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.  

Chinks in the AI Armor

We are starting to see changes in the hyperscalers.  META is following in the footsteps of xAI, cutting its losses on its AI ambitions.  META announced they are going to be renting out their excess AI cloud capacity, and it resulted in a big jump in the stock.  Maybe AI compute capacity isn't as tight as people assume.  Judging by the reaction to the META announcement, the stock market is rewarding hyperscalers cutting costs, and punishing those that mindlessly dump more money into AI capex.  Eventually, the hyperscaler CEOs will get the message and start reducing their futures capex plans.  It could come as soon as the next tech earnings reports in late July.  If MSFT, AMZN, GOOGL, and META signal lower than expected AI capex, that could be an atomic bomb dropped on MU, SNDK, NVDA, DRAM, SOXX, AMD, INTC, DELL, and a slew of other AI hardware/semi names.  

The AI momentum trade has hit a speed bump.  It is coming at the same time that the SPX has been in a short term rally.  This is the first time this year that the AI infra/hardware trade is underperforming the market as it is going up.  The biggest trade of the year is showing a change of character.  This comes after MU had a huge gap up after a big earning beat, and then sold off hard afterwards.  It has been only a few days, so I wouldn't say its game over for the AI trade, but it makes me more interested in shorting the index this month on a rally towards SPX 7600.  

It is interesting that CNBC seems to be more bullish on the overall market now that the non-tech stocks are rallying while the semiconductors are pulling back.  This is the exact situation where they should be more bearish.  Ever since 2020, rallies led by tech and momentum stocks have been much more durable than rallies led by a lot of non-tech stocks and small caps.  

Investors have poured in record amounts into US equities so far in 2026.  

Retail has been aggressively buying dips in 2026.  They've also been buying quite a lot on up days as well.  Both are above the lofty 2021 levels.  Given the underperformance of the heavily retailed owned Mag7 this year, it looks like they are doubling down as their holdings go lower.  Not a pretty picture.  Bad things happen when retail investors refuse to cut their losses and instead add to their losers.  The bigger your position, the weaker your hand.  Retail investors are in a weakened position right now, and it is beginning to show in their favorite names.  
 

Got some outflows from US equities this week, but money continues to rotate into tech stocks.  When tech stocks underperform while getting big inflows, it catches my attention.  

 

Only halfway through 2026 and IPO proceeds already above the SPAC mania 2021 levels.  More big IPOs are lined up for the 2nd half of the year.  Supply is coming out big time to meet the demand for US equities.  

SK Hynix has pulled forward its US IPO plans to July 10.  It was originally planned for August, but it looks like they want to sell shares while the iron is hot.  

The evidence is undeniable.  Equity supply is increasing at the fastest rate since the dotcom bubble.  Add to that the reduction in stock buybacks by big cap tech.  It creates a supply-demand dynamic that requires big investor inflows to keep the bull market going.  If we just get a normalization of inflows to just the average over the past 10 years, the demand won't be able to keep up with all the new supply.  Equities are priced at the margin.  With passive investing dominating equity flows, the stock market is much more price inelastic than before.  A small drop in demand, a small increase in supply could result in a big drop in price.  You could see some big time volatility if we see equity outflows when these big IPOs have their lockup expirations later in the year and into 2027.  

Last week, we finally heard the crowd recognize that Mag7 has been performing poorly this year.  The crowd just about nailed the bottom in the Mag7, as it has been outperforming the AI momentum names like MU, SNDK, etc. since then.  Either the Mag7 or the AI infrastructure momentum names have to lead to keep this aging bull going.  If they are lagging the indexes, the days will be numbered for this bull market.  There is no way that non-tech stocks will be able to lead this market for long.  Investors don't buy US stocks to get long low growth health care, financials, utilities, staples, or old fashioned industrials.  That doesn't drive a long term up cycle.  They are hedge fund pod shop rotation plays that have a limited shelf life.  

Bonds are trading much firmer in recent weeks as oil prices come down and the AI capex trade cools off.  CTAs are near max bearish on Treasuries and near max long on equities.  If you get equities selling off, that will trigger CTA sales in the indices, and probably CTA buys in Treasuries.  I expect a long awaited revival of Treasuries as a negative correlation, flight to quality asset.  It won't be like the QE years, but I do expect somewhat lower bond yields on equity market selloffs.  You saw a hint of that in late June.  

