In Episode 59, Benny & The Squirrel welcome Craig Coben to the show to discuss the SpaceX IPO.

Craig is a veteran capital markets banker. After starting as an attorney at Sullivan and Cromwell, he transitioned into investment banking in the 1990s at Deutsche Morgan Grenfell. He joined Merrill Lynch in 2005 going on to run the EMEA and later Global Equity Capital Markets businesses of what we now all know as Bank of America.

Since retiring from the sell side 3 years ago, Craig has turned his skills to the keyboard where he has been writing really fantastic columns for the Financial Times’s Alphaville team as well as for Global Capital Magazine, where he is the resident agony aunt for frustrated sell siders.

The show was recorded at 2am EST on Friday 6th June.

Become a paid subscriber to get the audio edition and the slide deck - as well as the rest of the 🐿️’s tireless work!

Read more

Episode 58 was recorded at 6pm EST on Sunday 7th June.

Reminder to join us on Thursday: Craig Coben (former Global Head of ECM, Bank of America) joins us 6 hours before SpaceX IPO opens for trading on Friday morning. SpaceX/Tesla merger Polymarket odds are still at at 5 cents…

Today’s Slides

Pdf Of Slides

2.22MB ∙ PDF file

Download

Download

Audio only version (for paid subscribers)

Pod Summary

Reverse De-Equitization & The Supply Thesis

• Rupert frames the core issue: hyperscalers moving from internal cash flow → private debt → public debt → equity issuance to fund AI CapEx, completing the full capital structure cycle

• Historical parallel drawn: Europe’s “de-equitization” post-2003 (buybacks, 401K flows, PE take-privates) drove US equity outperformance for 15 years — that tailwind is now at risk of reversing

• The 1990s analog cited: it was secondary offerings (not IPOs) that caused supply-driven market problems; COVID-era SPACs and restructuring paper offers another cautionary historical parallel

The Signaling Problem: Why Are the Richest Selling? (Minute 3)

• Core message: when Google, Meta, SpaceX/Elon are tapping equity markets, this is a signal, not just a supply event — the world’s richest companies are cashing out retail savings

• IG credit market is maxed out on capacity; hyperscalers know tipping it would be catastrophic, so equity is the only remaining lever

• The Friday jobs report gave a misleading “strong” headline (World Cup hires, birth/death adjustments) — Fed stays hawkish as a result

S&P Index Committee Decision & SpaceX IPO Demand (Minute 6)

• S&P index committee refused to fast-track SpaceX and other AI IPOs into the index (unlike Nasdaq/Russell) — this removes ~$19 trillion of passive benchmark flows Elon was counting on

• SpaceX IPO priced at $135 fixed offer; gray market showing only modest ~$170 pop; retail can get stock in “any size” — a bad sign

• Roadshow anecdotes: JP Morgan event described as bizarre (Asteroids games, bomber jackets); Morgan Stanley ran a “theater event” — frothy atmosphere raising red flags

• Gray market (HyperLiquid perpetual future) showing muted enthusiasm; caution urged on reading thin weekend markets as signal

• Brief crypto tangent: Benny quit crypto trading in 2020/21 because the 24/7 nature was unsustainable

S&P Valuations at Historical Extremes (Minute 10)

• WisdomTree data: over half the S&P trades above 10x forward sales — no historical precedent

• Equal-weight S&P (RSP) has <10% exposure to >10x price-to-sales stocks; bulk of exposure is <4x forward price-to-sales

• Trade idea: RSP (equal weight) vs. QQQ as a relative value trade — acknowledging 20 years of equal-weight underperformance but sees potential for violent re-rate; levered spread trade doesn’t need a huge move

AI ROI Shift: From Token Maxing to Cost Optimization (Minute 12)

• Key inflection point observed: the market is moving from “token maxing” AI agent hype → AI ROI and cost routing optimization

• Benny built an AI agent with a daily cost optimizer capping spend at 3 cents per task — this is what enterprise users will start doing

• AI cloud cost opacity compared to “heli-skiing with richer friends” — the bill shock is coming for enterprises

Bitcoin: Head & Shoulders (”Classic Clean”) (Minute 15)

• Bitcoin chart showing a near-perfect head and shoulders pattern

• Double problem identified: (1) PBOC/BOJ/Fed global liquidity tightening; (2) no narrative — former miners have migrated to AI infrastructure

• Novogratz’s Galaxy Digital pivoting to data centers (West Texas campus = half its value) signals crypto top-riders have jumped to AI infrastructure bubble

• This feels like 2021 crypto mania repeating in AI infrastructure — same “bubble riders” (Novo and The Mooch) now in AI; the data center bubble thesis worth watching

Brent Crude: Long Position Initiated (Minute 18)

• Rupert announces he took a large position at the front of the Brent crude curve last week — “enough was enough”

• Goldman Sachs energy sentiment survey at time of purchase: more bearish than COVID lows — classic contrarian setup

• Bull thesis: oil in the “wrong places,” 3–6 month product/crude shortages baked in; energy company earnings call transcripts showing 30–45 day inventory warnings

• Bear counterpoint acknowledged: JJ (veteran NYMEX trader) is in the “world has plenty of oil” camp

• Risk flagged: if energy rallies hard into July, it crushes everything else (reflationary shock)

The Four Horsemen Framework: Energy + Rates + Dollar (Minute 22)

• Energy flagged as Horseman #1; rates picture described as completely changed

• Dollar: Rupert is a medium/long-term dollar bear but has insurance against a dollar ripper; big bearish candle in the Euro noted

• Benny is a “massive Euro bear,” holds December Euro puts now back in the money

• Yen back above 160 — a 35-year technical line; Benny targets 200 if it breaks through

• MOF interventions keep pushing it down but it re-emerges (beach ball analogy)

• BOJ meeting in ~10–12 days: if forced to hike rates to even zero, it unwinds the global yen carry trade — massive FX vol risk

• Japan banks: still the standout trade — yield curve steepening is “a gift”; valuation normalization has played out, now it’s an earnings growth story

Volatility: Nasdaq/S&P Vol Divergence (Minute 26)

• Nasdaq vol / S&P vol ratio is “absolutely out of line” vs. Nasdaq/S&P price ratio — a dislocated market that must revert

• 25-delta skew also starting to turn — suggests the vol unwind is not done

• Thursday/Friday’s Nasdaq -5% move burned off gamma inventory quickly (due to daily/weekly options), limiting gap-down follow-through Monday

• Korea flagged as the “leader to watch” — massive leverage needs to be unwound

• Friday generated $55bn of levered ETF supply forced to come to market — amplified the Nasdaq -5% move

• Market not gapping down Monday despite Iran news = positive sign that leverage cleaned out on Friday

John Law / “The New Mississippi” Teaser (Minute 32)

• Benny teasing his latest essay “The New Mississippi” — parallels between John Law (Scotsman who bankrupted Regency France via the Mississippi Company/Compagnie des Indes) and current AI/Silicon Valley state-capital fusion

• Trump administration reportedly considering taking stakes in AI companies — government entering cap table of peak-valued assets is historically ominous

• John Law anecdote: the word “millionaire” was coined during his bubble; SpaceX may introduce the first “trillionaire”

This week’s ‘Start the Week’ was sent via email to existing paid subscribers at 4.50am EST on Sunday 7th June if you want to search your inbox. It will be put back in the online archive here once subscription billing is un-paused next week.

Start the Week note will be brief again this week. A few changes to the portfolios this week that we need to go through. I will finally make it home to Melbourne mid-week when normal publishing cadence will recommence (and billing will be un-paused).

Supply!

Last weekend’s note covered the 🐿️’s JOMO (Joy Of Missing Out) philosophy as regards the prevailing feeding frenzy in the memory semis market. Last week’s rout of the leaders frankly came much sooner than I had been expecting.

In last week’s Sunday Show, Benny & I discussed the shift in the sources of funding that the hyper-scalers were employing to finance AI data center (‘AIDC’) capex.

Quick plug: the interview that we recorded with Ameesh Agarwal last week was a fascinating deep dive on on AIDC credit markets. Do at least scan the show notes if you do not have the hour to spare. We have more work to do on this space - there will certainly be trade ideas.

