Volatility is Flatlining

Over the past week, we got all kinds of news and events, but the market has gone nowhere.  The AI news about slowing the pace of AI investment due to safety concerns, which seemed like a nothingburger, but people made a big deal out of it.  And it ended up being much about nothing.  Ever since that news came out last weekend, NDX has been outperforming SPX.  

The talk of the town this week was the Fed rate hike.  The market didn't seem too concerned, as it didn't really selloff much going into the FOMC meeting, and then you got a stop run after Warsh came out more hawkish than expected.  It seemed like an overreaction to the news.  What were people expecting?  A one and done announcement, a dovish hike?  The Fed never does that.   Especially non-forward guidance Warsh.  A dovish hike is like jumbo shrimp.  Fed reactions are usually overreactions and often reversed in the following day.  That's exactly what happened on Thursday.   

Those looking for a quick strike in this market have been disappointed.  It is quite unusual to see such a dull market when the market is making lower highs and lower lows.  Usually investors get worried when stocks are making lower highs and lower lows.  But the down moves have been small, so investors have shrugged off the weakness, expecting a rally to come to save the day.  And that's what's been happening.  Investors don't want to sell weakness, and they are not heavily hedged with puts, meaning gamma selloffs are unlikely.  So you get weak selloffs, which often reverse.  

One of the best indicators of retail investor flows is the DIX index, which shows dark pool short selling by market makers, which take the other side of retail orders.  They are going higher as the market is going sideways.  A high DIX is not necessarily a bearish signal, as DIX was high for much of the April to October 2025 rally in SPX.  But they are a warning signal of potentially more selling when retail is buying the dip and the market is trading sideways, like now.  Previous similar instances were early March 2026 and mid July 2026.

Last week, for the week ending September 11, BofA flows showed a big outflow from all investor types.  But this just partially negates the big inflows over the previous 4 weeks, as the 4 week flow is still positive.  

A note on the retail flows from BofA and implied by DIX.  BofA private clients are wealthy retail investors, while DIX is a representation of the payment for order flow retail brokers like Robin Hood, Schwab, Fidelity, E-Trade, etc.  

According to GS prime broker data, from Aug. 28 to Sep. 10, hedge funds were heavy buyers of tech stocks.  

Leopold apparently continues to splash around in the options market, buying OTM calls on high beta AI plays.  These options expire within 2 weeks, so he's basically buying expensive short term lottery tickets.  The guy just can't sit still.  Money continues to burn a hole in his pocket.  A degenerate gambler running a $10B hedge fund.  

You got a couple of investor sentiment surveys out this week, showing many more voting bearish.  In particular, the AAII investor survey showed a big jump in bears.  Yes, if you get these bearish numbers for a few weeks, its a good bottom indicator, but its just been one week, which often means more weakness ahead (see February/March 2025, mid March 2026).  

We've seen heavy inflows into TLT, as long bonds have been selling off for months.  Usually this happens in an uptrend, not a downtrend.  Being bullish bonds may seem contrarian, but the crowd has been buying, not selling.

SPX COT data as of 9/15 shows small speculators with a big reduction in net longs, ahead of the FOMC meeting.  

Short term, the market is lulling investors into complacency.  But we are seeing early signs that investors are starting to get worn out, as you have seen put/call ratios slowly rising, more bearish responses to investor surveys, and some investor outflows.  Now that September opex is behind us, we are in a very bearish 10 day period where investors have on less put protection and the stock buyback window closes.   One thing to keep an eye on is NDX outperformance over SPX since last Friday.  If it continues, it would make me less bearish and I will reduce my shorts.

I covered the rest of my shorts before the FOMC meeting, earlier in the week, and have put back on shorts on Thursday and Friday.  I want to be short going into the most bearish 10 day window of the year, as investors are starting to get worn out and more likely to lighten up on stocks.  

Drip Drip

This is the worst type of selloff.  For bulls, it is one paper cut after another, followed by quick relief, and then another series of paper cuts.  For bears, it is agonizingly slow progress to the downside when all the cross asset moves should have caused a bigger move lower.  The relentless march of time decay erodes the time value of puts, as there is no explosive upside for put buyers as the market drips lower, keeping IV contained.  