The next 2 weeks are a seasonally strong time period that aligns with start of the 2nd half inflows into equities.   Its almost time to put on long term shorts that could be held for months.  Seeing more signs of a meaningful top in the AI momentum trade, NDX trading weaker than SPX, and value outperforming growth.  All signs of a coming trend change for this bull market.  Holding a lot of dry powder, waiting to deploy shorts on rallies over the next 2 weeks.  Still think that non-AI infra tech is the place to be on the short side, even though the AI momentum trade looks toppy.  It may be less reward to short names that aren't up so much on the year, but its also much less risk.  There are plenty of non-AI infra tech names that are overvalued and underperforming the Nasdaq.  Lots of potential downside out there even in non AI hardware/semiconductor tech names.  

Sticks of Support

Post opex weakness came as advertised, even as oil prices and bond yields went lower. A blowout MU earnings report resulted in a huge gap up that was aggressively sold at the cash open.  Good news, bad price action.  

We continue to see a very split market, with lots of highs among the semiconductors and lots of lows spread out among a variety of sectors.  Semiconductors continue to outperform  the Mag 7.  The semis have held up the SPX, but their strength has wobbled this week.  Another Hindenburg Omen fired off this week.  There is growing evidence that we are making a significant market top.  

As the market is showing topping signs, $119B went into US equity funds for the week ending June 17.  Easily the biggest weekly equity fund inflow since 2015. 

Retail is aggressively participating in this market, as of June 15.  

Starting this week, retail investors have been on a buyer's strike.  You are seeing less buying in the dark pools, meaning less retail buying this week.  The DIX index has plummeted.  Low DIX readings have often marked short term bottoms as retail sells. 

This week, investors finally pulled money out of US equity funds for the first time in 3 months.  Its a drop in the bucket compared to the deluge of inflows over the past 3 months.   

 

As of June 16, hedge funds have historically high beta exposure.  

As of mid June, GS Prime Broker data shows hedge funds have accumulated heavily in US equities and macro products over the past 20 days.  

It is unusual to see such heavy inflows into US equity funds while the market is chopping violently from the top to the bottom of the range (SPX 7300 to 7600).  Perhaps it is the Iran-US peace talks and opening of the Strait that is getting investors more bullish.  It could just be investors seeing the strong uptrend since the end of March and feeling like 2026 will be like 2025, where the market bottomed in the spring and went up continuously for 6 months. They are feeling FOMO.  

Realized volatility is increasing, but the VIX is well behaved.  The put/call ratios have been increasing since early June, as the uptrend has turned into a violent, choppy sideways range.  Options traders have gotten much less bullish.  ISEE index is now back down to more normal levels.  

We had a big drop in the net position of commercial traders (dealers) after June opex.  Before June triple witching expiration, there was a huge amount of calls that dealers were short (customers were long) that they hedged by going long SPX futures.  The unwind of that position resulted in commercial traders having the biggest net short position in SPX since early 2025. 

SPX Net Position of Commercial Traders

The US stock market is going up because of the AI boom.  Hyperscaler AI capex spending is driving economic growth.  But the current situation is unsustainable.  The ridiculous prices for memory, GPUs, storage, etc. are not compatible with hyperscaler profitability.  MU, SK Hynix, and Samsung's profits are coming at the expense of OpenAI, Anthropic, SPCX, MSFT, NVDA, AMZN, GOOG, META, and ORCL.  These mega tech companies will not continue to burn cash in a bottomless money pit if their stock prices keep going lower.  

The Mag7 continues to lag the Nasdaq 100.  The Mag 7 is down 7% YTD while the SPX is up 8%. It is wild to see the Mag 7 down 7% this year considering how much money has flown into US equities YTD.  I finally heard CNBC commentators mention this Mag7 underperformance this week.  It probably means we are near the end of this trend.  I expect the hyperscalers to begin to rein in their spending on AI capex, as well as AI token use.  That's likely to happen as soon as the next earnings report season in late July.  That would be bad news for the semiconductors if that happens.  

