The End Of The Hype Cycle
This is the weekly market report for week ending February 9th, 2024.
My overall hypothesis is that extreme asset inflation is obscuring cycle risk.
At this latent juncture there are few if any signs of cracks in the economy. The jobs market and GDP are both strong at the same time, which is what one would expect at the peak of the cycle. However, economic strength is not enough to prove the cycle is ending. The economy is strong because the deficit is 6% of GDP. Had the deficit been this large in prior cycles since WWII, there would have been no recessions in the past 70 years. But, the dollar would be likely worthless by now. Therefore, proof of cycle risk at this juncture comes not via the economy but via the financial markets. I believe that deleveraging will begin in markets and then carry over to the economy, instead of the other way around. Financial markets, including stocks, Cryptos, and real estate are reaching a point of froth seen at key inflection points during the past four years. This latest melt-up is driven under the auspices of "Artificial Intelligence". The AI bubble follows in relatively quick succession the pandemic-driven Cloud/Work-from-home bubble which culminated in February 2021, the subsequent crash led to the total shutdown of the IPO market. That was followed by the melt-up of the Crypto/EV/Metaverse melt-up which culminated in November 2021. The crash led to the bear market of 2022. This latest AI bubble began over a year ago at the nadir of the 2022 bear market and is culminating right now. As I've pointed out on Twitter, February has been a locus of manic melt-up/crash four years in a row: 2020, 2021, 2022, 2023.
The USA Momentum Factor best exhibits the valleys and peaks that define this annualized pattern.
Of the several bubbles that already failed, the AI bubble is the least believable and fraught with the most market divergences. Outside of Nvidia, Microsoft, and Meta, there is almost no sign of tangible AI benefit. Bellwether Taiwan Semiconductor has seen a mere 8% increase in sales year over year. Total semiconductor global trade remains below the 2021 peak as of December 2023. As can be viewed here.
What's most ironic about the AI bubble is that there has been no read through to the pure play AI stocks. It's almost entirely a hardware driven bubble meaning "If we build it they will come mentality". The software and services AI stocks peaked back in 2021.
Tech companies are using AI as an excuse for mass layoffs, which of course is highly pleasing to Wall Street - to hear about massive job cuts AND imaginary AI growth on the same earnings call.
February 5th, 2024:
"The tech industry has started 2024 with another wave of job cuts, paring back even further after widespread layoffs last year"
Although economic factors are the main reason for tech layoffs, Lee noted that many companies are citing the race for artificial intelligence as a factor, as they are shifting resources to focus on AI talent. According to an analysis by CompTIA, which tracks employment trends in the tech industry, job postings in “artificial intelligence or requiring AI skill increased by about 2,000 from December to January, to 17,479.”
FR: "Shifting resources", great euphemism. Layoff 32,000 with the hopes of hiring 17,000 and the reality of hiring a few thousand, since there are no AI experts sitting around waiting for new jobs. It will take executives a while to figure out they're all competing for the exact same few people.
The punchline is this one:
"This year, “tech companies are still trying to correct for their over-hiring during the pandemic surge"
Companies are having to clean up the mess left by the last two failed Tech bubbles in order to make room for this latest Tech bubble. No pundit makes the connection as they all clamber aboard this latest roller coaster ride to nowhere.
All of which information is conveyed in the chart of Taiwan Semiconductor. Note call option volume in the lower pane as investors clamor for exposure. Note: single stock OCC data only goes back 2 years.
Companies that claim they are implementing AI are commanding ludicrous valuations both inside the Tech sector and in other sectors that have no use case for AI.
"As of Thursday’s close, investors are valuing Arm at close to 90 times forward earnings. That compares with a forward price-earnings ratio of 33 for Nvidia and 46 for AMD, which both have significantly higher multiples than other major chip stocks like Intel and Qualcomm"
The poster child for AI froth however is not Nvidia or Arm it's the best-performing $1 billion+ market cap stock in the entire market. I am of course referring to Super Micro Computer, which is up 140% year-to-date.
All of this froth and manic reach for risk is causing chasmic divergences in the stock market. The S&P 500 continues to make new highs this week which are unattended by the majority of stocks. Divergences are in breadth (stock participation, sector participation), volume, and of course new highs.
As it was in February 2020, the average stock is in no way confirming this new all time high.
On the Nasdaq, Hindenburg Omens which signal large numbers of new highs and lows occurring at the same time, have reached November 2007 levels this week. The other reading at this level came in October 2018 and preceded a -20% bear market.
In summary, this era now embeds 1990 style S&L small bank crisis. A Y2K Tech bubble, and a 2007+ level housing bubble all of which is likely coming to a head right now. Meanwhile, the Fed is constrained by 1980 style inflation. All of which makes the odds of a soft landing essentially nil.
