Title: The Anatomy of a Crash — When Rolling Blowouts Become Risk Off Show: Paulo Macro (Substack) — paid Guest: Paulo Macro ("Cloudbear") Date: 2026-MAR-14 URL: https://paulomacro.substack.com/p/the-anatomy-of-a-crash Length: written post (no timestamps) Note: Back-filled post (processed 2026-JUL-07). Paulo moves from "general bearishness" to "acutely concerned for a crash" as a Rolling VAR Shock converges on every hedge-fund strategy at once. He walks the market topology: positioning barely cut (private wealth/long-only at record-low cash, HFs only just degrossing), the Dispersion trade starting to buckle as single-stock vol gaps and correlation rips (his own AirBNB put order got re-racked above the mid, illiquidity worse than the $3mn top-of-book S&P tape), the Basis Trade leaking (swap spreads deeply negative), credit vol now LEADING Treasury vol (both a credit and a rate issue), and no marginal buyer left (retail going quiet, sovereign wealth + Asian surplus economies potential sellers into Japan year-end). Then he lays out "The Anatomy of a Crash": initial correction + relief bounce → retest with a minor breach/intraday tail → a 1-2 day final "all clear" bounce → collapse (1987, 1929, 2021 bitcoin, 1997 Hang Seng, now the Nasdaq 100). He bought VIX calls for the first time since April 2010 and shorted AirBNB via June $100 puts as a single-stock-vol expression. Body reproduced for personal study; Substack chrome removed, wording otherwise verbatim.
In Release the Crackups and the Era of Rolling Blowouts among other notes, I have tried to explain how seemingly unrelated ruptures in specific corners of the market run their course. The framework originally came to me in the summer of 2018 after the following sequence of events:
January 2018: bitcoin blew up.
February 2018: Volmageddon — short vol strategies and risk parity got torched.
March 2018: FAANGs, Facebook data scandal, Libor ripped.
April 2018: UST 10yr yield ripped through 3% for the first time in over 4 years.
May 2018: Turkey, Argentina, Brazil imploded.
June 2018: Italy BTPs and Spanish yields exploded.
My conclusion at the time was that even if the S&P would make new all-time highs (it did), a dramatic Risk Off lay ahead, because Rolling Blowouts have only ever ended in one way — Risk Off.
Today's fragility combined with the Rolling VAR Shock of recent weeks has transitioned me from "general bearishness since late last summer" to "acutely concerned for risk." However I had a realization in recent days that has notably affected positioning in my book, and to explain why, I want to paint a pretty comprehensive picture…so bear with me.
Setting the Table
Positioning
Credit and Collateral
Who is the Buyer?
Progression of Sentiment Anecdotes
The Anatomy of a Crash
Setting the Table
First, the a high-level backdrop as a reminder:
Back in late December, my friend Kevin Muir invited me on his holiday year-end Market Huddle special for a quick 10-minute year-end review and outlook. My contribution begins at 52:40 (link here). I gave Kevin a chuckle — and may have even scared him even though he has been bearish for a while now — with my take on markets as a Turducken of Market Risks:
Take 1973's Nifty Fifty concentration levels to an even greater extreme of 40% of the S&P in only 7 names [and yes, now an oil crisis, how 1973 of us];
Wrap that in 2000's US Exceptionalism of Growth + Tech/Dotcom/Retail mania;
Shove that into 2007's opaque credit structures and circular shadow vendor financing for datacenters in the hundreds of billions, alongside a decade-long bubble in mismarked/unmarked private credit misallocation (it can't be a true Turducken without hidden financial leverage);
Add the stuffing of 0DTE intraday leveraged instability a-la 1987 portfolio insurance.
Many thought the analogy funny, but I was being serious. Just one of these initial conditions would be sufficient for a pretty memorable bear market when equity valuations are trading at the highest levels in US equity market history. To have a superfecta like this is — historic? And would also suggest you don't wipe the table down with a run-of-the-mill bear.