What initially was being covered by internal free cash flow from the tech giants soon migrated into private then public debt capital markets. I am still struggling to understand the corporate finance logic behind financing a 10-year useful life asset with a 30-year JPY/GBP bond issue.

The 🐿️’s worst ever personal macro call was to leave an ECM job in Asia in July 2003 - on the eve of a decade long Chinese IPO mega boom. On the flight back from Hong Kong to my new job in London that summer, I read Rob Buckland’s note (below) with a degree of trepidation. What looked to be great outlook for equity prices (and it was!) and for M&A bankers was decidedly a ‘no bueno’ for IPO bankers!

Since the start of the year, I have been increasingly concerned that the tailwinds that have supported US equities in the past 15 years - namely share buybacks, LBOs and 401k flows - could be about to move into reverse just as the market faces a wall of AI-related equity supply.

Back to the Sunday Show. We were discussing this very topic. I commented that “…if we start getting MAG7 common stock supply on top of these IPOs, I’d be really worried”. Less than 24 hours later, the Alphabet $80bn capital raise news hit the tape. The physics of the SpaceX IPO layout just got considerably harder.

The 🐿️ had been booked by CNBC in Hong Kong to chat about the SpaceX IPO the following morning. But for me, the Google news was probably the bigger deal to discuss (link to the relevant clip via image below).

Google’s raise certainly robs liquidity from the SpaceX deal in the near term. More importantly, it has already potentially triggered its first ‘copy cat’ deal from Zuckerberg.

Who next? Amazon? Microsoft? The age of reverse de-equitization appears to be upon us! The hyper-scalers and foundation model companies could be issuers of primary equity for the foresee-able future.

In my view, the S&P index committee has 100% made the correct call to require SpaceX, Anthropic and OpenAI to satisfy the profitability and seasoning standards before inclusion in the benchmark.

I had assumed that they would cave alongside the Nasdaq 100 and FTSE Russell compilers. Am glad that they did not.

For me, we have the makings of a really interesting long equal-weight S&P versus short market-cap weighted S&P (or or ) trade that could work for the next couple of years as the largest stocks in the index become net suppliers of paper to the market.

This may now need some further refinement following the S&P announcement. More on this to follow. The chart below from WisdomTree has been doing the rounds lately on social media. Scary right?

Because the 🐿️ is a nerd, he decided to put together the same analysis of the S&P500 equal-weight / :

With less than 10% exposure to paper valued at > 10x Price to Sales, I kind of know where I want to rent my beta over the next few quarters -especially if it is those most expensive and largest components of the US benchmark are are planning on being net suppliers of equity paper to the market.

BUSHY™ and Acorn Review

Read more

In Episode 57: Ameesh Agarwal (CIO, University Place Asset Management) joins Benny and the 🐿️ to dissect the economics of AI data center investing. His firm underwrites the base infrastructure layer — shell, power, and location — avoiding GPU exposure due to uncertain 4–7 year depreciation cycles and negative unit economics.

The show was recorded at 6pm EST on Wednesday 3rd June, 2026.

I am not able to publish paywalled commentary while on a subscription billing pause - so for the next couple of weeks, enjoy these show notes for free!!

Audio only version (for paid subscribers)

Podcast Summary

Introduction

• Ben (Benny) introduces Ameesh Agarwal, CIO of University Place Asset Management (‘UPLAM’), for a special data center-focused episode

• Ameesh’s background: UPLAM has historically been an alternative assets investor — opportunistic credit and real estate across geographies

• Entry into data centers was organic: demand for capital outpaced supply, creating attractive risk-adjusted yields; LPs (primarily insurance companies and securities firms) had appetite to invest via infrastructure arms

• Data center underwriting mechanics are similar to real estate — long-duration, income-producing assets

The AI CapEx Supercycle & Scale of Spending (Minute 3)

• The 2023 ChatGPT/OpenAI “aha moment” triggered rapid AI diffusion and the concept of “a country of geniuses in a data center”

• Hyperscaler CapEx estimates: ~$600B initially for 2026, revised to ~$800B; total proposed hyperscaler spending is $1.1T

• Per NVIDIA’s earnings call: 50% of forward GPU demand is hyperscalers, 50% is neoclouds and others — total sector CapEx consensus is well north of $1T

• Capital requirements can only be met by the largest companies; smaller players are structurally disadvantaged

SpaceX / xAI Discussion — Colossus, Grok & TAM Claims (Minute 4)

• SpaceX’s upcoming IPO and the question of whether its terrestrial AI compute (Colossus I & II) is included in the $1T CapEx figure

• Ameesh is skeptical of xAI’s position: xAI/Grok claims a $27T TAM (close to US GDP), with ~90-93% attributed to AI, but xAI only captures ~3-4% of the actual AI frontier lab market

• Ameesh personally uses Anthropic every day but doesn’t know “a single person besides Twitter users who use Grok”

• Colossus is effectively acting like an Iren or CoreWeave — leasing GPUs, providing compute to Anthropic (paying ~$1.5B/year, representing $25B of revenue) — a pure GPU-rental model, not a frontier lab

• 🐿️: “There’s two trillion dollars riding on that being a reality” — skepticism is noted but the stakes are enormous

Data Center Economics 101 — Real Estate Analogy & Key Metrics (Minute 7)

• Data centers are not fundamentally different from real estate leasing: long-term leases (10–15 years, with 5–10 year extensions), similar to warehouse leases

• Key metric difference: real estate = price per sq ft; data centers = price per megawatt

• Location hierarchy: Tier 1 markets (Northern Virginia, downtown Chicago) > Phoenix/Austin > remote locations — just as Manhattan sq ft > New Jersey sq ft

• Location matters for: fiber access, “meet me rooms,” latency, network nodes, available/reliable power, and water for GPU cooling

• The four critical underwriting factors: (1) Location → (2) Power → (3) Developer/Sponsor quality → (4) Tenant quality

• Experienced developers (Vantage, Aligned, Compass One) vs. overnight opportunists — Ameesh sees clear differences in sophistication and underwriting quality

• Ameesh’s firm finances the base layer (shell + power), not the GPU layer — by analogy, the “roads” needed regardless of which car wins (Mercedes, BMW, train etc.)

• The shell and power have indefinite useful life (”Virginia is not going anywhere”)

• GPUs are the problem: useful life is debated — Michael Burry says 2–3 years, hyperscalers underwrite 4–5 years, Ameesh uses 6–7 years as a generous assumption

• Even aircraft leasing has a longer economic lifespan than GPU hardware

• Risk is not fully eliminated for the shell lender: if the tenant (who owns the GPUs) fails, the lease collapses; tenant quality is therefore paramount

• Hyperscalers (Amazon, Google, Microsoft) are the best tenants — fortress balance sheets, multiple downstream service layers that generate higher revenue per GPU CapEx than pure intermediaries

CoreWeave’s Unit Economics — The WeWork Analogy (Minute 16)

• CoreWeave’s business model: rent GPUs from NVIDIA, take on leverage, lease compute to frontier labs and hyperscalers — a pure intermediary with no proprietary software (all open-source) and no moat beyond GPU access

• Revenue generates negative unit economics: buying a GPU that depreciates in ~7 years and leasing it out does not generate enough revenue to repay CapEx at any reasonable WACC

• CoreWeave positions itself as a “systems integrator” — but every cloud provider offers the same capability

• WeWork parallel: WeWork leased office space, sub-leased it at a loss, dressed it up as a “software company” and “community platform”; CoreWeave does the same with GPUs; the buzzword is “AI infrastructure” instead of “community”

• Leverage amplifies the problem: “leverage makes any bad investment worse”

• CoreWeave’s debt position: ~$27B drawn, ~$13–15B remaining capacity across credit lines and delayed draw facilities

• Revenue target: ~$10B for 2026; even if quadrupled to $40B at steady state, the business remains problematic when measured on EBIT (not EBITDA) with 4–5x debt/equity ratio

The Valuation Methodology Problem — “Make DCF Great Again” (Minute 21)

