What is Unsustainable Will Not Sustain

The market is hanging off the edge.  Hyperscalers are going into negative free cash flow to chase AI dreams while semiconductor companies reap all the benefits.  The hyperscalers have bought into the Kool Aid, thinking that there is a big pot at the end of the rainbow for having the most AI compute.  But that's not how Wall Street is viewing the situation.  Their stock prices are going down the more they spend on AI.  Hyperscaler CEOs are rewarded when their stock prices go up.  They are hurting themselves chasing their AI fantasies.

The current situation is unsustainable.  The return on AI investment is uncertain, and likely to disappoint.  Hyperscalers' infatuation with AI is the foundation of the momentum trade.  If corporate executives finally realize that spending more will hurt their stock prices, then naturally the next step is spending less on AI capex vs. consensus.  That isn't in the forecast for many investors.  It would send out seismic reverberations to the semiconductor/AI hardware space.  It appears the market is sniffing out a sea change in the behavior of corporate CEOs, as more and more, token efficiency is being emphasized over tokenmaxxing.  The past couple of weeks has been quite weak for semiconductors relative to hyperscalers (GOOG, META, MSFT, AMZN).  

Without the momentum stocks leading, rallies are less potent and lack staying power. Since June 22, momentum has underperformed the SPX by 7% in an uptrending market.  

Momentum is the oxygen that breathes life into a bull market.  There was a historical outperformance of momentum stocks from the start of April to the start of June.  SPX went straight up, on the coattails of the momentum stocks.  The backbone of that momentum was AI capex.  It reminds me of the Nasdaq going parabolic from January to March of 2000, as semiconductors went bonkers.  History doesn't repeat, but it does rhyme.  From April to September 2000, the Nasdaq and the SPX both entered into a volatile, but range bound market.  It was the topping process that would lead to a 2 year bear market. 

Will we get something similar this time after the AI bubble pops?  Its possible, but unlikely.  The government and central banks are much more interventionist, much less laissez faire.  In a bear market, I would expect the Fed to rapidly cut rates, restart QE, and when those measures fail, start buying US equities with Congressional approval.  The US economy is heavily financialized, so a long extended bear market like the post dotcom bubble would be much worse for the overall economy.  

Let's not get ahead of ourselves.  Its still a bull market.  In a bull market, its much easier to make money being long than being short.  If you are going to short, you have to get very good entries.   Even if you are wrong about the bull market ending, good entries will not hurt you much, or even make you money.  For short-biased traders, the default position in a bull market should be to be in cash.  That is the only way to survive and thrive when the bear market finally comes.  

But the bull market is showing its age.  There is a growing pile of evidence that suggests that this bull is very close to ending.   

Since the start of June, you have seen some classic signs of a top.  

1. Choppy, high volatility price action in momentum stocks after a big move higher.   

 

2. Underperformance of momentum stocks/market leaders.  

3. Heavy inflows into equity funds, especially those with higher beta (tech).  

4. Heavy retail investor participation in higher beta stocks.  

5. Lots of equity issuance (IPOs, secondaries).

The extremely high two way volatility of the momentum stocks, while the overall market's realized vol is much lower, is a warning sign.  Momentum stocks lagging lower beta stocks is another.  Popular stocks/assets among retail investors underperforming the overall index is another.  But retail continues to buy more, even as their favorites stocks go down.  The STAX index, Schwab's index of retail buy/sell activity, shows the largest amount of buying since the bull market started in late 2022.  

STAX Index vs SPX

 

Let's get into the psyche of the retail investor.  They've made a lot of money buying the Mag7 since 2023.  But those profits are slowly melting away, as the Mag7 go down, even as the SPX goes up.  They are feeling some pain now for loading up on high beta growth stocks.  But they maintain their faith for the time being, because its worked for so many years, and the overall market is still doing OK.  

Since Trump go re-elected in November 2024, retail has gotten heavily involved in meme stocks and assets, like bitcoin, ether, quantum, nuclear, space, gold, silver, etc.  One by one, those holdings are going down, and not just small down moves, but big haircuts.  Bitcoin got cut in half, from $120K to $60K.  A lot of the meme stocks like IONQ, RGTI, OKLO, ASTS are down anywhere from 50% to 80% from their highs, in less than a year.  All of this is happening as the SPX is near all time highs.  These investors have gotten burned repeatedly chasing meme assets and meme stocks, even in a overall bull market.  They went for the most speculative assets to try to make fast money, only to lose huge chunks.  Eventually, retail investors will shy away from chasing the hype that isn't backed by earnings fundamentals.  It is already happening.  Look at how quickly the SPCX hype has died down. After having gone above 200 the week after the IPO, its back down under 150.  Almost all post IPO SPCX buyers are now underwater.  

You had the SK Hynix IPO in the US today.  In hindsight, the weakness in the AI semiconductor/hardware names could have been fund managers selling ahead of the IPO to raise cash to buy SK Hynix.  Just like you saw weakness ahead of the SPCX IPO, after it was over, you had a couple days of the market bouncing as the selling pressure abated.  I could foresee a similar situation here with the SK Hynix IPO out of the way.  Plus the first half of opex weeks are usually bullish.  Waiting for a really good entry point to short.  We are getting there.  If we get a further rally on Monday/Tuesday, I plan on putting on tech stock and index shorts.  We are getting close to the end of the seasonally bullish time period, and about to enter a seasonally bearish time of the year, from mid July to late September.  With all the bearish signals in recent weeks, it is about time to get more aggressive on the short side for the coming weeks and months.