In a week highly anticipated by markets, nothing much happened. US rates came down a little, as expected, even though the economic reasons for doing so are thin (see Schroders quote). UK rates were unchanged, even though the economic reasons for doing so are clear. The US central bank made its decision in the dark shadow of Trump. In the U.K. there is lingering concern over inflation.
Whatever, more important is the direction of interest rates through 2026. At least in that regard the UK is better placed, with a number of cuts highly likely. That is why the FTSE 100 didn’t react to no cut, nor stubborn inflation numbers, nor the headline “UK stocks dumped at fastest pace in 20 years ahead of 'terrifying' Budget”.
Otherwise Japan, Brazil, and India had decent weeks, up 1.5%-3%, plus the tech-centric Nasdaq, with AI still exciting some investors.
Over in commodities it was also a quiet (ish) week, but for uranium (Global X ETF up 9%), and palladium down 4%.
Thanks again to those who joined our latest webinar last week, and if you missed it, a recording is still available here.
Moving on, I have said many times:
“There is nothing new when investing. One constant is the behaviour of investors, swinging from extreme fear to extreme greed and complacency – these are the bookends of cycles approximating a decade. Fill your boots at one end, take care at the other end.”
It is widely understood that the pivotal US market is at the “greed and complacency” end of that behaviour spectrum, with the risk of a 50%+ fall from which it might take years to recover.
But timing of that event? Nobody knows. We have been saying that for some time. But that doesn’t make it less frustrating and stressful!
It’s worse if your less risk-conscious friend says “But I’ve made great profits”. As Kindleberger says in one of my favourite books:
“There is nothing as disturbing to one’s well-being and judgement as to see a friend get rich”
Of course we don’t have the benefit of hindsight. But why not use artificial intelligence (AI) so that we can fine-tune that timing issue to perfection, and get out the day before the market plummets?
It’s not as daft a question as it sounds, and answering it does shine a light on what investing is all about. A recent article on doctors and the use of artificial intelligence provides an illuminating parallel. The writer, Stephen McBride, began:
“At 5am our 7-month-old son had woken up with angry red blotches all over his body. Measles? Instead of going to A&E, I remembered my friend’s sage advice: You should ask AI.
I snapped two photos of the rash and asked ChatGPT to evaluate. In less than five seconds, ChatGPT said it was likely a common viral infection, nothing more serious. It also gave me specific warning signs that would warrant a trip to the A&E.”
We’ve probably all learned to be wary of Googling medical symptoms, but this is very different from what we might call “Doctor GPT”. The writer continues:
“Doctor GPT is already much better than a human doctor… This is not hyperbole. ChatGPT aced the United States Medical Licensing Exam, the grueling, three-part marathon required to practice medicine.
And when Microsoft pitted its AI against a team of human specialists on the toughest cases in the New England Journal of Medicine, the human experts got the diagnosis right 20% of the time. In contrast AI scored 85%.”
AI beat the human experts by a multiple of 4. Wow. Stephen continues:
“AI is almost perfectly suited to solving complex medical mysteries. Biology is staggeringly complicated. A single human genome contains three billion letters of code. The number of potential interactions between proteins in our bodies is infinite.”
No human brain can even begin to grasp it all. But AI can read every medical paper ever written and quickly spot patterns no human could hope to mimic, not even the most highly qualified and experienced doctor.
“Think about it this way. When you ask a doctor, “How serious is it?" you're getting an educated guess based on that one doctor’s personal experience. In contrast, when you ask AI, you’re tapping into a system that’s “seen” a million times more cases than the most experienced doctor.”
Investment is also “staggeringly complicated”, but there is no science to solve the “complex investment mystery”, because so much of what occurs in financial markets is random.
Medicine is governed by biology. Human bodies are extraordinarily intricate, but they follow rules of chemistry, physiology, and genetics. With more data and deeper understanding, AI can help by spot patterns doctors miss, cross-referencing huge amounts of medical literature.
In contrast, financial markets are not just complex systems but random ones. Prices reflect the aggregate decisions of millions of people responding to new information, expectations, and emotions. Much of the movement is noise, not signal. As a result future prices are not reliably predictable. The problem here is not merely incomplete knowledge, but fundamental randomness.
As ChatGPT put it to me:
In medicine, AI is like turning on brighter lights in a dimly lit room. The more light you have, the more of reality you can uncover.
In financial markets, AI is more like trying to forecast the roll of a dice. You can never eliminate the randomness, which is unknowable by its nature.
Sounds like our job is impossible! But it isn’t. There are enough precedents to guide us, and we know that at the “greed and complacency” stage it will invariably be a little frustrating.
But once the bubble bursts and the “reset” has occurred, it will be very exciting (for us all!) as the cycle starts over again.