• Investment banks model CoreWeave to its one peak EBITDA year, apply an EV/EBITDA multiple, and declare a valuation — this is fundamentally wrong

• EBITDA multiples assume perpetual cash flows; CoreWeave’s cash flows are NOT perpetual because they require constant CapEx reinvestment equal to GPU depreciation cycles

• Correct approach: normalize EBITDA by replacement CapEx based on GPU useful life; at any assumption (4, 5, 6, or 7 years), with 10% residual value, the business is insolvent

• The cash flow pattern: negative at CapEx → negative for years → briefly positive → CapEx cycle repeats; IRR is approximately 2–3%

• Rupert frames it: “It’s an equipment leasing business — what’s the value of the equipment, how long is there value to it? This ain’t ski rental” (skis pay for themselves in two rentals; GPUs take 6–7 years)

NeoCloud Taxonomy — Not All Are Equal (Minute 23)

• Ameesh distinguishes between fundamentally different neocloud business models:

  • Pure shell/power providers (Applied Digital, Cipher Mining/TeraWolf — former Bitcoin miners): provide power and shell only, lease space to hyperscalers or neoclouds; do NOT buy GPUs; power is their moat

  • Full-stack integrated operators (Iren, Nebius): own the power + shell AND buy the GPUs — end-to-end integrated

  • GPU-only neoclouds (CoreWeave, Fermi): lease power/shell from a developer (paying a substantial premium), buy GPUs on leverage, lease compute out — highest cost structure

• Why Iren and Nebius are better positioned than CoreWeave:

  • They own their own power/shell (no premium lease payments to developers, who target double-digit returns / 2x money over 6–7 years)

  • They carry the long-life power/cooling asset on the balance sheet (much longer depreciation)

  • Unit economics are “slightly better” — but GPU depreciation risk still exists for both

• CoreWeave’s structural disadvantage: must sign leases at a significant premium (vs. hyperscalers) because developers price in higher risk; all that margin is extracted before GPU revenues are even considered

• “The only reason CoreWeave exists is so NVIDIA has an alternate customer versus hyperscalers who will eventually come for Jensen’s margins”

• Ameesh agrees: CoreWeave is a “distribution shelf” for NVIDIA — not a business with proprietary IP or moat

• NVIDIA’s $2B investment in CoreWeave is effectively free optionality: NVIDIA received $2B+ back in GPU revenues the same year; if CoreWeave succeeds, the investment multiplies; if not, principal is recovered

• Historical parallel: Lucent financing its own customers during the dot-com era (keeping the build-out going, then collapsing)

• Anthropic now trains on Google TPUs (not NVIDIA GPUs) specifically because Google made early investments — NVIDIA failed to invest and lost the customer

• NVIDIA’s moat deepened by customer lock-in: high switching costs once companies build on GPU infrastructure; that’s the strategic logic behind financing distribution channels

Power Scarcity as the Durable Moat — Former Bitcoin Miners (Minute 36)

• “Never bet against engineers” — over time, efficiency improvements will reduce power and cooling requirements per unit of compute

• However, even as efficiency improves, the grid is backed up and cannot be built overnight; demand growth is outpacing efficiency gains for the foreseeable future

• Larger data centers are increasingly pursuing behind-the-meter on-site generation — Ameesh’s firm finances some of these

• 50% of US citizens wouldn’t want to live next to a nuclear power plant; the figure for AI data centers is 70% — permitting is getting harder, not easier

• Former Bitcoin miners (Applied Digital, TeraWolf, others) with existing power rights are being valued as scarce commodity plays by the market — tough shorts even with stretched valuations because power scarcity is real and durable

• Ameesh’s firm now shies away from deals where power is “in the backlog” vs. in hand; deals with power secured vs. speculative pipeline are fundamentally different underwriting scenarios

Capital Markets Deep Dive — Credit Pricing, Debt Structure & Risk (Minute 39)

• Credit spread differentiation: hyperscalers (investment grade) price at ~200bps over SOFR; neoclouds price at ~400bps over SOFR

• Neoclouds increasingly unable to access private credit; CoreWeave recently resorted to the junk bond market (7–7.5% coupon)

• Junk bond buyers are not sophisticated underwriters — they are bond traders relying on ratings, with no concept of GPU depreciation cycles; “they have no idea what they’re financing”

• Parallel explicitly drawn to 2007–2008: insurance companies, Asian and European desks buying paper without understanding the underlying — “Belgian dentists backing the truck”

CoreWeave’s Debt Structure — SPE Tranching & Capital Waterfall (Minute 42)

• CoreWeave has multiple tranches: revolvers at the corporate level + SPE-level debt secured by underlying contracts and GPUs

• SPE loans are written to depreciate in line with GPU lifecycles (~4–5 years), meaning revenue flows first to repay SPE debt before reaching the parent

• By the time revenues “hit the parent,” they are already collateralized

• At the corporate level: senior unsecured creditors are exposed with minimal credit enhancement

• LTV of 90–95% being written in some deals — “just be the equity guys” at 5% credit enhancement

• The system requires capital markets to remain permanently open: “any jam in the capital markets could cause major problems across the stack”

• Neocloud debt is material and growing in the high-yield index; insurance companies globally are buying this paper through products designed specifically for that market

The Offtake Agreement Dependency (Minute 45)

• Google previously guaranteed lease payments for tenants (e.g., FluidStack) as a mechanism to push Google TPU adoption — Google has since stopped providing these “credit wrappers”

• Credit wrappers are becoming less common — hyperscalers are pulling back from guaranteeing neocloud leases

• Neocloud financing is now entirely codependent on hyperscaler offtake agreements with no credit guarantee backstop

• If a hyperscaler pulls back or cancels a contract, the domino effect across the entire neocloud financing stack is severe

The Magnetar Trade Analogy — 2007 CDO Parallel (Minute 49)

• Ben and Ameesh connect the CoreWeave ecosystem to the Magnetar trade of 2006–2007:

  • Magnetar bought CDO equity tranches, owned senior tranches, and purchased mispriced CDS — effectively keeping the CDO market alive 18–24 extra months

  • Today, Magnetar is the largest shareholder of CoreWeave — same playbook, different asset class

• The circular financing dynamic: Magnetar/Google get paid via GPU revenues/cloud revenues while simultaneously being long the equity — structured to be OK either way

• Banks financing these deals retain zero risk (fee-driven), are not underwriting the assets, and are likely buying CDS against the same companies on different desks — “the same bankers retailing the risk”

Investment Analysis Quality — “The Gap Between Perception and Reality”

• Ameesh’s critique: most hedge funds, investment analysts, and sell-side do not understand what they are underwriting in AI infrastructure

• Analogy: a semiconductor analyst covering photonics (Lumentum), memory (SK Hynix), and CoreWeave is covering completely different businesses — requires fundamentally different frameworks

• The single most important input for CoreWeave valuation — GPU depreciation cycle — is routinely omitted or improperly handled in bank models

• Investment banks show the EBITDA-positive year, apply a multiple, and move on — ignoring that it is NOT a perpetuity

• Incentive misalignment: pod shop time horizons of 1–3 quarters, trend followers and passive funds dominate flows, bankers just want to get deals done for fees

How Does It Play Out? Catalysts & Positioning (Minute 55)

• CoreWeave has minimal economic value and debt spiral tendencies. The market is fragile: the gap between perception and reality is rarely this wide

• Math can be right but “you can still lose your shirt” going against momentum in open capital markets

• Goldman’s numbers show the year-over-year first derivative of CapEx growth is peaking right now — the rate of acceleration is topping out

• Framework: find the most robust (long) vs. the most fragile (short) in a capital market state transition

Episode 56 was recorded at 5.45pm EST on Sunday 31st May.

I am not able to publish paywalled commentary on a billing pause - so for the next 3 weeks, enjoy the show notes for free! Pod summary, show slides and audio edition are downloadable below.

