An important update to my 2023 Bull Thesis

Eventually, the wolf came.

It was lonely going into 2023 ignoring the warnings of Recession, and after having scooped up TSLA and META near their lows, I spent Q1 on SubStack pushing back on every seemingly negative Macro datapoint offered by “the experts”. Where most got it wrong was missing the fact that the Man-made Macro Cycle of 2020-2021 invalidated many of the YoY comparisons, creating many dislocations and false signals that had been reliable during normal market conditions. Fortunately, the Macro models I developed have proprietary methods for looking forward, not backward, and they showed Macro was actually Improving from SPX 3900.

For the prior 4 quarters, most TV analysts, FinTwit traders, and even The Fed cried wolf - the Recession is Coming! Of course, the market makes fools out of most, and as the SPX rose above 4300 in June and then 4500 in July, the “fool me twice” crowd flipped. Incredibly, even as The Fed pushed rates higher again last month, further tightening conditions that should increase the probability of Macro weakness, they also capitulated and changed their forecast to “no recession”. Today, “soft landing” is almost consensus. The boy is no longer crying wolf.

AI is a distraction

Of course, Price is the reason. SPX rallied above all the key moving averages, and the Macro pretenders had to come up with a good reason to abandon their negative views. The excuse is AI - the “next generation internet” that will result in a 1999 move. All of a sudden, the $25T economy was the tail of the AI dog. I believe AI is an incredibly important technology and has plenty of upside for business, but is not with many challenges as it scales. However, it’s not going to move the Macro needle as quickly as many expected, and will follow the adoption curve of any new, expensive technology.

Ironically, I have some interesting history with new technology, and see many parallels. In the 1990s, I started a video conferencing business that grew to 125 employees and did extremely well until I exited in 2012. Video conferencing had been a decent business since the 1980s, but really boomed in mid-2000 with the introduction of a new technology - the Multipoint Control Unit (MCU). The MCU allowed companies to connect multiple video sites simultaneously, expanding the number of applications (distance learning, telemedicine, corporate meetings) not previously possible. A new company, Tandberg, emerged out of nowhere, and caught the incumbents flat-footed. Within a year, big corporations and governments adopted MCUs and had plans to make these services available for their entire organizations. My company was awarded Tandberg’s #1 Federal Govt Sales Partner, which was easy since the MCUs were extremely expensive.

AI is following the same path. Incumbents like Google let AI slip away to start-ups, and although the cost is very high it didn’t stop buyers from writing big checks. But like MCUs, it turns out that providing these new services across the entire organization will require a much bigger additional investment. The laws of physics will not be overcome - MCUs that were purchased for the HQ and major offices were located far away from the end users. The speed of light governs how fast the technology can reach, and in order to serve the same level of service to all users, it would require additional purchases of more MCUs. Companies that wanted to scale the technology would have to buy many more MCUs and install them in more places. In other words, companies would have to write a lot more big checks, and adoption slowed. I believe the speed at which AI will scale is overestimated.

Why does this matter? Valuations are at insane levels (see NVDA), and Revenue won’t grow nearly at the same pace. A reset of these market leaders will shake the confidence of the overall market.

Follow The Data

While AI gets all the attention, contracting leading indicators, tightening credit standards and reduced lending, and all the “recession signals” that haven’t worked over the past 12 months are now ignored. “Mission accomplished” with Inflation coming off the boil, but failing to understand that consumers have done what they always do and spend down their extra savings and run back up their credit cards. Most indicators are already back to 2019 levels, when the cycle was getting closer to the end than the beginning.

Prices are still too high, and “only” growing at 3% doesn’t help. And even more importantly, the safety of the “fixed 3-4% mortgage” is a fallacy - home buyers are paying much more on escrow items (PMI, insurance, property taxes, HOA) and are finding that their “fixed monthly payments” are now growing. Add in rising cost of utilities and maintenance, and monthly budgets are stressed. Yes, new 7% mortgages don’t allow homeowners to move up, but overall rising monthly payments are also forcing some to move down. Home equity is still strong, and cashing out is an option that might accelerate as home values continue to decline.

In short, the consumer could care less about AI, and they are 70% of Macro.

Free MacroSPX information

I’ve got some great long-term (since 2015) Pro clients where I share proprietary Macro output and even my futures, options and equity positions. I don’t normally share this publicly, but I hope they don’t mind this time.

I’m bullish when Macro is Improving, and I’m bearish when Macro is Declining. Pretty simple, but it’s kept me away from the crowd. In June, my Macro models starting flashing some warning signs, and worsened further in July. MarketModel went to ALL CASH in June, and has missed the rally since. MarketModel has never had FOMO, and is comfortable sitting on 21% gains for the year. Since 2022, each time the model went to Long 0%, that marked a near immediate selloff in the market. But not always, and clients started asking for other periods where MarketModel was risk-off for long periods, and two came to mind - March 2022 and December 2017.

Interestingly, SPX has been unable to get above April 2022 highs on this latest move. At first, I looked at March 2022 as a similar early signal, but TA was not the same - 2022 was a Bear Market, 2023 is a Bull Market. Any model can be wrong, and in 2021 the model was fighting historic stimulus and Fed easing, so it is possible that it gets it wrong again. But December 2017 was starting to look like a more familiar Bull Market, when Macro Downgraded early and watched SPX rise through January. Before the Fitch downgrade, I shared the following article with Morgan Stanley clients and Michael Santoli.

Feb 2018

I see many parallels - falling Macro, persistent bid rising above the Rails, spiking interest rates, and a new XIV trade that backfired. Today, 0DTE has seemed to compress the daily range, and kill the VIX. Like XIV, it works until it doesn’t, and result in an unusual expansion of the range. Liquidity is falling in the bond market as QT continues, and a combination of rising yields and falling Macro are the ingredients of a deep correction. It didn’t seem like anything could stop the rise in early 2018, until the shock of a China Tariff set things off. Could the Fitch downgrade, which many are downplaying as a 2011 fakeout, be the trigger? No guarantees, and Support (20 and 50 day MA) have yet to be violated, a necessary break before Bears can start getting excited. But if that happens, another daily 1000 point DOW drop and a quick 10% correction like Feb 2018 would not be shocking.