You are already seeing a shift in the share of tokens used from more expensive US models to cheaper Chinese models.  The Chinese models are rapidly catching up with the US LLMs, and I expect more companies to use slightly worse models that can get the job done for much cheaper.  

Into the post opex weakness, I covered my short tech short positions.  We are now around the start of a seasonally strong period for the market, near the end of June into middle of July.  Given the big unwind in long call positions from speculators, evidence of retail selling in DIX index, and higher put/call ratios, we could see a bounce towards the upper end of the SPX 7300 to 7600 range.  Bought some gold into the weakness this week, looking for a technical dead cat bounce next week.  Big picture, seeing classic signs of topping.  But I expect the speculators will give one more push higher before we get the big move lower in late July/early August.

No More Excuses

After a long string of fake headlines and market manipulation, the US and Iran have finally signed the MOU.  Oil is now in the 70s.  The wall of worry over war, high oil prices, and supply chain disruptions has been obliterated.  There are no more excuses for this market to go down.  No, a hawkish Fed doesn't count.  Warsh put out some tough talk on price stability, but its empty words without real action.  Most don't believe he has the balls to hike rates.  I tend to agree.  But if this bubble doesn't pop this year, Warsh could definitely do a couple rate hikes.  The SOFR curve is pricing in about 2 rate hikes for the next 12 months, so even if he does hike, its mostly priced in.  

On Monday, after Trump confirmed that there was a deal over the weekend we got a huge gap up in SPX/NDX.  I was a bit surprised by the size of the gap up, and it held for all of Monday.  But just as I was getting ready to add on Wednesday FOMC day, it started to weaken on Tuesday and didn't give me the premium entry to add to shorts.  Most of the non-AI tech stocks that I was looking to add shorts to had topped out on that Monday gap up, and were lagging the market badly by Tuesday.  With some short exposure already, I will not chase weakness to add.  Still staying away from the strongest AI momentum stocks, as those are the most dangerous to short in this environment.  Stocks like MU, SNDK, WDC, STX don't trade like speculators have a full allocation yet. 

This market still trades narrow, and the winners continue to be concentrated in non-NVDA AI names that have gone parabolic since April.   You would have expected the market to broaden out with oil plunging.  But it continues to be the momentum stocks that continue to lead this market.  Unlike 2025, the Mag7 are not leading this market higher.  

The most popular and heavily owned names among retail investors are lagging.  That is definitely a sea change from how the market has behaved since the Covid lows in 2020.  Your average retail investor is loaded to the gills with Mag7 stocks.  There is no urgency to add more.  Big cap tech is spending all their free cash flow on AI capex, not stock buybacks.  Without buybacks and retail investors piling in, the Mag7 are a shadow of their former selves.  In the 2nd half of 2025, the Mag7 was outperforming the NDX.  But that outperformance has reversed to underperformance vs the NDX this year.  

 

In the meantime, we are seeing near record levels of dispersion in the stocks in the MSCI All Country World Index.  

April and May 2026 are right up there with December 1999 and February 2000 in dispersion rank.

 

This is what Robert Shiller described as the "gambler's excitement", where a few assets go up huge, attracting gamblers to the market.  It is those huge winners which drag the index higher, while the majority have mediocre returns.  

This gambler's mentality is causing more leveraged buying of equities, leading to surging funding rates.  

 

Funding spreads are spiking, something you saw around the local tops in January 2018, December 2024, and October 2025.  

Firsthand, I see the rampant speculation by market punters every day in premarket trading. Random low float stocks, many of them Chinese, have been going up 100s of percent in the premarket on no material news.  

Wall Street is feeding the ducks while they are quacking.  Huge equity supply projected for 2026 and 2027.   

We are seeing token costs drop significantly over the past 2 weeks.  There is news that the hyperscalers like META, AMZN, and MSFT are now emphasizing more efficient use of tokens, tightening token budgets, and no longer rewarding employees for mass, indiscriminate token usage.  Also hearing that more cost efficient LLMs, like DeepSeek are gaining popularity.  This has negative ramifications for the AI capex buildout, as the Anthropic and OpenAI LLMs are more hardware intensive.  