Today’s Slides

Pdf Of Slides

1.96MB ∙ PDF file

Download

Download

Audio only version (for paid subscribers)

Pos Summary

Q2 Performance Review & The Semis Trade

• Squirrel confesses to a poor Q2 — sat on his hands and missed the market rally

• The one dominant trade of Q2: semiconductors — Korea, Taiwan, and SOXX (SMH) all parabolic

• Everything else either flat or dragged higher by semis; EM is now effectively a semi ETF in disguise

Micron & Valuation Concerns (Minute 2)

• Micron trading at price/ sales levels last seen at the peak of the prior tech boom

• Debate on nominal vs. log charts for properly contextualizing valuations

• Key question: what multiple do you put on peak semi earnings?

• Darius Dale’s point cited: earnings expectations are a lagging indicator — analysts chase stock price narratives

• Risk of an earnings bubble (analogous to 2006–07 housing, not a 1999-style price bubble) — multiples aren’t crazy, but earnings may be over-earning

“Amnesia” Theme Explained (Minute 5)

• The episode title explained: Squirrel’s portfolio has “amnesia” — no memory (stocks)

• He closed the call leg of his SMH strangle, cheapening September puts by ~15%

• Congrats to those who rode the memory/semi bull - time to take chips off the table

• ~54–55% of the market is passive — non-thinking participants

• Shorts (like Benny in April with SMH) got run over, leaving few on the other side

• The “shoot it in the back” strategy discussed — warned it’s harder than it looks (silver analogy: down 37% in one day)

• Momentum vs. mean reversion trade styles debated — both hosts identify as inherent mean reversionists

• Institutional buyers of Hynix/Samsung running into single-name concentration limits

• Marginal buyer now coming from structured products — leveraged ETFs/ETNs

• Downside leverage works twice as aggressively — asymmetric risk building

Portfolio Performance & Energy Positioning (Minute 10)

• Squirrel’s YTD performance chart shown: outperformed S&P through Q1, now behind after the semi rally

• Only 30% equity weighting in multi-asset portfolio; over half of equities is energy

• Watching BNO ETF (front-month Brent) — oil inventories have been decimated

• “Barrel counters” all came into 2025 bearish and are now really worried about supply

AI CapEx Sustainability & Valuation Sanity Checks (Minute 14)

• Ben currently doing deep work on Coreweave and the broader AI CapEx numbers

• Free cash flow yield on Russell/QQQ is thin — market in “rare air” across multiple valuation metrics (Q ratio, Buffett indicator, trailing P, forward P)

• Combined with potential earnings bubble = “very interesting situation”

• Stocks as claims on the economy’s assets vs. actual income/GDP — market stretched at historic extremes

• References Michael Pettis’s “bezel” concept: paper wealth exists only if you can actually exit

• Post-GFC we’ve been in persistent overvaluation; 2022 was a small drawdown but “we’re back”

CapEx Funding Waterfall (Minute 18)

• Nomura chart referenced: decomposing AI CapEx into cash-flow-funded vs. debt-funded portions

• Capital markets funding sequence: cash flow → term loans → high-yield/IG bonds → equity

• Warning: if Mag 7 start issuing common equity on top of IPOs, that would be a serious red flag

• Cancellation of buybacks = first sign of equity dilution coming

• $36B private credit deal done to fund Anthropic chip purchases cited as example of debt-funded CapEx

SpaceX IPO Valuation Sanity Check (Minute 24)

• Compared SpaceX pre-IPO valuation vs. revenue multiples of prior major IPOs (Meta/Facebook, Google, etc.)

• Facebook IPO anecdote: priced at $38, closed at $38 (Morgan Stanley used the whole shoe), then dropped to $17

• Pre-IPO SPV deals with big front-loaded fees now circulating for SpaceX — reminiscent of Facebook retail frenzy

• Day-one IPO returns historically poor for hot names; mechanics mean insiders are waiting to sell

Coreweave / Magnetar / CDO Parallel (Minute 26)

• Magnetar made $10B+ on Coreweave trade; drew parallel to their CDO trade pre-GFC — same characters, same playbook (get in early on the debt-funded boom, get out)

No changes to discuss beyond closing the call leg of the strangle (retaining the September puts). BUSHY™’s assault on the front end of the Brent curve is still pending - but we are getting close.

In case you missed it, the 🐿️ made a guest appearance on The Compound with Josh Brown and Michael Batnick last Wednesday. They asked me on to question me about my work on the SpaceX last week.

Hynix JOMO

The 🐿️ is suffering from a grave bout of amnesia - he has NO MEMORY (in his portfolio)! The only recent exposure I have had to the investor feeding frenzy in DRAM and HBM semiconductors in the past 2 months has been the call leg of the strangle in BUSHY™’s hedge book - and we took that off last week!

As I have mentioned many times before, I have PTSD-related reasons for not participating in this cycle’s DRAM risk ‘orgy’ in Korea (and elsewhere). Significant personal history with multiple recapitalizations of Hynix in the early 2000s has left me with scar tissue and a core mantra that there is no such thing as a value trade in the highly cyclical semiconductor sector.

Read more

In Episode 54, we welcome Michael De Clercq to the podcast. Michael is the CEO and founder of China Performance Group and someone “who has forgotten more about China than a lot of people will ever know”, having been doing business in China since the late 1970s.

Michael has just launched a Substack on China and Geopolitics. As you will be able to tell from our conversation, I have a hunch that Michael’s thoughts will become essential reading. Sign up here 👇:

Reflections on China and Geopolitics

Insights and reflections on China, global supply chains, trade, and geopolitics from five decades of firsthand operational experience inside China.

By Michael De Clercq

The show was recorded at 7pm EST on Thursday 28th May.

I am not able to publish paywalled commentary while on a subscription billing pause - so for the next couple of weeks, enjoy these show notes for free!!

Today’s Slides

Pdf Of Slides

823KB ∙ PDF file

Download

Download

Audio only version (for paid subscribers)

Podcast Summary

China’s Rise: Setting the Scene

• Michael discusses China’s rise — from a country that couldn’t afford to send delegates to UN events to BYD becoming the world’s best car company

• Michael’s own journey has been an “an interesting ride” beginning with his first China trip in 1975, followed by time in Beijing in 1976 for the funeral of Zhou Enlai

  • Zhou was Mao Zedong’s right-hand man, Premier of China, educated in Paris, and survived the Cultural Revolution through intelligence

  • Zhou facilitated the secret Kissinger meetings and introduced Mao to the idea of meeting Nixon in 1972 — the first major unlock of China to the West

• Michael attended Zhou’s funeral as a 21-year-old on crutches, staying at the French Ambassador’s residence; he notes the genuine mass reverence for Zhou he witnessed

Key Milestones in China’s Rise (Minute 5)

• The Deng Xiaoping era: Rise of pragmatism; Deng famously said “there’s nothing wrong with being rich,” unleashing China’s natural mercantile instinct

• China was then on a coupon system — money had no real value; you needed rice coupons, mechanical coupons; real estate was not a capital asset

• Michael cites Chinese dominance of Southeast Asian wealth (top 10% of richest families in SE Asia are ethnic Chinese under Indonesian, Thai, Malaysian names) as evidence of this deep commercial instinct

• 2001: China’s accession to the WTO as the next big milestone

• 1989: The Tiananmen incident — Michael lived in Beijing at the time; Western companies threatened to leave but were “too committed already” and stayed

Michael’s Personal Commercial Journey in China (Minute 10)

• Michael studied Chinese at ANU in Canberra; fellow students told him it was “completely useless for business” — only useful for Taiwan trade; he was better off learning Japanese

• He graduated and opened an office in China in 1978 — “no logical rationale, incredibly risky,” no bank would finance the business plan

• Immediately created a two-tiered business: supply chain/sourcing (now the core of CPG today), and a second consulting arm

• He predicted China would become a massive manufacturing base for the West; when working with the US-China Business Council, China was doing just $350 million in exports to the US; officials doubted it would ever reach $500M — now it’s in the $450 billion range

• Chinese factories in the late 1970s weren’t particularly interested in selling — pricing was driven by the need to acquire US dollars, not by product cost

• All buying had to go through Beijing’s central foreign trade corporations; you couldn’t visit factories or negotiate with provincial entities

