Can AI stocks lead the market to a new bull?


I have had a number of discussions with investors and the question keeps coming up. Can the AI frenzy, which appears to be in its early stages, carry the stock market to a new bull?


The AI frenzy begins

A narrative is emerging that AI will become a highly disruptive force, much like the internet was in the 1990s. Indeed, the emergence of natural language processing like ChatGPT has the potential to transform the nature of work in the coming years.


A gold rush is developing. NVDIA has become the poster child for selling the picks and
shovels of AI, much like Cisco and Oracle were during the internet gold rush. As one analyst aptly put it, “There is a war in AI, and NVDIA is the only arms dealer in town.”


But market breadth has become extremely narrow, worse than it was during the internet bubble. Can an AI gold rush propel the stock market to a net bull?


I think that may be the wrong question to ask.


Combining momentum and trend following

I base my conclusions on a study by a number of academics at the Bayes Business School at the City University of London. The study is entitled, “The Trend is Our Friend: Risk Parity, Momentum and Trend Following in Global Asset Allocation”.


To make a long story short, the researchers found that investors can achieve superior risk-adjusted returns by combining price momentum with trend-following models. The strategy can be summarized as, “In a bull market, buy the hot stocks of the day”. Buy the flavour of the day. It doesn’t matter if they’re biotech, uranium, social media or AI stocks. 


The caveat is price momentum — a strategy of piling into winners — which works only with a trend-following filter. In a neutral or negative trend (bear) market, momentum doesn’t work very well.
My own research that uses sectors instead of individual stocks as a momentum strategy also confirms these results. Pile into the winners in an uptrend. Avoid momentum in a neutral or downtrend.
That’s why the question of whether am AI frenzy can carry the stock market to a new bull is the wrong question to ask. The right question is whether the stock market can achieve a bull trend and propel AI stocks to an internet-style bubble. For that to happen, a technical bull market where the index rises 20% from its low isn’t enough. Market leadership needs to broaden out.


One measure of the breadth of participation is the Advance-Decline Line. As the S&P 500 tests overhead resistance, different versions of the A-D Line are well below their highs. I would like to see these indicators strengthen further as signals of broadening participation.


Also don’t forget the venerable Dow Jones Industrials Average. Both the Dow and the Transportation Average are also well below their old highs.



An alterative explanation

If breadth were not to broaden in a definitive fashion, here is an alternative explanation of the current market structure. Recall Bob Farrell’s Rule #7: “Markets are strongest when they are broad and weakest when they narrow to a handful of names”.


An analysis of the relative returns of value and growth shows that growth stocks have dominated value in 2023. While the dominance continues within large-cap stocks, value has begun to regain relative strength within small caps. In other words, value/growth dominance is beginning to show signs of cracking.


The analysis of quality within value and growth is equally revealing. There are many ways of measuring company quality. One simple way is to measure the relative performance of the S&P and Russell indices. S&P has a much stricter profitability inclusion criteria than the FTSE/Russell indices. Consequently, the companies in the S&P indices are more profitable and have fewer money-losing companies than the stocks in the Russell indices.


While growth outperformed value, what we can observe is the more profitable S&P 500 value stocks tunderperformed but performed better than he less profitable and lower-quality Russell 1000 stocks (top panel). The middle panel shows that high-quality value beat low-quality value, but the bottom panel shows the low-quality growth stocks outperformed high-quality growth stocks.


In other words, the market structure is being dominated by a group of low-quality growth, while high quality is still dominant within the value universe.

Low-quality growth dominance and high-quality value? That should like a frothy market poised for a recession – which is the alternative and conventional explanation for the current market structure.


Investment conclusions

We began this publication with the rhetoric of whether AI frenzy, which appears to be in its early stages, carry the stock market to a new bull. Academic studies indicate that the right question is to determine whether we are in an equity bull so that AI stocks can take advantage of the bull trend. For that to happen, breadth participation needs to broaden out considerably.


The other and more conventional explanation of the current market structure is that this is a frothy market dominated by growth, while quality stocks are outperforming within the value universe. This is suggestive of an unhealthy and unsustainable theme-driven advance against a backdrop of a market that’s positioning for a recession.


1 thought on “Can AI stocks lead the market to a new bull?

  1. It all depends on the money flow. Right now there is this mania for AI.
    Computers are not new, the internet is not new, Chat GPT is new for me, but people have been working on AI since Turing, hence the name of the test.
    So what does this mean? Will AI change things this week, this month, even this year? Odds are no. Structural changes don’t happen overnight. How dramatic an effect on the world has blockchain been over the last 10 years?
    Will NVDA be selling an ever increasing amount of chips for 10 years?
    Why did GME jump?
    Frenzy is the right word. A mania is just that. There is no way to call the top, and one only knows that it is over when there is a lot of damage done.
    The effect of the internet over the last 30 years has been profound, but Cisco has not performed well since 1999. Perhaps 30 years from now AI will have had profound effects also (I imagine perfected translation apps that we wear and link to another person so that when I speak English they hear whatever language I choose. “C3PO move aside”). But perhaps they will also say that NVDA has not done well over that 30 years.
    We are overdue a recession and when it comes margins will suffer, earnings will suffer and at some point they will be screaming that the world is about to end. It won’t unless someone cooks up an EbolaMeasles chimera that wipes us out, at which point they will be saying buy gold!
    Call me stupid but I think we are still in a bear market rally, even if we get an ATH, we are in a bear market. It’s not prices that matter, but returns over time. Bear markets the returns are poor, bull markets the returns are good. Put another way, bear markets don’t offer good value for what you pay.
    There has been a huge distortion because of monetary policy. Does anybody see this distortion ending? What did they do about the debt ceiling? Oh yeah, they are pretending it doesn’t exist until after the election. This could make for a frenzy of spending, so maybe the market is not irrational but rational in it’s appreciation of our irrational leaders.
    So, who knows? Anything is possible.

Comments are closed.