Late Stage of the AI Bubble

The internet is to AI what the airplane is to auto pilot.  Yet here we are, with investors comparing the current AI bubble with the internet bubble of the late 90s/2000.  The use cases for the internet were enormous.  The practical use cases for AI are limited and more specialized.  The internet actually saves energy, while AI consumes huge amounts of energy.  As the years go by, technological innovation continues, but the productivity gains get smaller and smaller.  Mankind have mostly picked the low hanging fruit when it comes to productivity gains through technology.  The internet was a huge productivity booster.  AI will be much less so, and at much higher costs due to the high energy use for unreliable, flawed results.  

While this AI bubble is much smaller than the internet bubble, it is more ridiculous because none of these big tech companies spending multi billions on AI have a clue on how to profitably monetize the technology.  Considering how much they have to pay to NVDA for the chips and the high maintenance costs, it requires a lot of revenue to cover the expenses.  We are still in the grace period where spending willy nilly on AI is still rewarded and even encouraged by the stock market, but that won't last forever.  Look at what happened with Facebook stock when they spent all those billions on the metaverse.  It got crushed and only recovered when Zuckerberg gave up on his dreams.  I see a similar situation where the stock market will start expecting results from all the AI spending, and will start punishing companies that keeping pouring money into the AI fire pit.  This is a controversial statement, but AI is closer to the metaverse bubble than the internet bubble.

Since the AI bubble began in 2021, first as a small niche and now as a burgeoning sector of the stock market, the hype, mentions, buzz, and excitement about AI is as intense as ever.  It is getting irrational out there.  NVDA exploding higher was the main catalyst, but its spawned "satellite" AI plays in semiconductors like TSM, AVGO, etc., in hardware like DELL and HP, utility companies that would benefit from greater electricity demand, and now, end users like AAPL, which ended up being a buy the rumor, buy the news reaction on their AI plans at the WWDC conference.  That is how big this AI bubble is becoming.  Even with AAPL just using OpenAI and not getting paid for any of the extra AI features, its getting a big boost and up 25% in less than 2 months.  The chase for anything AI related, based on extremely optimistic projections has infiltrated the stock market, and gotten retail investors buying into the hype.  Like all bubbles, retail investors are the last to get the memo and buy the most right before the bubble pops, add on the way down, and ride it all the way to the bottom.  

It is uncommon for such a bubble to happen so soon after a previous bubble, the Everything Bubble of 2020/2021, burst just 2 and 1/2 years earlier.  But investors have short memories, and the FOMO mentality is pervasive among retail investors as well as institutional investors.  The US stock market, in the form of SPX and NDX, are the gift that keep giving, which reinforces the uptrend.  It is fascinating to see investors so confident even as bond yields stay higher than most expected.  Here is a look at the household percentage of financial assets held in equities as of end of March.  This number is higher now as the S&P 500 is up another 2% since then.  

These are nosebleed levels for equity allocation among households.  Even during the dotcom bubble in 2000, the equity allocation wasn't this high.  Its slightly higher than those crazy times in 2021 when it seemed everything went up huge with investors all in on stocks.  This high of a level of equity ownership among households magnifies the wealth effect of stocks.  Don't be surprised to see consumer spending vacillate with the ups and downs of the SPX in the coming years.  The US economy is now heavily financialized, making the stock market the most powerful leading indicator of the US economy.  

As you can see from the above chart, when equities held as a percentage of financial assets gets high, a bear market is just around the corner.  

We have the CPI report and FOMC meeting today.  There seems to be quite a bit of complacency going into these events.  Based on what I hear on CNBC and Bloomberg, most are expecting a dovish Powell and a rally after the FOMC meeting.  Since the Powell dovish pivot in October 2023, the market has rallied at almost every FOMC meeting, as Powell has talked dovish at each one.  Will he repeat his dovish performance again?  I wouldn't be surprised if he did, as he seems more worried about his reappointment by the next President than he is about appropriate monetary policy.  

There continues to be breadth divergences as we got another Hindenburg Omen in the Nasdaq and more signs of a split market where a select group of tech stocks keep going higher, while a majority of stocks are unable to rally and keep up with the SPX and NDX.  Global risk appetite appears to be abating as Europe and Asia trade much weaker than the US.  Bitcoin, gold, and silver are starting to falter.   The leadership is getting narrower, and other periods where SPX kept making new highs with so many laggards led to a sudden pullback, e.g September 2014, July/August 2015, October 2018, April/May 2019, July/August 2019, and November/December 2021.  The only time within the past 10 years where the market ignored these divergences and kept going higher was in 2017.  That cannot be ruled out, but 2017 had much lower valuations than now and investors had much lower equity allocations, leaving them with much more room to add equities.  

Remaining short SPX and holding.  Looking for a 4-5% pullback from current levels to reset the bullishness and consolidate the big gains so far this year.