• Products were garments and footwear; you bought “hundreds of thousands of dozens of shirts” nearly as a commodity, specifying sizes, colours, yarn, dye — you didn’t know where the factories were

• By the mid-1980s, private Chinese enterprises were allowed to trade directly with the West — a pivotal change

Industrial Clusters: China’s Natural Organizing Principle (Minute 17)

• We discuss regional specialisms and when this cluster model formed

• Michael explains clusters formed organically in the 1980s through copycat behavior — factories saw neighbors making money and replicated, which drew in support industries (yarn, buttons, lining) nearby, amplifying competitiveness

• This contrasts with the Western obsession with USP and moats: “The Chinese government actually promotes competition and is against monopolies to the extent they can control” — a key reason Chinese equities haven’t tracked China’s economic growth

Why Chinese Equities Have Underperformed (Minute 20)

• Chinese equities haven’t done well despite China’s growth; Michael explains Chinese companies (e.g., BYD) are brilliant at penetrating their own market but the West’s analytical framework excludes them

• The reason: difficulty valuing Chinese assets (all within Chinese geography), lack of transparency, and China’s own economic moat — the controlled currency and protected market

• Historical lesson: China’s awareness that isolation caused its fall (Opium Wars 1842, loss to the East India Company; defeat by Japan — a former vassal state) drove Deng’s opening-up logic

The CCP’s Optimization Strategy vs. the West (Minute 27)

• The West optimizes for profit; China optimizes for full employment and retaining power

• Michael identifies 2013 (Xi Jinping’s ascension) as a new milestone — a “left turn” from free-market capitalism, which had become “predictably quite corrupt” (pay the right people, get things done)

• Xi’s anti-corruption campaign cleaned up political competitors and the corruption culture — economic potential wasn’t significantly damaged

• Michael’s core thesis: the CCP sees itself as “the new imperial dynasty”, seeking to rule for a thousand years, driven by a genuine belief they are the solution to making China great again

• In the 1700s, China was the largest economy in the world by a long shot; the Opium Wars, Boxer Rebellion, and defeat by Japan reduced it to “the poor man of Asia” — leaving a deep chip on the shoulder of Chinese patriotic leadership

• Distinction from MAGA: China’s desire to be great again is genuine historical restoration; “America was great to start off with and continues to be great”

• Unlike Cuba (hold power regardless of national welfare), the CCP holds power because they believe they can make China great — so they invest in education and poverty reduction even though a nation of 10 million unemployed university graduates creates dangerous social discontent

The Chinese Consumer Market: Why the Engine Never Quite Starts (Minute 36)

• We discuss what’s preventing the “Chinese consumer engine” from turning on — every foreign business has dreamt of 1.2 billion consumers

• Michael identifies two structural barriers: the lack of a welfare system and the hukou system

• Welfare paradox: Despite being communist (which Westerners assume means generous welfare), China’s social safety net is no better than the US system — Chinese people hoard savings as insurance against hospital costs and life events; they are “great savers and poor consumers”

• Consumer spending is only 25–45% of Chinese GDP vs. 65%+ in the US

• Xi Jinping is personally afraid of welfare-induced laziness — so he won’t expand it

• The CCP’s substitute for welfare: manufactured holidays — when Michael started working there, people worked 7 days a week; the CCP discovered holidays drive spending and now creates multi-day holiday blocks (e.g., tomb sweeping day + weekend = 5-day block, but workers make up days the prior weekend)

• Hukou ties your rights as a citizen (free schooling, kindergarten, social services) to your birthplace; moving to an urban center means losing those rights

• Historically 80%+ of China’s population was rural; urban migration happened through informal “turning a blind eye” because cities needed workers

• Second- and third-tier cities (e.g., Hefei) are beginning to waive hukou requirements to attract workers and boost local consumption — an early-stage reform

• The core reason: social stability — the #1 metric by which every Chinese provincial governor and city mayor is measured

• Hukou is a social control mechanism; abolishing it risks instability in urban centers

• Michael agrees that technology (AI, surveillance) could replace the hukou’s control function; second- and third-tier cities are already evolving it; but it will take 5–10 years to phase out entirely

Taiwan: The Risk/Reward Analysis (Minute 46)

• Ben asks for Michael’s view on the real risk of a Taiwan invasion — “the number one thing market participants love to talk about”

• Michael is very confident there is no imminent military risk of PRC invasion of Taiwan

• Historical background on why Taiwan matters to Beijing: In 1949, Chiang Kai-shek and the Kuomintang fled to Taiwan “temporarily to regroup” with China’s capitalist elite — it remains “unfinished business” in the CCP-KMT civil war

• Taiwan is also an ideological embarrassment: an average Taiwanese is roughly 11 times richer than an average mainland Chinese — a Western capitalist system that directly contradicts the CCP’s narrative that communism is the only path to Chinese greatness

• But historically, Taiwan didn’t even “belong” to China until ~200 years ago — it was Japanese for a long time and is ethnically Polynesian at its origins

Why Invasion Won’t Happen: Risk > Reward (Minute 54)

• The reward: Taking back the island (primarily a political win, not about chips — “that’s not the real reason”)

• The risk: The CCP loses power — Xi Jinping’s #1 fear is not Trump, Putin, or the Uyghurs; it’s the Chinese people rebelling

• If China attacks Taiwan and fails, the CCP faces potential loss of legitimacy; if it succeeds, it invites global sanctions — destroying China’s role as “the factory for the world”

• It would essentially be a civil war — Chinese armed forces killing Chinese people, “never a good look”

• Taiwan as a deterrent card: The CCP uses Taiwan as a brilliant negotiating card to be pulled out when useful; the moment they execute on it, they lose that card forever.

• Ukraine lesson: Russia had to cross a line in the sand — not 85 miles of water — and look what happened; “the Communist Party pays attention to that”

The 🐿️ and Benny’s RMB Bet (Minute 59)

• Rupert asks Michael to adjudicate their long-running bet (18 months old) on the RMB: Rupert believes the RMB will strengthen vs USD (strike price: 6.50); Benny believes it will devalue

• Michael’s analysis: Everything the CCP does is politically driven; the RMB is a tool to dethrone the dollar — “quixotic to say the least” given the dollar’s dominance in global trade

• But: if any currency could challenge the dollar, it’s “the currency of the country that makes everything for the world” — the supply chain logic is real

• Belt and Road, soft power initiatives, and Trump’s antagonizing of trading partners all inadvertently help the RMB’s reserve currency ambitions

• The euro is ranked #2 but won’t displace the dollar; the yen won’t either; the RMB’s potential is huge

• CCP wants the RMB stable (not volatile like a yo-yo) to project confidence in China’s economic power; they indirectly manage it without directly striking a price

• Michael sides with Rupert on RMB strengthening — Rupert quips: “Firmly in the squirrel’s camp”

Blind Squirrel Macro paid subscriptions are paused until early June. I am unable to accept new paid subscribers until then. While I am on the road, I am continuing to publish the ‘Start the Week’ note on Sundays. Please get in touch if you recently subscribed to the letter and are interested in a trial / review of the archive.

This week’s ‘Start the Week’ was sent via email at 4.30am EST on Monday 25th May if you want to search your inbox. It will be put back in the online archive here once subscription billing is un-paused in June.

If you had not had a chance to look at the new SpaceX piece I put out on Sunday - I think this IPO is the biggest risk factor facing US equity markets this quarter. You need to understand the potential flow of funds impact even if you have no interest in investing in Elon’s rockets.

At the moment, BUSHY™ is underweight risk assets out of choice. The only risk that the beta portfolio has added since the start of the war has been in agricultural commodities and energy equities.

As such, the portfolio is always destined to underperform in market conditions like we saw last week - with energy producers being the weakest US equity sector.

A Quick Note on Crude

I am writing this on Monday afternoon Asia time (c.4am EST). Sunday night futures appear to be paying tepid lip service to the ‘Iran deal proximity’ being sold over the Memorial Day weekend. Equities are well bid and $4-5 per barrel has been shaved off front month Brent and WTI crude prices (about half of what was suggested by Hyperliquid markets over the weekend).