After this week's FOMC, the market has changed their view on Warsh.  They think he will try to talk tough on inflation at the beginning to gain some credibility from other Fed board members.  That could mean that the July FOMC meeting could be hawkish and lay the groundwork for a September rate hike.  Warsh also will issue less forward guidance, making the FOMC meetings that much more volatile.  There will be more uncertainty going into those meetings, so likely more selloffs going into FOMCs than in the past.  

The Iran deal bull catalyst has now been used up.  This is significant as there is no longer positive headline risk to deal with if you are short.  It also makes investors more complacent.  We are now in the post June opex seasonal window of weakness.  Post June opex and post September opex are the 2 most bearish times of the year for stocks.  A lot of put protection has expired.  If we do get a dip next week, will look to cover shorts to re-short in early to mid July.  

There will be a substantial partial lockup expiration of SPCX shares after their earnings report in early August, so that could be another catalyst for an August decline.  Base case is that we chop up and down between SPX 7300 to 7600 for the next 3-4 weeks.  I expect that chop period to end by mid July, and a potential waterfall decline towards SPX 7000 from mid July to mid August. 

Nonchalant Investors

Last Friday, after the NDX went down 5% in one day, I was expecting to hear some concern from those on CNBC, Bloomberg, and Twitter.  Instead, most were quite nonchalant about the big down day.  They were sticking with the same sales pitch, that I keep hearing over and over again.  They said earnings are growing faster than stock prices, that the AI capex boom has a long ways to go, economy is strong, etc.  You had a similar sized pullback from late February to mid March, and traders were much more fearful back then.  Its quite clear that investor positioning has gotten much more bullish vs. 3 months ago.  You did get a bit of fear on Wednesday as SPX broke below 7300, and US/Iran exchanged some strikes but it disappeared right after you got another Trump TACO on Thursday.  

They don't ring a bell at the top, but the amber flashing lights come on more often at tops.  One of those amber flashing lights is the Hindenburg Omen.  Its one of the few multivariable indicators that have a good hit rate in picking meaningful tops.  They come when there are lots of new 52 week highs and 52 week lows on the same trading day with negative market breadth, with the index above the 50 day moving average.  Since the start of the month, we've had a cluster of Hindenburg Omen signals.  

The Hindenburg Omens are popping up just as speculative breadth is getting weaker.  Speculative breadth is the breadth of speculative stocks and assets, such as AI related names, meme stocks, bitcoin, gold, etc.  Basically anything that is popular among the Reddit wallstreetbets crowd.  Speculative breadth has gotten much weaker this year compared to 2025.  In other words, only a few giant companies that are putting up big earnings growth due to the AI capex boom are rising.  Stocks like MU, SNDK, WDC, STX, INTC, AMD, DELL, etc.  One of the exceptions is NVDA, which continues to lag its AI hardware/seminconductor peers, despite strong earnings growth.  NVDA has reached investor saturation, so those looking to buy it already have a sufficient allocation in their portfolios.    

Looking at how much inflows the DRAM ETF is gathering, as well as levered single stock ETFs in stocks like MU, we are quickly getting to the saturation point for the recent parabolic names as well.  

Speculators are also looking for more juice when buying the popular momentum names.  Leveraged long and inverse single stock ETFs AUM are up huge over the past 2 months.  Most of the AUM is in leveraged long ETFs.  Inverse ETFs only make up a small portion of the single stock AUM.  

Once you get the speculators fully loaded up on the recent AI parabolic movers, its probably game over.  I can't imagine investors rotating back to previous speculative favorites like bitcoin, gold, and non-profitable speculative tech (space, quantum, etc.).  So many momentum investors have gotten burned chasing those assets and stocks, I just don't see them repeating the same mistake they made less than a year ago.  Plus, there is lots of technical resistance above.  There are plenty of bagholders in those assets/stocks looking to unload their bags on any strength.    

Trend following CTAs are now heavy net long SPX and MSCI EAFE futures, and near max short Treasury futures.  This is even after the big pullback over the past week.  DBMF, the biggest trend following ETF, is very long stock index futures.  I don't remember seeing them this long SPX futures, or this short Treasury note futures.  

The bigger their position, the more they have to unwind when the trend turns.    

