Simmering or bubbling?

When preparing to write the Weekly Digest, I’m usually thinking about the potential subject matter during the previous week. Last week, I was happily gathering data to discuss whether or not equity markets are in some sort of bubble and everything was going well… until President Trump announced the threat of incremental 100% tariffs on China. Cue a rapid sell-off and an unscheduled diversion to a different track. The result is a two-for-the-price-one-deal, because now I am going to discuss both of them!

Trump turns up the gas

Let’s look at the tariff threat first. Trump justified the move by saying it was a response to actions by China, and it certainly looks as though both sides can take some of the blame. China tightened controls on the exports of rare earth minerals, which are key raw materials in various high-tech products, not least industrial magnets. A supply shortage brought the US automobile industry to the brink of shutting down earlier this year. There are other restrictions too, but rare earths are the key ones, and it is this card that looks like the strongest in President Xi’s hand at the moment, owing to China’s dominance of the rare earth refining industry. Trump responded with the much blunter weapon of tariffs, still ignoring (wilfully or not) the fact that it is American companies and consumers who will have to pay most of the tariff costs.

What’s going on here? Presidents Trump and Xi are scheduled to meet at the Asia-Pacific Economic Cooperation (APEC) summit at the end of this month. This is just days before the current deal to pause the existing tariffs on Chinese exports to the US expires. It seems plausible that Xi wants to exert a little leverage over whatever negotiations take place, but Trump cannot be seen to take that lying down and so reacted as he usually does. Having also threatened not to meet Xi in South Korea, he has since made more concessionary noises, even allowing that Xi was just having a “bad moment”! The phrase used in diplomatic circles is “escalate to de-escalate”, and that is how investors are reading the current situation. The US needs Chinese rare earths and China needs certain US semiconductors and, for example, aerospace components. They are mutually dependent but will never admit it, although both are working towards a world where they might not be. For now, that suggests the odds favour a renewal of their current agreement on trade, but one that is somehow engineered to allow both sides to look like the winner to their respective populations.

The bubble in bubble warnings

There is no shortage of august bodies and individuals saying that the AI “bubble” is ready to burst (and more than a few full-time Cassandras, too). Last week both the World Bank and the Bank of England drew attention to the risks, as did the CEO of the world’s biggest bank, Jamie Dimon of JPMorgan, and the legendary hedge fund manager Paul Tudor Jones. This is high-profile stuff and probably set nerves jangling among mere mortal market participants.

There are at least two problems to solve here. The first is defining whether this is a bubble; the second is identifying what might be the catalyst to burst it. The necessary conditions for a bubble are definitely present. The authors of Boom and Bust, William Quinn and John Turner, both finance professors at Queen’s University Belfast, came up with the concept of the “bubble triangle”. As a fire needs fuel, oxygen and a spark to ignite, a bubble needs speculation, marketability and credit/money. Speculation, even if often well-informed, is rife in the field of artificial intelligence (AI); marketability is ticked off through the presence of a strong narrative and easy access to investment products (often leveraged); and credit/money is being supplied by private equity and credit funds, other companies (look at chip-designer Nvidia’s string of investments in some of its own customers) and central banks in the form of interest rate cuts.

Case closed, then? Not so fast. First we have to think about the opportunities in AI. How big is the future market going to be? How long might it take for the industry to become commercial? And who are going to be the winners and losers? If your answers are respectively “huge”, “quickly” and “everyone”, then there is a long way to go. Of course, that would be the rosiest scenario, and one that does seem improbable. It seems more likely that at least one leg of that stool will wobble from time to time. Right now, I would say that the potential for widespread adoption is huge, but there is still some uncertainty about who is going to pay for it and how much. Accessing, say, ChatGPT for nothing is one thing; signing up for a paying subscription is another, although, to be fair, most households now pay for streaming services that not too long ago were “free to air”. Investors are still struggling to sort the sheep from the goats in terms of winners and losers, with the “picks and shovels” companies being the primary beneficiaries so far. The beauty parade is going to run for a while yet, with stock selection set to be more important than just backing the field.

Then there is valuation. Very simply, the current two-year forward price/earnings ratio for the Magnificent Seven leading US technology shares (all of which are exposed to AI) is around 27x. That compares to the top seven at the height of the TMT boom in 2000 being on a multiple of 52x. Japan’s leading business groups in 1989 were on 67x and the top 7 of the Nifty Fifty leading US shares in 1973 were on 34x. Furthermore, the current leaders have a much higher return on equity (46%) vs 2000 (28%) and look better on net income margins (29% vs 16%). They also have very strong balance sheets. The key point is that a new industry is being born out of existing cash flow (at least for now) and not solely from a combination of debt and new equity. Equity capital to fund growth is not being extracted from shareholders. In fact, US share buybacks are still expected to top $1 trillion in 2025 and there is more equity being retired from global equity markets than issued.

Given the technology-led nature of the current US bull market, it is not unexpected that parallels are being drawn with the period from 1996 to 2000. There are uncanny similarities that lend credence to the bears’ narrative. For the Russian debt crisis and collapse of Long-Term Capital Management in 1998, perhaps read Liberation Day. The Federal Reserve’s restart of its rate-cutting cycle does not compare with the aggressive response of Chairman Alan Greenspan in 1998, but it signals the beginning of a new trend lower and futures markets are anticipating a lot more cuts next year, especially once current Chairman Jerome Powell vacates his seat in May. But there was still a long way to go in 1999, and the house of cards was built to much higher levels even after the Fed started raising interest rates again in 1999.

No doubt there are pockets of speculation in equities, especially in the US. But the overall market looks less vulnerable at this time, last Friday’s sell-off notwithstanding. The S&P 500 index had just recorded 47 consecutive days without a one-day fall of 1% or more, and so a fall of 2.7% was somewhat overdue. The fall was exacerbated by an unwinding of speculative and leveraged positions, and that’s a risk that’s ever-present these days. A more prolonged sell-off requires a stronger catalyst, such as sharply tightening financial conditions or possibly the sort of exogenous event that, by its nature, is very difficult to forecast, especially in terms of market timing.

The number of payrolled employees in the UK has been generally falling over the past year

Economic Highlights