The current headlines out of Washington and the Middle East are not exactly screaming FINAL RESOULUTION! at me but I am prepared to at least concede that we are closer to the end of the war than the beginning.

Read more

In Episode 54, Benny & The Squirrel run though our latest report on the SpaceX IPO which was published on Sunday morning. It is not behind a paywall so please share it far and wide.

Share

The show was recorded at 6.45pm EST on Sunday 24th May.

I am not able to publish paywalled commentary on a billing pause - so for the next 3 weeks, enjoy the show notes for free! No pod summary this week (it’s all in the note) but the show slides and audio edition are downloadable below.

Today’s Slides

Pdf Of Slides

2.79MB ∙ PDF file

Download

Download

Audio only version (for paid subscribers)

Share

Subscribe now

The ‘Start the Week’ note will be out for existing paid subscribers on Sunday night or Monday morning EST. Please get in touch if you recently subscribed to my work and are interested in a trial / review of the archive.

The Physics of SpaceX

Well, the SpaceX S-1 finally dropped this week. As you know, the 🐿️ (together with podcasting partner Benny) have been following this one very closely. Even if you (like this rodent) have no intention of placing an order in the IPO, all investors and traders have no choice but to pay attention to the circus.

But right now the ducks are quacking - Polymarket odds are pricing a 74% probability of an end of listing day market capitalization for SpaceX in excess of $2trn.

And if the SpaceX common stock is insufficiently racy for you, the ETF Sponsor community has already filed some completely degenerate alternatives for you. I imagine that listed weekly options will be also on the menu within days of listing.

Once you have browsed through the glossy rocket photos and voluminous risk factors of the SpaceX S-1, you soon stumble upon the document’s ‘money shot’ - almost 80% of SpaceX’s TAM has nothing to do with rockets or satellite broadband subscriptions!

Most of “the largest total addressable market in human history” is AI Enterprise Applications. The largest estimates of AI Enterprise Applications 2030 TAMs that I have been able to find online range from $50-300 billion. I guess that the inclusion of Mars more than makes up for the exclusion of China and Russia…

Close your Eyes and Ears

You are currently being submitted to a continuous barrage of propaganda. You must tune out CNBC BubbleVision (“Rockets!, Mars!, wow!”)! SpaceX has also co-opted the entirety of Wall Street. The assembled syndicate below is probably looking at a fee pool of $850m, assuming a (well below market and yet to be confirmed) underwriting commission of 1%.

That number buys a ton of cheerleading (top tip - keep an eye out on Jefferies - they have a highly-ranked and now un-gagged Aerospace & Defense analyst and were not included as an underwriter).

Next, please please please ignore these 👇clowns - I have engaged (🙏 Jason) a full-time stunt double to listen to (and summarize the bullshit contained therein) the insufferable All-In podcast each week.

Below sets out the abbreviated SpaceX fundraising history. Elementary grade math will tell you that the basis of majority of the SpaceX cap table is in the trade at less than 5% of the current ‘price talk’ valuation.

These investors - who have collectively put in less than $11bn of equity (versus $2trn of proposed market cap) into SpaceX - will be all over the TV and podcast circuit over the next few weeks encouraging you to embrace their bags.

Does the layout math math?

I am revisiting my SpaceX IPO layout model from April.

The index committees have - as Benny and I feared - also been co-opted by ‘Team Elon’. SpaceX will enter every major benchmark index at the earliest possible opportunity. Nasdaq - with its 5x free float multiplier rule - is particularly egregious. But does it really make a difference?

On current price talk, let’s assume (a very long putt) the following. Musk manages to:

• capture 100% ($30bn) of the captive index demand for the offering (more on the mechanics of this below)

• enfranchise 50% of the demand from active managers that ‘closet benchmark’ the major indices

• find a home for 30% of the deal ($26bn) with retail investors - a staggering absolute number to raise from this cohort (I remember the days when a 5% allocation to retail was considered aggressive)

Assume all of the above and Musk still needs to find another $16.5bn of demand just to get the deal to 1x covered.

Have a play with the numbers yourself 👇:

SpaceX IPO Layout Model

12.5KB ∙ XLSX file

Download

Download

Even if you toggle for the IPO valuation and the ‘compliance’ of closet benchmarkers Musk’s bankers are still left short of allocable demand to the tune of 40-50% of deal size.

The Hedge Fund Passive Warehouse Trade

Let’s return to that $30bn of demand from the passive index funds. Guess what? The likes of Vanguard, State Street, Blackrock and even Invesco (asset of manager of ) are not putting orders into the IPO order book. They cannot buy SpaceX until the day that the company is admitted into the various indices they track.

The idea is that hedge funds will step in to act as temporary liquidity providers, anticipating that they can sell the stock at a premium to the index buyers ( et al ) shortly after listing.

In theory, the execution of this strategy would look like this:

1. Hedge funds bid for large allocations in the IPO order book. To fund these massive purchases without blowing up their prime brokers’ balance sheets (under Basel III) capital requirements, they will lean on synthetic leverage via swaps.

2. Because warehousing billions of dollars of SpaceX stock exposes the hedge funds to broader market drawdowns, they will likely also need to hedge (via S&P 500 or Nasdaq futures).

3. As the fast-tracked index inclusion dates arrive, the funds will sell their warehoused shares to the passive funds - hopefully at a markup.

The problem is that if these hedge funds misjudge the timing or if passive demand is weaker than modeled, they could be caught holding massive, highly leveraged positions in an illiquid market. It also does not help when most of your peers have the same trade on.

Top-of-Book Liquidity Constraints

Then there is a ‘making room’ problem that faces the broader market.

As passive investors and closet benchmarkers mechanically sell roughly $44 billion of existing holdings to fund their SpaceX bids, they will execute against a market where “top-of-book” liquidity (the ability to transfer risk quickly at the current bid/ask spread) is severely depleted.

These days, investors can transact far less capital ‘at the touch’ without moving prices. Large, directional and simultaneous sell orders across passive players will shred through this thin liquidity. Traders will be forced into deeper layers of the order book where liquidity is even thinner. Wider bid-ask spreads, increased transaction costs, and heightened volatility comes to the broader market.

At the single stock level - even for the giants of the Mag7 - top of book liquidity is even thinner as well as heavily fragmented across dozens of different exchanges and dark pools. During periods of stress, algorithmic market makers will likely pull their quotes - individual name liquidity could evaporate.

On the face of it, $44bn does not sound like a great of money. However, students of friend of the podcast will know that there is a price impact multiplier effect to consider. According to his most recent work, this amount of selling could feel $1 trillion of price impact. Do check out our conversation from January.

And it’s not just Day 1

Assuming that the $86.25bn gets successfully allocated, the market absorption is still not done. A wall of supply is coming from those insiders that bought SpaceX at a fraction of the price.

The S-1 makes clear that Musk (approximately 42% economic interest - with 85% of voting power via super-voting Class B shares) is subject to a normal lock-up but a lot of SpaceX insider stock joins the free float remarkably quickly.

“Early Release Eligible Shares” have not been quantified in the draft S-1 but the lock-up schedule and triggers are set out below and I have seen estimates that 80-90% of the non-Founder shares will be freely tradeable by November:

Will it Even Happen?

Back in April, Benny and I were skeptical whether the IPO would happen at all. We were wary of the obvious trade which would have been to short Tesla in the run-up to the SpaceX IPO as the Musk acolytes sold their shares to fund the switch into more direct play on the Muskonomy.

The eventual combination of SpaceX with Tesla seems inevitable. What if the challenge of finding a home for $86.25bn of stock laid out above turns out to be beyond the skills of the greatest stock promoter in human history?

The obvious ‘Plan B’ would be to reverse SpaceX into Musk’s existing listed vehicle. RIP the Tesla shorts if that happened! Back in early April the odds of a merger announcement by June 30th were priced at 5%. They are now at 1%.

Indulge the 🐿️ in a bit of wild speculation. Many column inches last week were devoted to the appointment of Goldman Sachs as ‘lead left’ on the IPO. This included a fun piece from my pal Craig Coben in the FT (linked below). Craig, a former Global Head of ECM at BAML, will be joining Benny and me on the podcast in early June.

