Back
Equity markets in July
After a bumper Q2, equity markets are (more or less) flat for the month as I type. But I wouldn’t say they have been calm. The Korean Kospi index, home of two of the largest semiconductor manufacturers in the world, fell almost 40% from its June peak, but is now up 24% from its Wednesday 29th lows (today is early on Friday 31st to date stamp all this).
Not financial advice, but my instincts are that people like shares that go up. Certainly, this looks to be true in Korea where two and three-times levered ETFs on Samsung and SK Hynix became some of the largest single-stock ETFs in the world (going from a rounding error to maybe $50bn of market cap by late June). These ETFs borrow money to buy shares. If you do this in enough size, you should push the share price up. Great news! But, of course, what happens if and when the shares start to go down? You then have to sell in a falling market to repay the debt you borrowed. Not such great news. And this is, I think, the story of Korea (and plenty of other AI related businesses) in the last few weeks:
And it is not just Korean retail investors who have had a wild ride. Leo Aschenbrenner is (was?) a wunderkind hedge fund manager who went from zero to $20bn in AUM in a couple of years by betting aggressively on AI winners (including semiconductors, energy providers and Anthropic itself). And he was also borrowing money to generate some of his extraordinary returns. But when the markets turned, he too had to sell to pay the margin loans back. In the end he couldn’t sell fast enough, and, I’d guess, his public market investments have been wiped out. CNBC was reporting that Citadel had taken over all the public equity positions (and associated margin debt) for (I would guess) a token $1 price. But this also meant a large, forced seller was done selling. Yesterday’s rally in the AI names began when this news came out.
I mention this because stocks can go up and down for fundamental or technical reasons. Much of the last few weeks has felt pretty technical to me (on the way up and the way down). But if this is the end of the forced selling from leveraged borrowers the market may go back to being a cleaner reflection of AI fundamentals. This would be a good thing.
But maybe you wish for a quieter world without all this AI driven speculation. If that is the case, I offer you the UK FTSE 100 index. This benefits from the tailwind of higher energy prices, has almost no direct technology exposure and is up a respectable 4% for July. Large parts of emerging markets come and go with the AI investment cycle. Whether this is a good thing or not, it is the UK that offers some diversification today.
AI fundamentals (and the impact on jobs)
Sam Altman, the boss of OpenAI, had a pretty interesting conversation with Patrick O’Shaughnessy this week on the Invest Like the Best podcast. Here he is talking about the cybersecurity breach where an OpenAI model (in development) broke into a different AI platform (called Hugging Face):
This is, of course, a glimpse of a Terminator-style future of man vs machine. But it also shows, I think, why AI adoption has actually been slower than a lot of the techno-optimists were predicting a couple of years ago. Letting an AI give you information is one thing. But letting it loose so it can actually do useful things (like book a flight, update a database, reply to client e-mails) is something completely different. The risk of reputational or financial damage from an AI making an error or going rogue is just too great today. This means humans need to be very much in the loop. Here is what Sam said on where AI intelligence is today:
We have, in aggregate, seen very little impact on the jobs market so far from AI. And I think this is part of the reason why. An employee who is a genius, but also a toddler, and who cannot be trusted to be left alone to get things done only has a certain, limited use. The companies investing trillions of dollars into data centres today are, I think, making a bet this picture will improve. We shall see. But, for now, we do seem to be in a sweet spot where the economy is benefitting from AI investment but AI is not smart enough to replace most real-world jobs. I am not smart enough to predict where this will end up, but I can tell you that the gyrations we have seen in markets in the last few weeks have very little to do with this (much more important) fundamental story.
Chris Brown, CIO
cbrown@ipscap.com
The value of investments may fall as well as rise and you may not get back all capital invested. Past Performance is not a guide to future performance and should not be relied upon. Nothing in this market commentary should be read as or constitutes investment advice.