With all the IPOs planned for this year, equity issuance will almost be equal to equity buybacks/purchases for 2026. That hasn't happened since 2021, when we had the big SPAC boom.  Add to that the huge increase in the free trading float as insiders at SPCX, Anthropic, and OpenAI sell stock after lockup expirations.  Just those 3 stocks will probably unlock over $3 trillion in equity supply from now till 2027.  

The put/call ratios went up after last Friday's selloff, so you have cooled down the call option punters looking for big fast gains.  But the greed runs deep in this market, and just today, the put/call ratios have fallen right back to levels where they were before the 5% drop in NDX.  The BTFD forever mindset remains as they come right back looking to catch a rally.  Only after they've been burned a few times betting on quick rebounds in SPX and NDX will they get more cautious, setting up a waterfall decline later in the year.  

We got Iran deal headlines again, and it appears we are finally closing in on a deal this time.  There have been several head fakes regarding a deal, but the crude oil market trades quite weak, and that is the biggest signal that those putting money on the line are betting that the Strait will be open reasonably soon.  Even if we do get a deal announced, they've squeezed so much juice out of a potential deal that I doubt you get a big up move after the deal is formalized.   

Inflation fears and optimism on the US economy are driving a big push to short SOFR.  SOFR futures are pricing in 37 bps of rate hikes by early 2027.  Speculators have the biggest net short in the history of the contract.  

SOFR 3 Month Speculator Net Position

 

The market is heavily leaning towards further weakness in bonds, and strength in stocks.  The US and Asian economies are dependent on the AI boom to continue to maintain growth.  Any disappointment in AI capex growth could fuel a massive short squeeze in bonds and a market going from pricing in Fed hikes to pricing in Fed cuts.  

Overall, it feels like we are now in the topping process.  During the topping process, I expect lots of volatility but in a range bound market.  We could see marginal new highs in SPX and NDX in early to mid July, perhaps 1-2% above the early June highs.  But I expect any new all time highs to be a false breakout setting up a big down move.  

Bottom line, lot of things are lining up now for the short side.  

1. Very bullish investor positioning.  

2. Unfavorable supply/demand picture starting from now, well into 2027.  

3.  Bearish price action underneath the surface with many speculative stocks trading weaker than the index.  Hindenburg Omens firing off. 

4. Extremely high valuations in a bull market that is now almost 4 years old, dependent on an AI capex boom, which has yet to generate profits for those building the AI infrastructure. 

Covered some shorts into the dips this week, still short some speculative tech.  Looking to put some of those shorts back on sometime next week, if SPX > 7480.  I expect a choppy month, so should get another playable dip later this month after this short term rally plays out.  

An Exclusive AI Party

As the bubble gets bigger, the speculators are getting more selective.  In 2020 and 2021, the government handed out trillions over 18 months and the result was a giant speculative frenzy.  During the Covid money spew, SPACs were popping up left and right. Most of them acquired sub-par companies and took a big slice off the top, making the after market SPAC buyers absorb all the risk.  Crypto was going crazy, as were meme stocks, bought up by a bunch of herd chasing, inexperienced investors.  Stocks like GME and AMC were the poster childs of that bubble.  You had NFTs, which was a ridiculous idea in the first place, going for huge amounts.  Sports cards, Pokemon cards were rocketing higher.  That all ended with the bond market rout in 2022.  

Then the second wave of this never-ending party started in late 2023 with AI.  The first beneficiaries of this AI boom was NVDA and the AI power infrastructure plays.  They were the most popular names in 2024 and 2025.  This speculative wave got another boost that was non AI related, with Trump getting elected in 2024.  Crypto pumpers and HODLers got all excited about a "crypto-friendly" government.  I remember hearing about how stablecoins were going to go mainstream, about how crypto would allow financial transactions to be tokenized, and you could buy parts of a share of stock, or even a house.  The idea is ridiculous, but speculators were coming up with any reason to buy bitcoin, ether, and alt coins.  The real reason they were buying was because it was going up, and they thought it would keep going up.  

In 2025, everyone was participating in the party.  At the core, you had AI and AI adjacent names going straight up.  Crypto was going crazy with bitcoin/ether treasury companies raising tens of billions to buy more coins.  People were talking about alt season.  Super speculative concept stocks in quantum, space, nuclear were up over 1000% in less than a year.  The speculation was broad, indiscriminate, and caused massive short squeezes.  That phase of the bubble topped out last October.  