All it took for GS to displace Musk’s traditional bankers at Morgan Stanley apparently was for David ‘DJ Sol’ Solomon to slide in to Musk’s DM’s on Twitter. Really?

The 🐿️ had a front row seat to the intense rivalry in Equity Capital Markets between GS and MS. It was a blood match. To lose any deal was bad enough. To lose it to Morgan Stanley was a cardinal sin!

MS banker Michael Grimes staked his reputation (and a great deal of firm capital) financing Musk’s buyout of Twitter. He even followed Musk into government for a while. Surely the top slot in the ultimate Muskonomy mandate was nailed on for Mr. Grimes?

What if Musk has done the same demand math as the 🐿️ and really wanted to keep Grimes focused on his job as lead M&A advisor on ‘Plan B’?

Perhaps this theory is a bit ‘tin foil hat’ but 100:1 odds on Polymarket for an emergency ‘shotgun’ Tesla/SpaceX marriage by the end of June do not look like a crazy flutter. Frankly probably better EV than buying SpaceX at a $1.75 trillion valuation. Just a thought. You have been warned.

Benny & The Squirrel Out!

Subscribe now

****Bonus Extras

I have released all of the work that Benny & I have done on this situation from behind the paywall. This includes the full April note and podcast as well as the latest full valuation model that we created with Manus and which agreed with our April assessment.

SpaceX Valuation Model

1.36MB ∙ XLSX file

Download

Benny and I prompted the Manus agent model to build this full valuation model for SpaceX. Without any input from us - using the S-1 filing only - the multi-agent model came up with a valuation that was pretty much in line with the Benny & The Squirrel model from early April. Have fun playing around with it.

Download

If you are a free subscriber and cannot wait for June to subscribe and get access to the paid content, send me a note and we will see what we can do about a trial.

The 🐿️ celebrated entering the back half of his 50s by taking viewer questions with Benny. The first in our series of regular shows.

Show was streamed live at 7pm EST on Thursday 21st May, 2026.

As explained before, Blind Squirrel Macro paid subscriptions are paused until early June. I am unable to accept new paid subscribers until then but please get in touch if you recently subscribed to the letter and are interested in a trial / review of the archive.

I am not able to publish paywalled commentary on a billing pause - so for the next 3 weeks, enjoy the show notes for free! No slides this week.

SpaceX IPO: Valuation Deep Dive

Key points:

• Manus AI model test: Benny fed the SpaceX S-1 to Manus (the Chinese AI, blocked from Meta acquisition), which produced a 15-tab model arriving at a similar conclusion to their own research — hard to get past $500–600B; stretching to $1T requires “letting the right tail in”. It pretty much matched our own model from 2 months ago.

• Anthropic/Colossus deal: Anthropic leasing the Memphis data center (~$1.3B/month) turns xAI into a “neocloud landlord.” The Colossus center cost ~$28B in infrastructure + ~$25B in chips (~$50B capital business) generating ~$15B in rental income — a good return on capital, but it’s a cap rate/REIT business, not an AI-multiple business

• The valuation bridge problem: At $7B EBITDA in 2025, a $2T valuation is simply not justifiable by fundamentals. Rupert: “I can’t see the bridge between $600B and $2T.” The concern is that retail and unsophisticated investors will apply “space AI multiples” to what is effectively REIT income

• Market structure risk: At $2T with a 5% float, you need to place ~$115B of paper in one day (the greenshoe alone would be the 10th largest IPO ever). Even assuming 100% of passive/closet indexer demand, you’re still 40% uncovered. This creates massive potential for market dislocation

SpaceX IPO Mechanics & How to Play It

• Is it a market top signal? — “A narrative I could get behind.” Rupert stays out of the way and watches for signs of market structure weakness from the primary raise itself

• How to buy big IPOs: Don’t buy on the open. The Facebook IPO anecdote — when a Smith Barney colleague offered Rupert “as much as you want,” that was the tell. High retail allocation = bad aftermarket. Wait 6–18 months for the “seasoning” reset to $200–300B before getting interested

• Insider dynamics: D1 and Coatue are already 20–25x baggers on SpaceX paper; they’ll show up in the book to “keep the party looking like it’s raging” but they’re really just managing average cost — not buying to hold

• Comparison to Bitcoin ETF launch: History rhymes — Grayscale → ETF → retail 401Ks put money in at the end. SpaceX is the same dynamic: VCs have pre-sold to family offices and IRAs, inverting the traditional IPO access model — “complete exit liquidity”

Were Walmart Earnings More Important Than NVIDIA?

• NVIDIA is more important for the market (it drives AI capex and S&P leverage to spending), while Walmart is more important for the economy (a bellwether for the US consumer)

• The “K-economy” split: AI capex is booming while most consumers are “completely screwed”

• Key concern: if Walmart starts doing negative comps, that’s a five-alarm fire — but currently ignored because the market is driven by the AI trade and Russell 2000 is “completely driven by AI capex”

• If faith in the AI trade breaks: XLP/defensives are already too expensive to offer real protection; rate-sensitive stuff (Rocket Mortgage, UWMC, Home Depot, Pool Corp) is the unloved contrarian long

Market Top Signals & Paul Tudor Jone

• PTJ is “having it both ways” — simultaneously most bullish and calling IPOs a top signal. Both hosts are wary of macro legends on TV: “there’s always an agenda”

• Rupert’s philosophy: he prefers confirming turns and pressing them; Benny prefers calling turns. Neither is superior — holding is actually the hardest skill

• Technical tell for a top: when a market stops making new highs and goes flat — that signals distribution. Key indicator: policy change (Fed, political support for the market disappearing) or 10Y Treasuries breaking 6%

How to Decide When to Sell a Winner

• Rupert’s framework: Enter with a predefined stop but target a “fair multiple.” Once you hit it, redo the work. If the thesis still holds, become a pure trend follower with a trailing stop. Never try to flip to short (that’s Benny’s game)

• Benny’s framework: Think of equities as options by delta. If a 5-delta stock moves to 25 delta, start taking some off. Take your initial capital off the table (”play with house money”). Three approaches: take capital off, be systematic, or revise valuation (which is “arrogant but has merit”)

• Covered calls debate: Rupert used them on Rheinmetall last year at a 30% premium to elevated levels — worked perfectly, banked premium, didn’t get called away. Benny more skeptical: “If you think it’s late-stage parabola, just sell it or buy puts”

• The long-term compounding point: Dave’s chart showing 75% of the S&P 100-year CAGR is explained by the top and bottom 2% of observations — the cost of missing GE Vernova or NVIDIA-type runs is enormous

Turkey Sells Gold & Sovereign Reserve Dynamics

• Turkey sold 58 tons of gold and all its US Treasuries — forced selling to fund government obligations with one of the world’s weakest currencies

• Broader point: sovereign buyers of gold and reserve assets face “enormous bills to pay” — this is a key driver of the gold pullback

• Gold has been “detached from real rates since the Russia-Ukraine war.” The real driver now is PBOC/Chinese & Indian retail demand

• Gold inversely correlated to oil — uncomfortable trade right now, but “the absolute best buys are the most uncomfortable”

Gold, Oil, SOFR & Bonds Positioning

• Rupert’s gold allocation: ~6.5% of portfolio, sitting on hands all year. Bought platinum/palladium last week (”probably a silly trade but sticking with it”). Getting close to topping up gold

• Benny’s SOFR pain: lost money on SOFR calls (front-end rates trade) this year. Pain partially offset by a levered short/long bond position

• Trend followers are now outright short the entire yield curve — same as late 2021 setup before the 2022 bond bear market

• Portfolio construction idea discussed: bonds + gold + energy as a recession hedge that also captures upside if oil spikes to $200-300 (at which point gold and bonds “go just like that”). Rupert prefers gold over bonds on the other side of the energy trade

Equities as Inflation Hedge & Bond Market Dynamics

• Are equities now the “safe haven asset”? Both hosts note that when bonds sell off, equities are holding — suggesting investors (consciously or not) believe equities keep up with inflation