With that table set, the problem today is the convergence of vulnerabilities in market topology among participants right here that has moved me far beyond my cave of general bearish skepticism since last August-September (when the Cloudbear Zyn Indicator began to rally hard) to now outright concerned for a significant market crash in the short term.
Positioning
By now we know the story. For all "the bear talk is everywhere," what I sense is actually a growing nervousness very recently (past 2-3 days, I will come back on this in the Anecdotes section farther down) along with and a lack of appreciation for just how little fuel there is for a sustainable bid in the market in relation to a potential offer.
A quick reminder just how much risk has really been "cut" here so far:
Private Wealth has barely cut according to yesterday's BofA Flow Show:
Hedge funds are just starting to lower gross exposure while aggressively cutting net exposure by reaching for hedges. I love the context of "largest single weekly gross decrease in more than four months" below from Goldman…and yet this dropped gross exposure to the 86th percentile of just the past year, while still sitting at the 97th percentile on a 5-year lookback:
Source: GS
So L/S strategies are just now starting to degross, but over the past few weeks have been netting down initially by layering in lazy macro hedges because they don't want to sell their core holdings and risk missing a TACO upturn that everyone keeps banking on in Pavlovian fashion. The idea is hedges like index futures and ETFs can be quickly cut, but single stocks take time to buy and sell. This point was confirmed yesterday by GS Head of Trading John Flood (I have always considered John a permabull cheerleader, so the fact that he is not his usual megabullish is another confirmation of the nervousness of the past two days — I'll get to this later).
For now, the resistance to degrossing while buying hedges has perversely kept the market from releasing lower because implied volatility is elevated due to hedging demand but hedges are not really helping while participants are still working with a Rotation mindset (keep cash low, just find me a place to hide for now while staying invested). You can see this in these comments from Goldman:
So guys are not in "get me out" mindset — they're in rotate mindset while the tone of the market has clearly changed. What's holding it up?
For this, I come to probably my single biggest concern right now. What's the one strategy that has not yet felt pain amidst the blowouts across hedge funds?
Dispersion.
For those unfamiliar, please read this primer. The short version is this strategy takes advantage of harvesting premium by typically buying single stock volatility (which pays off as individual stocks move a lot day-to-day, but independently of each other when correlation has been low), while selling inflated index volatility which has been realizing far lower than implied since the index has not really moved around much. You can see how the "short" leg of the Dispersion trade is paying off quite well as implied index vol has risen yet the index itself is not moving a lot:
What about the "long" leg? Implied vols on single stocks have been inflating a lot, but they are still realizing (moving around) a lot, meaning there has been alpha to harvest. But there is a growing stress here via rising correlation. On Thursday in Making Macro Great Again in 2026 The Year of the Reckoning, I flagged in the section Volatility and Hedges vs Degrossing (please reread if you missed it) that
Macro hedges were indeed somewhat unwound on the weak Taco Tuesday;
I mentioned correlation vs. volatility. Here's what I wrote, with the charts updated as of Friday night:
Here's 1m-3m implied correlation... 1m is getting jacked, and the spread is up at Liberation Day and March23 SIVB crisis levels [actually as of Friday the difference between 1m and 3m correlation exploded through to multi-year highs!!! See bottom pane]:
You would look at that and say "oh clearly a bottom, buy stocks." But then look at 1m-3m VIX … we haven't seen The Fear:
Correlation is screaming of a growing problem under the surface, and rising correlation is always going to be the signal that the low volatility index regime could flip.
Then I had a personal trading experience that brought it home. I decided on Thursday I needed more downside, but index vol is tough to buy in the 20s (especially out of the money which is higher). So I looked around for single stock situations. Some of you will recall this stock has occasionally landed in my book on the short side… AirBNB. I figured:
It's consumer/tourism related, which will get destroyed in a global recession from an oil crisis;
It's a crappy customer experience with tons of hidden fees;
It's tech adjacent;
The chart looks horrendous.
