Preface: Explaining our market timing models
We maintain several market timing models, each with differing time horizons. The “
Ultimate Market Timing Model” is a long-term market timing model based on the research outlined in our post,
Building the ultimate market timing model. This model tends to generate only a handful of signals each decade.
The
Trend Asset Allocation Model is an asset allocation model that applies trend-following principles based on the inputs of global stock and commodity prices. This model has a shorter time horizon and tends to turn over about 4-6 times a year. The performance and full details of a model portfolio based on the out-of-sample signals of the Trend Model can be found
here.
My inner trader uses a
trading model, which is a blend of price momentum (is the Trend Model becoming more bullish, or bearish?) and overbought/oversold extremes (don’t buy if the trend is overbought, and vice versa). Subscribers receive real-time alerts of model changes, and a hypothetical trading record of the email alerts is updated weekly
here. The hypothetical trading record of the trading model of the real-time alerts that began in March 2016 is shown below.
The latest signals of each model are as follows:
- Ultimate market timing model: Buy equities (Last changed from “sell” on 28-Jul-2023)*
- Trend Model signal: Bullish (Last changed from “neutral” on 28-Jul-2023)*
- Trading model: Neutral (Last changed from “bullish” on 24-Jan-2024)*
Update schedule: I generally update model readings on my site on weekends. I am also on X/Twitter at @humblestudent. Subscribers receive real-time alerts of trading model changes, and a hypothetical trading record of those email alerts is shown here.
Subscribers can access the latest signal in real time here.
Priced to the point of insanity?
You may have seen the charts of the relative performance of the NASDAQ 100 to S&P 500. The ratio has already exceeded the dot-com peak in 2000. In addition, NYU professor Aswath Damodaran, who is regarded as the dean of company valuations, went on CNBC to say that Nvidia is priced “to the point of insanity”, while the other Magnificent Seven stocks are roughly fairly priced.
While the latest AI-driven mania may seem stretched by historical standards, we would argue that it has a lot further to run before the AI bull is done.
Where’s the froth?
From a technical perspective, here are some key differences between today’s tech boom and the one in the late 1990s.
First, the short-term momentum of the NASDAQ 100 is nothing like the 2000 market top. The black line in the accompanying chart shows the 12-month normalized NASDAQ 100/S&P 500 ratio, which is in the middle of its historical zone. By contrast, the same ratio went off the charts and became extremely overbought in 1999 and 2000.
Where’s the froth?
The market top in 2000 was characterized by a flood of low-quality IPOs which swamped the market that soared to “to point of insanity”. Remember pet.com, or all the B2B and B2C plays? Today, small-cap technology stocks, as represented by QQQJ, are lagging QQQ, the NASDAQ 100 ETF.
Where’s the insanity?
Putting it another way, the dot-com bubble top saw leadership by low-quality stocks. Today, the low-quality stocks are lagging the market. The relative performance of QUAL, which is an ETF of high-quality stocks, is leading the market. As well, the S&P 500 is beating the Russell 1000, which is a proxy for large-cap quality as measured by profitability. Standard & Poors has stricter profit criteria for index inclusion compared to the FTSE/Russell indices.
High-quality leadership is not a sign of market froth.
An extended advance
While I am intermediate-term bullish on large-cap technology and AI stocks, the market is extended in the short run and it can pull back at any time.
Magnificent Seven leadership is narrowing. Only four of the Magnificent Seven has made new highs in the last week. Two, namely Apple and Tesla, are weakening. Alphabet skidded after its earnings report, but it’s trying to rally but hasn’t reached a high.
Other negative breadth divergences seen in the NASDAQ 100 are equally worrisome. Past positive divergences in the percentage of NASDAQ stocks above their 50 dma have worked out well in the past, how will the latest negative divergence resolve itself?
Steve Deppe studied past instances when the S&P 500 finished a calendar week with a trailing 15-week return of 20% or more and a new all-time high. While the sample size is small (n=5), forward returns weren’t promising.
