24 Nov 24th November 2025

Index Movement Last Week

Finally we have some excitement and a bit selling to talk about. Last week’s market action has been portrayed as the AI bubble bursting but that is simply not true.

The action closely resembles a classic risk-off move. Probably triggered by the strong employment numbers that puts a question mark around future interest rate cuts.

The S&P500 was down -2.0% but MSCI Europe was down more at -2.5% as was Japan -3.3% and Emerging markets were down -3.4%.

If the AI bubble burst story was true then you would see the S&P being the one that falls the most. Rather, we saw US treasuries rally and the US dollar gain against all majors.

Then on Friday odds of a rate cut increased again and stocks went back up with it.

One cannot find a more classic set of risk-off moves than we saw last week. So it was nothing to do with AI and was just a run of the mill adjustment like we see regularly.

Main stream media doesn’t know what it is talking about. But it does love the bear’s story as it makes better false headlines.

The bear’s story is the AI trade is over. The bubble has burst. Everything AI is overvalued. We don’t yet have any proof AI will increase productivity and earnings. Hyperscalers are spending too much money on AI infrastructure. AI cannot make enough money to pay back this upfront investment. AI is a circular spending cycle to prop itself up without any real income. Data Centers will use too much energy.

Did I miss any other reason?

The Bear’s problem is they do not understand the dynamics of disruptive technology and Moore’s Law

Originally made by Gordon Moore in 1965, Moore’s Law is the observation that the number of transistors on an integrated circuit doubles approximately every two years, leading to increased computing power and lower costs. 

More transistors means more computing power. Over the last 20 years or so, computing power has increased by about 100,000%. That is not a typo.

Even if Moore’s law is slowing, we can safely assume AI has the potential to double computing power every 3 years. That means in 10 years time, computing power will be around 1,000,000% more than it was in 2005. (also not a typo)

What that leads to is great difficulty in markets being about to discount future earnings.

To explain why this makes a difference let me try and explain how markets work. Let’s say there is a company that has very predictable earnings and a very predictable growth outlook. That makes it easy for analysts to predict future earnings and everyone arrives at the same number which leads to the same stock price calculation and hence everyone knows what the stock price should be and so it hardly changes.

AI (and Tech more broadly) is very hard to predict just how much earnings will be made in the future which leads to less accurate predictions. What we have seen in the last 25 years is the market is not good at pricing Tech and Tech tends to grow much faster and become much more profitable than analysts predict today. Which leads to rapidly increasing stock prices.

But, of course, all this unpredictability leads to swings both ways. Which is what we are seeing now.

I don’t know if prices will be higher next week, or next month. But I am 100% certain they will be significantly higher in 3 years because Moore’s law will have doubled computing power. And more power leads to more profits.

So I am a buyer right here.

Plus I know this fact

Since the inception of the index (1985) there have been 72 corrections of 10% or more. On average it has taken 40 days for that correction to occur and the average size of the drop is 16.4%.

But, one year after each of these drops, the index has been up an average of 19.7%

This game really is very simple. Just buy the dip and hold for a year or more.

Warning

Stock values can go down as well as up. It is possible to lose 100% of your investment in a stock. Any advice given by Capital 19 is general advice only and does not take your personal circumstances into account and might not be suitable for you.