So I went to price some $100 June puts. I start pricing… the historic vol is in the upper 30s, and a 40 implied vol seems reasonable, let's try it. Knowing the options are illiquid and quoted wide like most single stocks, what I do now is use a retail account like a Schwab which I know will initially route the order to an internalizer like Citadel that pays for my flow, and I'll try to "wake up the machine" by bidding just right of midpoint for a small 5 or 10 lot to get started and wake up the market maker. For illustration, let's say in this case the market was $1 bid, $2 offered. Ok let's bid $1.60 for 5 and get him to tighten up. Usually the machine will fill me a few times, wake up after the first ten or twenty lots, and tighten the market toward say $1.40 x $1.85, and I'll continue to bid just right of the midpoint and get filled on bigger size.
For the first time ever, something else happened. I bid right of mid… and the entire quote gapped above me to $1.65 x $2.40. Now the implied vol was 55+.
This has never happened to me before. I have never seen the machine rip the entire market quote above the mid and rerack a quote range and stay wide.
And then it occurred to me… the illiquidity at the top of the S&P book which John Flood mentioned … that $3mn top-of-book when it's normally >$15mnn? It's even worse in options now than people think, and the implied vol in the screen is understated if you want to transact in any size.
In other words, over just the past few days, Dispersion Bros are having to pay up far more for single stock vol than before, amidst rapidly widening spreads, meaning their "long side" of the harvesting strategy is starting to buckle. If correlation continues to rise, then we may move into a world where constituents move 2-3% together, but if implied vols for single stocks are say 50, but all realizing 40 together because correlation is going up and they are all moving the same amount together, then the long book loses money. And once the long book starts failing as correlation starts to move the index 1-2% a day (meaning less harvesting of the short index vol book), and this persists for a few days… they have to cut risk and degross. If they do that, they have to buy index vol to cover the short leg… and then index implied vol starts to rise, feeding back into realized constituent volatility. And then your S&P starts dropping 1.5-3% per day…
I believe the Rolling VAR Shock has the Dispersion strategy in its sights, and this will release index vol.
And there's still another problem elsewhere. Remember the infamous Basis Trade that everyone was going on about in 2023, and has oddly become invisible? I talked about this at the end of the bond note last week here. The ending is the important part:
What you see above is asset managers (the big institutional long money in green) getting more long over the past five years because it is more capital efficient to own duration in futures than to buy cash treasuries. The main seller of those futures has not been the banks but rather hedge funds (red, inverted) who buy slightly-off-the-run cash bonds and short futures to asset managers. These hedge funds collect the arbitrage between the two by delivering their 'cheap' cash bonds into the futures when they expire. Notice what has been happening in recent months in the chart above: their net short has declined by over 25%. Why? I'm not sure, but I think swap spreads have something to say about this:
What are swap spreads? They are the difference between the fixed rate on an interest rate swap and the yield of a government bond of the same maturity (SOFR swap minus Cash Treasury). It effectively captures the relative pricing, credit, funding, and liquidity conditions between swaps and cash bonds. Swap spreads today are structurally negative because when they used to be positive, they reflected higher credit and liquidity premia in bank/derivative space relative to "risk free" government bonds (along with balance sheet and funding costs to go on swap vs cash).
In the chart above you can see above that these spreads — especially the 10Y and 30Y — are becoming more deeply negative and approaching Liberation Day levels. A more deeply negative swap spread means the Treasury cash bond yield is high relative to the equivalent swap rate, i.e., Treasuries are cheap versus swaps. What can this mean in practice?
Heavy Treasury supply or selling pressure cheapens cash bonds (higher yields) when private balance sheets are reluctant to absorb it at prevailing rates.
Or a deeply negative swap spread could indicate strong demand to receive fixed in swaps (e.g., asset-liability hedging, duration demand from large funds via over-the-counter with banks), which pushes swap rates down relative to Treasuries (swaps are rising so this is less likely).