My main takeaway from Deppe’s analysis is the market advance is obviously extended, and this study is just an illustration of what might happen if it were to pull back. The low sample size is insufficient to forecast market weakness, and only highlights possible probabilities. There is now ironclad law that says stock prices have to weaken next week.
Instead of focusing on the technical conditions of the non-tech part of the S&P 500, which is not extended, let’s just consider NASDAQ stocks. The NASDAQ recently flashed a cluster of Hindenburg Omen warnings, which is an indication of a bifurcated market in an uptrend that could be ready to break down. There were 16 such signals in the last 10 years. 10 (pink bars) resolved in downside breaks of different magnitudes while six did not.
What to watch
Here is what I am watching for signs that the rally is faltering. Semiconductor stocks have been a proxy for AI-related strength. The SOX Index remains in an absolute and relative uptrend. Breaks of either channel would be a warning that a corrective phase is about to begin. Nvidia’s earnings report, due on February 21, could be pivotal to the health of the current advance.
Also keep an eye on the regional banks as a barometer for the health of the overall stock market. Anxiety is building in this industry and a break of relative support (bottom panel) could be the trigger for a bearish episode.
In conclusion, I believe the current AI-driven rally has a long way to run. Market mania tops are characterized by excessive froth, which is not in evidence today. However, price momentum in technology stocks is faltering, and these stocks can pull back at any time. We would regard any pullback in these stocks as a buying opportunity.
Here is the longer-term view. The NYSE FANG Plus Index, which is a proxy for large-cap technology and AI-driven stocks, has staged a decisive upside breakout through cup and handle resistance on an absolute basis and breakout of a long base on a relative basis.
What’s not to like?
Was it Chekhov who wrote about happy and unhappy families? Maybe it is the same with market insanities, just like individuals, they are all insane in their own way. What does FOMO have to focus on today?
The fang+ S&P chart is beautiful, so it is a better place to be than the SPY, but if the SPY were to crash, there would be a downdraft in the fang+.
There is no way of knowing when this will end, only that some day it will, and then many will hang on the slope of hope, and will double down on the 2nd wave of the down move setting up a brutal 3rd wave…or the Elliott guys will say.
It’s tough, volatility is great when one is on the right side of the trade, but awful when on the wrong side. It is so much easier to be greedy when one is winning, than withstanding the fear when one is losing.
NVDA is insane, but it can become way more insane. It reminds me of CSCO in the 90s, running from 12 cents to 56 bucks in under 10 years . Volume did drop off in the later part and most importantly had a blow off top that broke through the upper channel, but it’s on a monthly chart so even then it’s months possibly before it breaks. I’m not saying NVDA will do this, but if it breaks out above the upper channel on a monthly chart this could be a sign that it is nearing a top.
What might offer a great risk reward at that point is a put spread…for example, right now a Dec 24 spread on NVDA the 250/300 costs about 1.60 for a max yield of 50 and max loss of 1.60. Of course right now NVDA has not blown out through the upper trend line and perhaps never will. Waiting for the monthly MACD to turn negative on the histogram would add to confidence in the trade. One thing you can count on, is they won’t let you forget, there will be a lot of NVDA noise at the time.
This is the best newsletter I have read. Thanks Cam.
I have studied bubbles since being in my first with the oil bubble of the 1970’s. I took one of my first clients from $50,000 to $1.4 million in four years when oil went from $4 to $45 and then back to $400,000 after oil crashed back to $8.
This A.I. bubble is early stages. Progressing they have multiple times when expert warn of insanity and overvaluation. Anyone who listens and gets out, regrets it when the bubble stocks go higher. Eventually stocks go parabolic when experts are not listened to any more.