When investors are not willing to pay up for put protection, dealer option books are short fewer puts than normal.  This reduces the index delta hedging / IV rise feedback loop (gamma/vanna effect) when the market goes down, making selloffs look meager and lifeless.  But this type of market condition also breeds a dangerous situation where investors have less put protection while also holding large net long positions.  Investors are skydiving without a backup parachute. 

SPX put deltas for customers are showing very light customer hedging.  Very different picture than September 2025.

 

Apparently Leopold of Situational Awareness is back in a big way, and was single handedly responsible for the NDX outperformance last Friday and on Tuesday.  Talk about a price insensitive buyer who has no patience.  Now that he's back in the water, NDX has been underperforming SPX since Thursday.  Probably going to see another Leopold moment, but this time will probably be quieter with call options bleeding premium and deltas going into October opex.  Makes me more bearish NDX than SPX for the next few weeks, especially considering the Anthropic IPO coming up.

 

Schwab STAX index dropped from 59.8 to 57.5 in August.  Still at a very high level vs. average of last 3 years.

 Biggest net buys and sells among Schwab clients in August:

I believe that stock market bubbles don't burst because of external causes, such as higher bond yields/higher energy prices.  Bubbles burst because perception gets way ahead of reality, too many pile in looking for speculative gains with minimal support from long term fundamentals.  The bursting of the AI bubble will result in the vol event, not a vol event causing a bursting of the AI bubble.  

The price action this week just reinforced my belief that in order to get a sustained selloff, it has to come from AI.  Oil prices surged higher, along with bond yields, as the war drums got louder, but the market could still not selloff more than a fraction of a percent each day.  It was a drip drip slow bleed that was profitable, but frustrating to watch.  In the back of my mind, I know that there is a relief bounce just around the corner.  Oil prices don't grow to the sky.  Same for bond yields.  You could get another BS peace deal rumor.  It doesn't matter.  These kind of moves in oil and bonds are not sustainable.  As soon as they reverse, or even pause, shorts are going to be squeezed.

To get a real correction, you need to see AI stocks continue to lag the overall market, and after that, the AI capex boom to end.  The use of AI (at much lower prices) will probably still rise despite the AI capex boom ending.  It just means that the demand for semiconductors and AI infrastructure drops, which can easily happen while there is still growth in AI usage.  That is what happened in 2001 when the dotcom bubble burst.  I expect a repeat of that in 2027.  

Percentage of stocks in a bear market is very high with the SPX within 3% of an all time high.  

Fed flow of funds data out today.  Shows a big increase in household percentage of financial assets held in equities as of June 30, 2026.  Blasting to a new all time high.  They are all in on stocks.

The inflows into equities continues, according to BofA.  6th largest weekly flow in BofA history (since 2008). This time, its hedgies and institutions leading the buying, while retail sells.  

 

The latest COT data show mixed flows.  Asset managers have been reducing longs in SPX futures, while small speculators have been adding to their net longs in both SPX and NDX futures.  No clear signal from the speculator positioning there.

Asset managers have continued to trim their large net long position in SPX futures, now down to levels last seen in March.  

SPX Asset Manager Net Long Position

 Dark pool index (DIX), which shows dark pool market maker short activity (retail buy activity) shows retail buying the dip this week.  Retail is showing very little fear.

Put/call ratios remain near the middle of its range.  Investors don't seem to show any fear on selloffs, and no real excitement on rallies.  

Investors are buying the dip in Treasuries, TLT chart with 3 month rolling flows:

After the CPI, the odds of a FOMC rate cut next week shown by the CME Fed funds futures went up to 90%.  The bond market cares, as it continues to selloff, but the stock market has been very resilient despite the march higher in bond yields.  With the STIRs market already pricing in a hike, its likely that you could see a relief rally after the Fed hikes 25 bps.  Add the rise in crude oil prices this week and the stock market "should" be down more.  But you have to trade the market that you see, not the market that you want to see.  

The big picture remains the same, a very late stage bull market where the uptrend is starting to flatten out, but with investors remaining very complacent.  Investor inflows into equities continue to keep the stock market afloat, even as the enthusiasm wanes, the TINA (there is no alternative) belief remains even though bond yields keep going up.  