Binding balance sheet, capital, and SLR constraints that make the classic arbitrage (buy cheap Treasuries, pay fixed swaps) capital-intensive, so hedge funds need a much more negative spread to justify using balance sheet.
So when swap spreads move more deeply negative beyond what you'd ascribe to structural SOFR effects, it often suggests:
Treasuries are under pressure / relatively illiquid vs swaps.
Dealer and intermediary balance sheets are tight, limiting arbitrage capacity.
Markets are tolerating a larger "mispricing" between swaps and Treasuries because the marginal balance sheet needed to close it is scarce or expensive.
A combination of #1 and #3 above could lead basis traders to degross their books by selling their longs (cash bonds) and buying back their shorts (bond futures). Swap spreads going more negative reflect a basis trade potentially undergoing strain and confirms the positioning leakage above — precisely at a time when people stopped talking about it.
More concerning still is that the "plumbing" on the surface in terms of the overnight rates directly under the Fed's control seem well behaved. It's hard to say there is an immediate liquidity issue like we were seeing in Q4 before the Fed began to inject money via RMPs:
Which brings us to the pushback of how the Fed and Treasury can just keep the plates spinning like they always do. Just how does the Fed print oil? Moreover, how would rate cuts help exactly in an inflationary recession to ease the situation? The Fed already told us they are of two minds on an oil crisis and undecided on what to do. They are also lost on AI. And in any case, Powell is in charge until May. How does Trump's guy Warsh actually save the day, assuming his appointment is confirmed? We are in a multi-month window of institutional inertia that is extremely dangerous for the Pavlovian "Fed has our back" mindset.
Credit and Collateral
All this would be bad enough if we weren't already barreling into a credit cycle. I won't rehash all the cockroaches we have read about just in the past two weeks, but none of this should come as a surprise to readers. As far back as October I was writing about the disaster coming in CLOs, BDCs, and Private Credit here, here, and here as a big shoe to drop (I never got around to writing up my Big Short on life insurers during the holidays, but since everyone is now aware of the situation, it seems besides the point). The important part to what we are seeing today was this:
You could be the biggest private credit fan in the world, but if I show you two funds — Fund A and Fund B — where the portfolio of holdings is practically the same, but where Fund A is liquid and trades at a -20% discount to Fund B where you can sell (redeem) 25% higher — you would have to be mentally challenged in a truly special way to decide to stay with Fund B and not reallocate your funds from B to A. I'm not saying "waaaah sell private credit" (ok I'll get to that) — I'm just saying if you like private loans, Fund A is on sale. And this is a problem. Maybe many investors with long time horizons have enjoyed the benefits of infrequent marks and "make believe," but as my buddy Erik recently laid out in his excellent Private Equity Train Smash Update, we know that increasingly public pensions and university endowments are requiring cash to make payouts, to the point that we are seeing some PE or PC stakes being marketed in the secondary markets for…yep, 80-85% of NAV over the past few quarters.
The problem is the sequencing that comes next. It only takes a few % of private clients, family offices, or other money that moves faster than "let's talk about it next year at the meeting" to see this and ask for a redemption at NAV thinking the following:
The unlisted fund now needs to raise cash, so they sell some loans that are trading near NAV or otherwise unimpaired — the better portion of the portfolio (sell what you can, not what you want to). That means that what's left in the fund is of worse quality, less liquid, potentially troubled… stuff that at the margin is harder to sell.
You can see the problem now. The investors who don't redeem get stuck with garbage that is likely to be marked down, or at least more likely to slip than the holdings that were sold to meet redemptions. This is why gates exist. This is how credit funds in 2007 got stuck. The first one to panic gets all his money back. The last one is left picking up the pieces.
And in classic Minsky fashion, as investors are unable to get their money out of illiquid credit, they are starting to sell what they can, not what they want. And that means public credit. And they are doing it, as is typical, before S&P and Moody's have even started to really pile on with downgrades that pose a potentially systemic risk.