Generative A.I. is not just an advanced search tool like novices see. That is a toy compared to where A.I. application experts are taking it. We will see more and more applications in our daily life and hear of amazing happenings that will fan the flame of investor interest. Altman is looking to raise seven trillion dollars for chip making factories. That’s with a T. That is not just for a new powerful Google. People like Black Berry thought the consumer applications on the iPhone were a toy.
It’s extremely hard, almost impossible, to stay with a new, huge thing. It always seems to high. This is why market indexes beat managed money in the long run. Apple goes into the index and thirty years later it is up a hundred fold. The index makers keep holding while real people lighten up.
So the answer is momentum investing. I mentioned the momentum ETFs many months ago and you will find that they have all done extremely well, even outperforming the S&P 500. I expect that will continue from now until the final bubble top.
The last few years had momentum ETFs continually underperforming because leadership kept swinging from Growth to Value every six months. So every time the ETF rebalanced into the new momentum leaders, those stocks lagged as styles reversed.
I thought those leadership swings were over as Growth would keep strong momentum and Value would lag. That has happened. Value took a two week lead after the Powell Pivot mid-December but that was snuffed out and Growth, especially Large Company Quality Growth reclaimed the momentum lead since the start of the year. Banks were a popular Value pick that is now flaming out with CMBS problems.
I track the DJ Market Neutral Momentum ETFs and amazingly it shows almost all industries are experiencing high momentum continually gaining on the same industries over their laggards. The current business environment is separating winners and losers even within industries. A.I. will accentuate those business conditions in future. Momentum-style investing captures the alpha of that.
Holding a passive ETF of weighted outperforming companies shields you from worrying if a company is insanely too high, you sell and then it justifies the price by doubling and tripling without you. One of their holdings might plunge 30% on a bad earnings report but five others go up 10% on positive surprises. One can sleep at night during earnings season. Note how Microsoft had a brief dip on earnings day and now is powering to new highs.
That’s the thing about bubbles, you don’t know when the parabolic rise will end. Altman wanting to raise 7 T for making chips makes me think of the fiberoptic craze of the Dotcom where they made way more fiberoptic stuff than they needed.
Well, if they want to build all those chips, how will they get the electricity? Just another reason to think of picks and shovels like uranium. People may not want power plants near them, but somewhere in the middle of nowhere they might be ok with a small modular powering a massive array of chips. After all it doesn’t matter where the computing is done, it could literally be on the moon.
Waiting for a pullback means you are expecting optimistic investors to swing to fear and pessimism as stocks roll over. That historically true but…..
Markets are more and more driven by unemotional algorithms.
You are right, Ken. 80-90% of daily volume is algo-driven. There are some dumb algos at certain periods of the year. Most of the time the alogs are getting better and more focused, backed by big raw data and then first and second derivative data from self-learning (pretty much today’s AI jargon, but it has been in existence for a long time). Think of recursion or feedback loop in school days.
The outcome is Pareto Principle being enhanced again and again. So the concentration is getting higher. A lot of hedgies exist but they have very little R&D. They simply buy algos from certain shops. The resultant price movement is what we called “piling.” An example is Tiger funds. They are notorious momentum chaser. At this moment what we are seeing is tons of call buying. So we have sharp rise in the share price because of delta nd gamma hedging (1st and 2nd derivative hedging). Many price charts are going vertical aggressively.
The second aspect of Pareto Principle is reflected in performance in market caps. It is getting more and more expensive to operate a company. Starting up a company today is almost impossible. Big corps have all the advantages, plus they can buy politicians to legislate in their favor or get certain handout.
Extended further we get to the international level. Altman’s 7B chip idea is not practical. Don’t blame him because he is from CS ground. CS guys typically are not grounded in reality. You know what I mean. Many corps doing real manufacturing hire very few CS graduates to do software/firmware, instead it’s EE/ME/CompE doing it. He is thinking 100% coverage but that is not possible. Aside from this there exists the fact that AI game is very expensive. And who can get it done today will accelerate the strength divergence among nations. The gap will be insurmountable as time goes on. A true quantum leap experiment.