock is worth considering for long-term investors.