Not only are Investment Grade and High Yield spreads rising…
… but so is IG and HY implied volatility:
Look at this last one again. Notice in prior volatility spikes like Liberation Day 2025 or SIVB 2023, volatility spiked with MOVE (Treasury vol). This time credit vol is leading Treasury vol. Notice also what happened in 2023-24 — the volatility in IG and HY credit collapsed much faster than Treasury vol. This was telling us that the problem coming out of the 2022 bear market in equities and bonds was a rate issue, not a credit issue. Now credit vol is leading Treasury vol higher. This is both a credit issue and a rate issue!
Coming back to what finally forces a proper hedge fund degrossing: it's rising bond volatility. If the asset base at a giant multimanager is an upside down pyramid with Treasury collateral at the bottom and asset book at the top, you can play rotation games on high gross so long as the collateral remains stable. If the collateral starts to whip around — say because the 2yr Treasury yield exploded over Fed Funds on Thursday and rose almost 10bps that day…
…then more collateral in relation to the top of the pyramid is required, and that forces the top of the pyramid to narrow. That's Degrossing — when people stop buying hedges and start selling to raise cash.
Who is the Buyer?
So we have multiple hedge fund strategies that may need to raise cash via outright degrossing.
Retail we all know: insatiable demand that never ends… the new force. Which is why these comments are particularly concerning, as retail always buys the big top and sells the big bottom going back to the days of Livermore:
And then there are the warnings from Middle East sovereign wealth who, like their US pension and endowment brethren, are stuffed full of illiquid alts and real estate, and suddenly have a pressing need for cash amidst collapsing revenue…
At the same time, surplus economies of the Far East like Taiwan, Korea, and Japan are getting absolutely rocked by the current crisis. Here is the Brent crude front month price in local currency — we are at Russia/Ukraine highs, and that's before touching on LNG availability:
At what point do these economies see they have less petro/trade dollar to recycle in foreign assets due to plunging trade volumes, and simultaneously realize (as their currencies are now sliding) that they may have to liquidate Western assets to support their FX and local economies, lest skyrocketing energy costs touch off an inflationary spiral?
And we are two weeks away from Japan's year end…one of the biggest creditors in the world…what if the Japanese need to take their ball and go home in the next few weeks?
So let's review:
Most hedge fund strategies from L/S to rates and basis traders (and soon Dispersion?) are all losing money in accelerating fashion, and on the precipice of a potential forced degrossing driven by escalating volatility in Treasury collateral.
Retail is suddenly going quiet.
Long only/private wealth are holding some of the lowest cash levels in history.
Sovereign wealth and Asian surplus nations are potential sellers (or a much less motivated buyer).
Did I miss anyone? Who exactly is supposed to buy stocks down 5-10% and keep an accident from happening — hedge fund macro short unwinds? Bessent and the Fed? Have people actually thought through the size of the numbers in these various cohorts in the context of a $3mn top-of-book S&P liquidity profile and options liquidity that is vanishing? That "wall of money" cuts both ways…
Progression of Sentiment Anecdotes
A month ago I had to open a chat with only 3 other people called "The Only Grumpy Old Men Left" because the vast majority of people we knew were either very bullish or bullish, and we needed a safe space for our bearporn.
When I wrote Tom Hanks Has Covid ten days ago, I was struck by the fact that one of the most intelligent, skeptical, cynical people I know in the industry was his usual skeptical self — but still pretty nonchalant because they will always print money, backstop an accident, and keep the plates spinning. Personally I'm sure how we print our way out of an oil crisis, but if that bro was not intensely bearish, where is the rest of the market… imagine how alone I felt while my chats were overrun with "no big deal" comments. Quite literally nobody is mentally prepared for a market accident. The most uneasy investors I know have only just very recently moved to cash.
On Wednesday I hear of a lone pod PM who was at his wits' end because the "Pretty Complete" Taco on Tuesday led most of the managers around him to cut aggressively hedges and once again jam the market (I'll come back to this in Positioning).