Have been very patient with the short positions but there just hasn't been enough progress and my short term conviction has gone down.  Less short term conviction = smaller position.  I trimmed my short position after the CPI came out, seeing bad price action after a slightly hotter than expected number.  It feels like the market will just trade sideways until FOMC, and then  have a relief rally/squeeze into triple witching opex.  I will look to re-add shorts if SPX goes back towards 7750-7780.

Earnings Distortion

There has been a lot of talk recently about bond yields and whether the Fed will hike or not.  The market has been jumping on Jackson Hole and nonfarm payrolls.  Those are a distraction from the main driver.  Whether the Fed hikes in September is basically meaningless, regardless of what you hear from pundits who will overreact to a single FOMC meeting.  When volatility is low and the markets are trading in a narrow range, investors will often take their eye off the ball.  The ball is AI.  

The AI boom has created a surge in earnings that has been the justification for moves higher in SPX and NDX for 2026.  Among the hyperscalers, the majority of the earnings growth is coming from non-operating one-time gains from the revaluation of private equity AI holdings.  The private equity valuation increases for Anthropic and OpenAI are the main reason for the increased earnings.  These type of earnings are not sustainable.  Two companies whose main source of revenues are from LLMs which are quickly becoming commoditized, thanks to China.  Two money losing companies that have huge upcoming debt payments when you get a wave of AI data center completions starting next year.

For the semiconductors, the earnings are real and come from operations, but depend on continued spend from Anthropic and OpenAI which fuels hyperscaler and neocloud capex demand.  Hyperscalers/neoclouds are going increasingly into debt to fund their AI data center buildout so the Street has been flooded with AI data center paper.  A look at ORCL 5 year CDS (around 200 bps) shows you the consequence of going all in on AI data centers.  Before ORCL's spending spree, their 5 year CDS was around 35 bps.

The earnings growth story is dependent on this AI capex boom continuing.  It all flows back to Anthropic and OpenAI continuing to spend beyond their means through equity and debt offerings, or getting enough profits to pay for it without financing.  But a recent Substack post from Groundbreaker shows the similarities between AI financing and Subprime mortgages in the mid 2000s.  

Instead of a mortgage rate reset wall, you have a compute commencement wall from AI data center completions.  When the data centers are completed, that's when OpenAI and Anthropic are obligated to start paying for the compute they ordered 2 to 3 years before.  The bulk of the completions happen in 2027 and 2028.  That is when they really need to start paying for their compute with profits or go back to the Street for more financing.  Eventually, the Street will balk if they remain unprofitable.  If the AI labs can't get more financing or become profitable enough to fund themselves, they will default on their debt obligations to the hyperscalers, and could force NVDA to backstop them, leading to big losses.  

Already seeing reduced demand for AI data center debt.  From a recent FT article:

So far, with the rapid revenue growth from selling their services below cost, the AI labs have been able to get all the financing they need.  This is the honeymoon period where the hyperscalers book huge order backlogs with each signing with AI labs, raising their revenue outlook.  But the hard part comes later, when the AI labs' payment obligations really ramp up with data center completions in 2027 and 2028.  The AI labs need to profitably sell their compute at increasing scale or borrow more/sell more equity.  If the AI labs aren't sufficiently profitable by next year and the financing window closes, then defaults start happening, hyperscalers have to use non AI cashflow to pay for their AI capex debt, and the shit hits the fan.  

Few are bracing for this.  It all depends on demand for AI compute at high prices that can make Anthropic and OpenAI profitable enough to meet their growing debt obligations.  In order to improve their poor operating margins, the AI labs will have to use less compute for training, meaning their lead over Chinese AI labs will shrink even more, or completely disappear, or they will have to raise prices, reducing demand for compute.  In either case, there will be less demand for compute at unsubsidized pricing.  Less compute demand is bad news for semiconductors, the most important sector in the stock market.  Only if the demand for AI compute continues to grow even at unsubsidized pricing will you see the AI boom continue.  Otherwise, AI investors will be heading full speed towards the compute commencement wall in 2027/2028 without airbags.