2026 is a different animal.  On the surface, it may seem like more of the same animal spirits as 2025, with the Nasdaq going up almost every day, and SPX following along.  But this market is much narrower than 2025.  The breadth of the speculation is poor.  Not as many names are working as 2025.  This year, its been the previously overlooked AI beneficiaries that have squeezed higher.  MU, SNDK, WDC, STX, AMD, INTC, MRVL, DELL, etc.  Semiconductors of all types are working.  AI hardware companies are experiencing huge earnings growth as the AI capex boom continues.  But beyond that, buyers are notably absent.  

The 2 main barometers of US stock speculation are NVDA and TSLA.  When they are stronger than the market, you know that your average stock investor is doing well.  While NVDA is holding up relatively well, its not rallying with the other AI semiconductor names.  It trades heavy.  NVDA had great earnings a couple of weeks ago and it sold off on the news.  Having a market cap of over $5 trillion will do that to a stock.  

TSLA trades worse than NVDA.  It is another household name that is one of the staple tech holdings among your average retail investor.  It is lagging the Nasdaq badly since it topped out last October.   NVDA has been popular among retail since late 2023.  TSLA since early 2020.  Investors already have enough NVDA and TSLA in their portfolios.  There are not enough new investors coming into the market to bid up those names.  

TSLA vs NDX

Same can be said for bitcoin as for TSLA, but to a bigger degree.  Unlike TSLA, which gets a steady stream of buying from inflows into index funds, bitcoin is a pure spec play with no passive flows coming in.  In fact, bitcoin miners are a steady source of bitcoin supply.  In 2025, you had bitcoin/ether treasury companies, in particular MSTR and BMNR, buying up huge amounts of bitcoin and ether.  That caused a big rise, but it wasn't sustainable.  The bitcoin flash crash on October 10, 2025 changed everything.  That was the event that marked the end of the post Trump election crypto boom.  After that, the bitcoin treasury company stock prices got crushed, making it difficult for them to continue to sell stock to buy more coins.  That iterative cycle of investors buying the crypto treasury company stocks forcing more buying of the coins was frozen in its tracks.  At that point, investors who were going to buy crypto had already bought it.  There were not enough new buyers (suckers) coming in to prop up an inflated meme asset.  Bitcoin got crushed this week and went down to below $60K, as the Nasdaq is still up big on the year.  That is what happens to a pumped up meme asset that is saturated with bagholders.  

As the market is getting more selective on what tech stocks to pump, you had GOOG coming to the market this week to raise $80B of new equity.  And on Friday, you had META headlines about looking to issue tens of billions of equity.   

With all this liquidity getting drained from equity markets with equity issuance, you have a very different supply/demand situation than 2025.  From a supply/demand perspective, this AI bubble is rapidly catching up with the dotcom bubble.

This is the backdrop as the SPCX IPO is coming up next Friday.  They are looking to raise $80B.  Preliminary IPO pricing is at $135/share and a $1.8T market cap.  If that's the IPO price, there will be very little upside for retail investors buying into this IPO, as the valuation is already so high that its hard to imagine buyers bidding it up to even a bigger nosebleed valuation.  $1.8T is a bigger market cap than META.  The earnings gap between those 2 companies is enormous, and SPCX is the more expensive stock!  SPCX is a space play, which doesn't have the earnings growth that speculators want these days.  It is a pie in the sky stock with xAI attached to it, which is a glorified AI data center company.  xAI has essentially given up on competing vs. Anthropic and openAI, and are just renting out their data center capacity to others who have the demand to actually use it.  

The put/call ratios continue to show lots of call buying.  In particular, retail traders are buying a lot of calls.  

The COT data as of June 2 showed both asset managers and leveraged funds reducing their SPX net exposure.  After the index went up.  That is uncommon behavior.  

 

Dealers took the other side of that trade, as they significantly reduced a big short position. They are going back to similar sized positions of the 2nd half of 2025.  It just so happens that the put/call ratios are also down to 2nd half of 2025 levels.  Dealers are likely delta hedging their short call positions by buying futures.  This matches the recent retail options activity data which shows lots of bullish call positioning.  

SPX Dealer Net Position

This week, you finally saw the AI hardware/semiconductor trade get taken for a big hit.  