• Historical bond bear market triggers reviewed: oil shock (1970s), taper tantrum (2013), COVID money printing (2022)

• Bonds disconnect from fundamentals is “funny to see equities holding” — but “it’s a great plan until you get punched in the face” (Mike Tyson)

Portfolio Construction: Beta vs. Active

• Rupert: 75–80% of capital is in a “beta portfolio” (runs with Bushy/VTTHX); remainder in single stock, macro, and thematic ideas discussed on the show

• Benny: Biggest position is Energy Transfer (ET) — held for 6 years, his “pipeline company” core long. Won’t sell due to tax consequences and strong distribution yield; described as “key infrastructure with unrealized optionality” (polar vortex quarters can generate $2B)

• Both hosts dislike Apple on valuation but acknowledge the “technofeudal” lock-in

• Key insight on passive flows: “more and more money is flowing to the top of the cap stack” — people don’t trust governments, so Apple (which didn’t succumb to the capex boom and keeps buying back stock) becomes a quasi-sovereign store of value

• Tax rule: Never make an investment decision solely on tax grounds. A delayed sale for tax reasons is “often an absolute curse.” Uzbekistan IPO shoutout — priced at $25, already trading at $28

Benny and I will be live on Substack and on YouTube at 7pm EST on Thursday 21st May.

We will be taking questions from readers and viewers. If you have one you want to ask, please drop it in the comme…

Read more

Show was recorded at 6.45pm EST on Sunday 17th May, 2026.

As explained last week, Blind Squirrel Macro paid subscriptions are paused until early June. I am unable to accept new paid subscribers until then but please get in touch if you recently subscribed to the letter and are interested in a trial / review of the archive.

I am not able to publish paywalled commentary on a billing pause - so for the next 3 weeks, enjoy the show notes for free!

Today’s Slide Deck

Today's Slides

3.29MB ∙ PDF file

Download

Download

Pod Summary

Trend Following Review

• Rupert explains his 15% NAV allocation spread across seven CTA/trend-following funds (6 ETFs, 1 mutual fund). These have been “the solid offensive midfield” of portfolio performance over the past year.

• Backstory: the allocation was made in response to inflation volatility breaking the traditional stock/bond hedging relationship.

• Full 15% allocation was made in May 2024 at “a spectacularly bad time,” right before a painful drawdown. Recovery from drawdown occurred in October 2025, with the allocation finally charging ahead in recent weeks.

Greenlight for more Energy (Minute 5)

• Key purpose of the CTA review was to determine where trend funds are positioned relative to 🐿️’s discretionary views — specifically in energy. We have green light.

• Plan is not to add more futures, but to load up on energy equity closed-end funds — a potential “double win” on both energy cash flows and NAV discount compression.

• Discussion of the sell-side convention of applying a zero multiple to “one-time” energy cash flows during price spikes.

• Energy Transfer (ET) used as example: $1–2B windfall cash flows from extreme weather events shouldn’t be valued at zero given their network edge.

• Modern energy companies are better capital allocators (returning via dividends/buybacks vs. drilling more rigs).

S&P Earnings Revisions (Minute 13)

• Energy sector has had a 48% upward earnings revision since 12/31/25, contributing +150bps to S&P 2026 EPS; Tech with a +14% revision contributed 500bps.

• Together, energy and tech account for essentially all of the 7% total S&P change. Both agree ‘26 and ‘27 energy estimates still “feel light” given the structural dynamics.

Volatility & Dispersion Markets (Minute 16)

• 30-day realized dispersion vs. 30-day implied correlation — current reading is a five-year outlier completely “off the map.”

• On 10-year lookback: also an extreme outlier, comparable to early 2023 (Feb 2023 junk rally), summer 2024 (NVIDIA blowoff into the yen-carry unwind), and the Rivian November 2021 top.

• Ned Davis Research correlation chart — Trailing 250-day median correlation of S&P 500 stocks to the index is near dot-com era lows and pre-Volmageddon levels. The high-beta/low-vol ratio (inverted axis) tracks this closely. Key prior extremes: pre-Liberation Day, summer 2024, August/September 2020, Rivian Nov 2021 top. Each preceded a sharp, fast correction — “not the end of the world, but crowding that set up something sharp.”

• Dispersion vs. put-call ratio — uncharted territory — Dispersion divided by put-call ratio is at 10-year highs (only comparable to COVID on a 1-month basis). Translation: the market is overwhelmingly positioned in calls, nobody owns downside protection, and “if it moves, it’s going to move a lot.”

• Skew data from Nomura — Call skew at the 100th percentile (1-year trailing); put skew at the zero percentile. 25-delta risk reversal “outrageously out of line” post-Liberation Day. A broker’s proprietary model flags a 75% probability of a 5%+ down-draft in the next 23 days. The “gas is in the tank” from levered ETFs, vol-control funds, and option dealers — a negative gamma environment where selling pressure accelerates as markets fall.

• Negative gamma dynamics explained — In a normal environment, dealers sell puts; here they are selling calls (”spot up, vol up”). This means that on a down move, dealers have to unwind calls AND sell the underlying simultaneously, creating sharp, fast left-tail moves. Historical example: NVIDIA June 8, 2024.

• Historical case study: Summer 2020 SoftBank “option whale” — The original dispersion/gamma squeeze was run by an ex-Deutsche Bank team working for SoftBank out of Dubai. They drove Apple up 23% in a few days on no news (August 4, 2020), before it collapsed 20% in two to three days.

• Tactical opportunity: what to watch — Recommend QQQ over SPY expression for put positioning; also EWY (Korea).

Blind Squirrel Macro paid subscriptions are paused until early June. I am unable to accept new paid subscribers until then. While I am on the road, I am continuing to publish the ‘Start the Week’ not…

Read more

In Episode 51, Benny & The Squirrel welcome Larry McDonald to the show. Larry is the Founder of Bear Traps Report and bestselling author of A Colossal Failure of Common Sense and How to Listen when M…

Read more

Show was recorded at 6pm EST on Sunday 10th May, 2026.

Read more

While the flagship weekend note will not be published while I am on the road, this week’s ‘Start the Week’ note was just sent to paid subscribers under separate cover.

In addition to the regular weekl…

Read more

In Episode 49, Benny & The Squirrel welcome to the show. Brent is President of Spectra Markets and author of AM/FX and the . Brent’s 3rd book, Alpha Trader 2: Outside the Box will be available on May 22nd. The first 2 were bangers - this one is more targeted towards professional traders.

The show was recorded at 6pm EST on Thursday 7th May.

Become a paid subscriber to get the audio edition, full show summary and the slide deck (it’s a fantastic one this week) - as well as the rest of the 🐿️’s tireless work!

Read more

Quick update: I am hitting the road this weekend for just over 3 weeks. I am working on an important project that will make it almost impossible to keep up with my usual research and writing schedule.

As such, I am pausing subscription billing for this period. Monthly and annual subscriptions will automatically be extended for the time that I am away. You do not need to take any action. I will also not be able to accept any new paid subscribers or process any gift subscriptions during this period.

During this time, I will continue to publish an expanded ‘Start the Week’ note for paid subscribers. This will contain the usual BUSHY™ portfolio review, Acorn updates as well as my thoughts on prevailing markets. This will hit inboxes at 7am EST on Sundays, starting this weekend.

I will also (I promise) be ever present in The Drey - our Discord server (several hundred smart folk sharing stock and theme ideas every day with oversight - and snark - from the 🐿️).

Episodes of Benny & The Squirrel will continue to drop on Sunday and Thursday evenings (EST). The show on the evening of Thursday 21st May will be a LIVE show during which Ben and I will be taking questions from viewers - I hope to see as many of you as possible in the chat room on YouTube!

Read more

Podcast now back to its normal ‘Monday at the close’ timing cadence. This week we cover the 🐿️’s new concentrated China equity portfolio.

The Weekend Note:

Paid subscribers who would like to listen to these pods on their favorite podcast app (instead of via a browser or the Substack app) can set this up very easily. Click this link while on the device you use to listen to podcasts.

Read more