If the AI labs remain unprofitable and keep burning cash, as I expect in 2027, then they will need more financing rounds to pay when the compute commencement wall hits.  I have a feeling that those financing rounds will be much tougher to get than the ones this year.  I doubt there will be much appetite to fund money losers when they have so many future debt obligations coming down the pike.  2028 will be even worse for AI data center completions and debt payments.  If the AI labs can't get the financing, then they will burn through cash and have to default on their loans.  

Those loan payments are going towards hyperscalers and neoclouds.  Then the hyperscalers will have to write down huge losses on their books, along with the steady depreciation of their AI assets.  They will have to repay AI infra debt from non-AI data center cash flow.  This will impair their ability to do stock buybacks and be a huge hit to their earnings.  The worst of the storm will hit the semiconductors, as their order books will dry up, along with their profit margins which will plummet.  Then people will understand why semiconductors are considered cyclicals.  

You will need at least 150% revenue growth in 2027 for openAI and Anthropic to be able to meet their debt obligations without additional financing.  From a demand perspective, I am highly skeptical that AI compute at unsubsidized pricing grows 150%+ vs. 2026 levels.   If AI was so great, why aren't we seeing call centers all replaced with AI automated chat bots?  If you ever called a customer service phone number and gotten an AI voice bot trying to solve your problem, you know how annoying and unhelpful it is.  You just end up having to waste time going through various chatbot menus instead of speaking to a call center worker directly.  You have to wonder why there haven't been more job cuts at the tech companies which are the heaviest users of AI agents.  If these AI agents were so productive, why not do more drastic job cuts?  Many are ignoring the limitations of AI use cases as they project huge growth rates for years on end.

Instead of scaling laws leading to exponentially increased capabilities, its been more like a logarithmic increase in capabilities that show less progress for each new model vs previous ones.  ChatGPT is a clear example of this.  Their first models hallucinated a lot.  The next iteration vastly improved these shortcomings, but with each new model, the low hanging fruit having already been picked, you get less incremental improvements.  You are already seeing pricing pressure at the AI labs, as demand for the latest expensive models has been subpar.  There is very limited demand for frontier models when previous models get most of the AI capable jobs done for a fraction of the cost.

BofA clients have been aggressively buying tech for the past 4 weeks.

 BofA clients bought heavily last week.  

Back to the markets.  You are seeing a lot of discussions about bond yields these days, and the kneejerk reaction would be to say that we are nearing a top in bond yields because people are worried about them.  If investors were really worried about bond yields, the MOVE index or VIX wouldn't be so low.  You have been getting very little volatility in both stocks and bonds, and thus investors are complacent.  There hasn't been a 1%+ down day in SPX in over a month.  Investors tend to extrapolate recent history into the future.  Hedging demand for SPX is way down from earlier in the year.  Investors are increasingly unhedged going into 2 big catalysts:  Anthropic IPO and midterm elections.  Gamma fueled selloffs may seem scary, but they don't last for long.  The scarier selloffs come from real money selling, not dealer delta hedging.  With customers light on put protection, they are more apt to panic sell in a correction.  

SPX avg stock call skew is high, avg stock put skew is low.  Investors are leaning heavily towards calls. 

DBMF, the biggest trend following ETF, substantially increased their long SPX and MSCI index futures positions to a total of 63% of AUM.  They now have one of the biggest net long equity positions for the history of the ETF.  

The imbalances in the market are huge.  So many are leaning towards the long side here, aggressively long, just when the risks are about to be the greatest.  Maybe I am off my rocker, but I think the next 2000 points in SPX is down, not up.  It is a long term view looking out into 2028, so I try not to let those views affect my shorter term trades.  

That being said, the probabilities on the short side are lower than trading on the long side, but the payoffs are more explosive and come in waves, all at once.  Shorting the index is a low probability/large payoff structure.  I see a big wave coming before the midterm elections.  It could start next week, it could start in late September.  If we are still above current levels by the last week of September, then my thesis is incorrect and I will exit my position.  In the meantime, I will hold a full position short, with initial target of SPX 7400, and depending on price action and how the crowd reacts, may look for a move down to 7000.  Looking to hold the trade at least till the last week of September (may close out some earlier to put out again on a short term bounce), and will look to close it all out at the latest in the last week of October.