On Friday, MU, SNDK, AMD, INTC, DELL, INTC as a group were down more than double that of the NDX.  For the more speculative AI stocks, they were down more than triple that of the NDX, which was down more than 5%.  Realized volatility went through the roof this week, but VIX only closed at 21.5.  This as oil was going down, which is a new development.  Up until Friday, the SPX/NDX and bonds have almost always rallied with oil going down.  That changed in a big way on Friday.  A change of character for this market.  

With the strong NFP number, bonds took a hit, and that seemed to be the initial trigger for the selloff.  But SPX kept going down, an extra 150 points even after yields had stopped going up.  Short term Treasuries look interesting after the selloff on the stronger NFP.  I will consider slowly building up a long position as I believe the Fed will be late to try to hike even with high inflation.  As the Fed delays rate hikes, the AI trade likely goes bust in the meantime.  And there is no way the Fed hikes as the stock market is imploding, I don't expect Warsh to do what Powell did in late 2018 by hiking into a falling stock market.  

I added to non-AI tech shorts earlier this week, and remain short.  I am still not interested in index shorts, as I would rather target the weaker speculative tech stocks than the whole market.  I will look to hedge some of those shorts with AI longs next week.  While Friday's selloff looks scary, they usually don't last for long, especially after such a strong uptrend.  I expect the uptrend to flatten out and get choppy with big down days followed by big up days.  That could last for a few weeks before you get the big move lower.  By July, it will probably be time to go short NDX along with speculative non-AI tech.  Until then, I will look to put on more non-AI tech shorts after rallies, and hedge with AI longs on dips. 

Pokemon Bubble Redux

The 2000 parallels are eerie.  

In 1999-2000, there was a mania in Pokemon cards that crashed in 2001.  It just so happens that Pokemon cards are in a even bigger bubble than 2000.  Ever since the Covid stimmies, Pokemon card prices have been rising, but lately, its been going parabolic.  Wages aren't keeping up with inflation.  White collar jobs are hard to get.  Those looking to make money without getting their hands dirty are turning to speculation.  Its a continuation of the Covid mania where asset prices went crazy, taking a break in 2022-2023, and restarting in 2024.  Meme stocks aren't as popular as they were in 2025, but Pokemon cards are as hot as ever in 2026.  

Even CNBC has noticed, as it made a recent video on Pokemon cards.  

Now, according to Pokemon card collectors, the mania has reached the crazy stage.

 

We also have the Beanie Babies bubble in the late 1990s, which popped in 1999.  You can draw a parallel to the Labubu doll bubble which burst in late 2025.  

 

Many may think that its nonsense to draw parallels between Pokemon cards and Labubu dolls with the US stock market.  But 2000 and 2021 has shown that when the stock market is in a bubble, the speculative animal spirits spill over into collectibles.  These ancillary bubbles tend to start after the stock market has already been going up for quite some time.  Unlike the stock market, these bubbles in collectibles like Pokemon cards are much briefer in duration.  And they tend to top out a bit sooner than stocks.  So they are a helpful timing tool in determining what stage of the bull market we're in at the moment.  Looking at what's happening in the Pokemon card market right now, it just reinforces the 2000 bubble replay thesis.  When the prices of Pokemon cards top out and star dropping, that could be a warning sign for the bigger casino, the stock market.

The money keeps pouring into AI.  On Thursday, Anthropic raised $65B in equity, and are looking to add $36B in debt to finance more AI infrastructure.  

 

It looks like the VCs are already spending the money that they plan to get from the Space X IPO sale.  They are all in on AI.  AI is the only source of growth in the global economy.  Without AI spending, the global economy is stagnant with very low growth.  That is why the popping of the AI bubble will have a much bigger economic impact than most expect, and will trigger a bigger bear market than 2022, IMO.  

But corporations are starting to rein in their AI usage, looking at their AI costs skyrocketing.  