On Thursday afternoon I received multiple inbounds from a mix of institutional investors with very similar tone and subject matter [side note: this rarely happens — I don't get a ton of inbounds in the first place as I am pretty private, and most people either just read or hit my group chats]. The message subtexts were that they are growing uneasy at what they are seeing (and reading from me), they know things are going wrong, but are limited to a certain level of max cash and wondering where to rotate…where to hide. Think about that for a moment: we are back to hedging rather than degrossing. But in between the lines, I sensed a growing nervousness — handwringing while wearing handcuffs against getting out of the market and going quickly to cash.
On Thursday night I went to drinks and received confirmation from a friend that that everything I laid out about a rolling VAR shock (touched off by gold in late January) proved correct as allocators are now seeing this in their hedge funds: different strategies blowing up at different times (quant, L/S, rate PMs ten days ago)…and then suddenly everyone got hit this week. The time between strategy blowups is accelerating and converging as they struggle all at once.
The Anatomy of a Crash
Point #4 particularly concerned me above because I thought back to how crashes happen. In sudden moves like 1987's Black Monday, the market didn't just crash out of nowhere; it had been leaking lower for weeks, and investors were growing uneasy. Markets don't crash from euphoric sentiment. The crash from hyperextended positioning and unstable conditions is when people begin to grow concerned:
I am going to be unapologetically bearish here. If I'm wrong and it proves to be a giant contra for just how bearish everyone really is out there somehow — so be it.
There are consistent patterns exhibited by the significant market crashes in recorded human history. I call it The Anatomy of a Crash.
I believe the US market is now starting to display this pattern, and I believe the probability of a crash is both very elevated and close (days/weeks, not months).
The Anatomy of a Crash is:
An initial correction and a relief bounce
A retest of the low, often with a minor breach and an intraday "tail";
A 1-to-2 day final bounce ("all clear, we can rally because everyone is bearish");
Collapse.
Some examples…
1987 Black Monday shown above. I have marked the four stages:
The 2021 bitcoin crash from $60k → $30k:
The Granddaddy of crashes in 1929 which set off the Great Depression…unfortunately intraday data unavailable but the pattern is the same:
This is not a US-specific phenomenon. Behold the Hang Seng crash of the 1997 Asia Crisis:
Here is where the Nasdaq 100 — the most concentrated of the major indices with 40% in the Mag7 (vs the S&P's 32% currently):
A reminder that the divergences across different indices is the most notable I think I have seen in my career.
All time highs:
Nasdaq 100 — Oct 29, 2025
Russell 2000 — Jan 22, 2026
S&P 500 — Jan 28, 2026
Dow Industrials — Feb 10, 2026
Dow Trannies — Feb 11, 2026
NYSE Comp — Feb 12, 2026
I would argue the "real" high in risk was October because that is when is when beta peaked out:
Oh one more thing… October was also when Mag7 peaked out:
Again, addition to the initial conditions of the Turducken, what most concerns me is the convergence of the various flashpoints all at once around everything — all strategies, with seemingly nowhere to hide while investors are still asking not "should I sell" but rather "what can I rotate into to ride this out." A confluence of sellers, an illiquid tape, and a potential crash pattern.
Notice that at no point did we discuss the escalating oil/supply chain crisis in everything. You have read enough about that from me over the past few weeks.
Needless to say, I am at full attention for how the indices respond here in the coming days.
While I know they are a Blood Sacrifice and even mentioning their existence in my book banishes them to the mass graveyard of losses, on Friday morning I bought a significant amount of VIX calls. You may think "well that figures, he's a Cloudbear," but this also deserves some context: I have not bought near-dated VIX calls in nearly sixteen years, since late April 2010. "IYKYK."
This is not financial advice, just a diary. Remember I'm a Cloudbear, not your advisor.
Stay frosty…
As always, kindly yours,
Paulo aka Cloudbear