#219: Deep Momentum

micksonmatch
January 27, 2025

Happy Australia Day to those celebrating.

S2N Spotlight

Here is one approach to see how deep the momentum in an asset class is. I measure the gap between the current price and the 200-week simple moving average. The study is not trying to forecast the probability of a particular asset continuing up. It is rather trying to compare how deep the trend is at the moment in comparison to other assets.

It is fitting to start with the star of the last year, Nvidia, and work our way down. The last 3 charts represent downtrends; remember the 10-year Treasury note is showing the yield, which is the inverse of its price.

I think it is important to realise that catching a trend is an enormous gift. However, too much of a good thing can be fatal. We are experiencing very large gaps in the charts presented below. Newton’s first and second laws of motion will ensure that the gap will close.

S2N Observations

I have often shared my Shiller PE or CAPE (same thing) models on US returns. I also share a regression showing that future returns are much lower over 10 years the higher the current Shiller PE is. I came across this chart that I thought I must share that does the same analysis across multiple countries stock prices and earnings. As I don’t have the data to do this study, I thought it was useful to share and reinforce that overvaluation in the end leads to a return to normal. This is a continuation of the theme in the Spotlight.

Last week I shared how blown away I was with DeepSeek, the latest AI powerhouse.

This weekend the news has all been about DeepSeek and its impact on the AI space, with competitors like OpenAI (ChatGPT) and others seriously worried.

“Deepseek R1 is one of the most amazing and impressive breakthroughs I’ve ever seen—and as open source, a profound gift to the world.” (Marc Andreesen)

“To see the DeepSeep new model, it’s super impressive in terms of both how they have really effectively done an open-source model that does this inference-time compute, and is super-compute efficient.” Microsoft CEO Satya Nadella.

Over the weekend, Sam Altman was in panic mode reversing a decision by the company. “ok we heard y’all. plus tier will get 100 o3-mini queries per DAY (!)”

The reason I am sharing this is because we have witnessed one of the biggest booms on the back of the AI revolution. Companies are trading at valuations that would make Benjamin Graham sob like a Taylor Swiftie. The question that I have asked and people far cleverer than me is how will these companies make money?

We are witnessing a new dawn. Previously there were only 1 or 2 competitors at most, but the game has been blown wide open, and we as consumers of the AI tech will benefit, just like we did as purchasers of computers, smartphones, and other technologies that soon became commoditised and prices dropped.

I foresee the same thing happening here with AI. However, the drop in prices is going to see multiple AI heavyweights go bust. I would not be buying into the current AI hype at nosebleed valuations. I would in fact be taking some chips off the table.

I want to share one last anecdote on the subject. If you have spent enough time with AI models, you will see that the models themselves demonstrate an aptitude. Some models are good at some things and weak at others.

ChatGPT is fast losing the advantage that it held at the beginning of this revolution. I am not nearly as impressed with it as I am with Claude.ai and DeepSeek. We are entering a new phase in this AI war, and I am not sure OpenAI, the parent of ChatGPT, is going to remain the world superpower.

S2N Screener Alerts

The Chinese currency has been weakening over the last few years; Friday saw a 3-sigma strengthening. I think one of the reasons is DeepSeek. The Chinese may be further along with their AI development than was previously thought.