This week, you saw a lot of the space related stocks go up, in anticipation of the SpaceX IPO.  Speculators are overestimating the retail demand for the SPCX IPO.  Investors are looking for pure play AI plays, not meme stocks with low value add AI exposure.  If it wasn't for the Elon Musk name value, SPCX would be valued way below $1T.  And since 2021, TSLA has been lagging the Nasdaq.  Eventually, you have to deliver earnings and free cash flow if you want to maintain a giant market cap.  This market has been so frothy for so long, its given Musk and TSLA a lot of rope to mess around, and not really affect the stock.  Now with SPCX adding a lot more Musk equity into the hands of retail, its going to be interesting to see how the supply/demand dynamics play out as SPCX insiders dump to the public.  

The meme stock/asset era peaked last year with bitcoin FOMO.  Now speculators are laser focused on AI, in particular hardware and semiconductors that weren't pumped in 2024, with a few AI beneficiary software companies thrown in.  That is why MU, SNDK, WDC, STX are surging while NVDA is stuck in the mud.  Investors are already heavily invested in NVDA.  They've only recently started accumulating MU, SNDK, etc, and haven't gotten nearly as big an allocation.  

With an estimated market cap at IPO of $1.75T-$2T for SPCX, it would already be valued greater than META.  It will take an enormous amount of buying from institutions to absorb all the insider selling that will happen once the lockup expirations occur, as they are spread out over 6 months, with a big chunk being freed after Q2 earnings (early August).  I would not be surprised to see retail bear the brunt of the selling, as they will be allocated a huge chunk (about 30%) of the IPO.  

Looking at positioning, according to GS Prime Broker data, L/S equity hedge fund net leverage is at 59%, the highest since the 2021 peak.  Net leverage has been rising fast since the March 30 bottom.  Mutual fund cash as % of assets is also historically low.

Retail continues to pile into this market.  

On the Iran War front, all week, you've had hints of a deal being made, then headlines walking that back, rinse and repeat.  On Thursday, it finally seemed like the market believes that a deal is imminent, just waiting for Trump's approval.  I would say a deal is basically priced in, and I would expect any news that a deal is done would only cause a small pop higher that would quickly be faded.  That positive catalyst is now all used up. 

We are finally seeing palpable signs that investors are feeling bullish, and positioned that way.  The daily put/call ratios are low.  You don't hear about private credit anymore.  You don't hear about AI killing software companies anymore.  Now you hear a lot of talk about how tech earnings are growing, justifying the move higher in the NDX and SPX.  This is exactly the kind of complacency that provides good risk/reward short opportunities.  

I am sure some of the paper napkin chartists will compare 2026 with 2025 and note that the market bottomed in early April in 2025, and kept going higher till October with barely any pullbacks.  Big difference was that you didn't see positioning go down as much during this March 2026 correction vs. the March-April 2025 correction.  So you don't have as many investors looking to buy back to get to a full equity allocation.  And you also didn't have a gigantic $76B IPO with an accelerated lockup expiry waiting to flood the market.    

The breadth of the speculation is more limited than last year.  Last year, you had bitcoin/ether making huge up moves, quantum, nuclear, AI data center, AI energy plays, AI software (APP, PLTR) all squeezing higher.  This year, the speculative focus is narrower.  Its direct beneficiaries of the AI capex spending such as semiconductors and hardware.  Throughout May, bitcoin has been struggling to go higher even as the Nasdaq has been going up almost every day.  I don't expect any meaningful rotation to the speculative favorites of 2025, as there is a ton of overhead resistance, with too many bagholders looking to exit gracefully.  It may look even more wild now than 2025 with the big moves in stocks like MU. But the speculation is going towards companies that are getting a huge earnings boost from all the AI spending, not just any meme that happens to be the flavor of the month like in 2025 (crypto alt season, ether mania, nuclear, quantum, etc.)

Big rally in bonds over the past week.  I was considering a long position if there was still Iran deal pessimism, but it played out the other way.  If you get another selloff in the bond market in June that takes it back towards the May lows, I will look to buy short term Treasuries.  I expect a 3-4% pullback in SPX in June, then probably another breakout to a marginal higher high in July, which could be the time to go for index shorts.  

I put on a starter short position in some speculative tech names on Friday.  I plan on adding to shorts throughout June.  We are close enough to the top that it looks better to short the higher beta tech sector than the overall market.  However, I will avoid shorting the most recent parabolic moves in stocks like MU, SNDK, etc. preferring to short stocks that are not direct beneficiaries of the AI capex boom.