Benny & The Squirrel

Benny & The Squirrel

Benny & The Squirrel is a twice weekly show about finance and markets. The show notes for members are hosted on Substack. www.bennysquirrel.tv

  1. 1d ago

    Volatility is a Gift!

    Recorded at 3pm EST on 19th November. For Episode 105, we are joined by Brad Lundy of Butler Hall Capital, a 20-year long/short veteran who started as a floor options trader and now hunts cash-generative businesses in the $1 to $25 billion market-cap no-man’s-land. Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. Show Notes How Butler Hall Played AI * Firm style: private equity approach to public markets; buys cash-generative businesses at an inflection toward accelerated growth, mostly in old-economy sectors. * Early AI winners bought cheap: Vistra Energy at ~5x cash flow (AI power demand), Vertiv at ~8x earnings for a cooling business growing 50%+ per year. Effectively free options on AI TAM. * Capex trajectory: a couple of hundred billion in 2024 to a Morgan Stanley estimate of roughly $1.64 trillion in 2027. AI capex is now effectively the crux of the economy. * Current stance: neutral. Too tough to call; requires underwriting capex into 2029 and 2030. Key distinction: recurring revenue from installed data centers is preferable to pure capex-cycle exposure. * Funding mix has shifted from hyperscaler cash flow to private markets and debt, adding credit-market dependency. Cyclicals and Valuation (Minute 6) * Best industrials entries historically occur when multiples look high because earnings are at trough, not when they screen cheap. * NVIDIA: cheap-vs-expensive depends entirely on what AI capex does over the next decade. If earnings go up 100x, paying even 10x is risky. * Brad is bullish on AI for society, bearish-to-neutral on underwriting AI equities. * Employment impact: an entry-level analyst can now be replaced for about $1,000 a month, which has real value but raises questions about who trains the next generation of analysts. * Token costs are declining ~90% while pricing rises short term; not obviously sustainable. Benefit likely flows to society at large, not one dominant company owning the next internet. Decline of Active Management (Minute 9) * Post-2020s active managers left, moved to platforms or launched their own shops; the next generation is effectively absent. * Old regime: valuation anchored capital allocation, so a 20% drawdown in a cheap stock attracted buyers and self-corrected. * New regime: a cheap stock down 20% picks up negative momentum factors, index ejections, and platform selling, and can fall 70 to 90% before recovering. * Consequence: invest with a 5 to 10 year horizon; short-term price action can do almost anything. * The key determinant of success is now the ability to hold a position through pain (liability structure) - the horizon required is shorter than five years in practice but the principle stands. * Butler Hall book: 30 to 40 longs, roughly 100% long / 60% short. * Sizing driven by capital at risk, not expected return: targeting max ~150 bps portfolio loss per long, so a 50% downside name maxes out around 3%. * Result: low portfolio-level volatility despite frequent 50%+ single-stock drawdowns, which they treat as exploitable. The Cash-Return Mechanism (Minute 13) * 10% free cash flow yield used to be attractive; now 20% yields are findable, but cheapness alone is not enough. * Without buybacks or actual cash return to shareholders, there is no timing mechanism for self-correction; a double that takes five years is a poor return. * Focus: businesses inflecting from ~3% growth to 10 to 15%, plus accelerating cash return, to survive the volatility. Broken Earnings-Revision Correlation (Minute 15) * Last five months: AI/data-center names posted big beats and upward revisions, yet the trade broke down, inverting the usual correlation between revisions and price. * Rarer behavior that historically predates major unwinds (dot-com analogy). * Host counterpoint: sell-side numbers could simply be wrong. Brad agrees the market is telling you capex comes down, and if it does, AI-exposed names at ~50x pre-euphoria earnings with margins 2 to 3x normalized are precarious. Signs of Life: Housing-Exposed Stocks (Minute 17) * Small and mid caps are in a bear market, but housing-exposed names (renovation, furniture, pools) started rising on negative revisions from September. * Pool demand is ~35% below normalized levels (~60k installs vs a peak of 117k in COVID), so earnings are terrible, yet stocks bottomed. Suggests markets overshoot on both ends. * COVID-era parallel (Q3/Q4 2020): stay-at-home winners crushing earnings were not indicative of long-term value. Market now over-prices near-term earnings relative to long-run cash generation. Private Equity and Private Credit (Minute 23) * Private credit: structurally asymmetric (limited upside, full downside on impairments); expects subpar returns for a period but not the existential crisis feared 6 to 12 months ago. * Anthropic/SpaceX IPO proceeds likely recycle into private markets; legacy PE vintages may struggle to monetize at marks. * Decade-long lockups demand a meaningful return premium over publics to compensate for illiquidity. * Google, Meta, Amazon and peers have effectively compressed into one thematic AI bet, concentrating S&P risk in a single trade. * Goldman tracking: share of S&P 500 names inversely correlated to the index on a daily basis hit ~20% around Q2 and ~40% by September; never before have ~200 index names been negatively correlated. Echoes of 1999 to 2000. * Reading: it is possible to build genuinely uncorrelated portfolios today, and the anomaly is likely an omen. Bear Markets and Recession Risk (Minute 28) * Allocators likely need 2 to 3 years of subpar index returns before changing strategy; 2022’s drawdown reinforced buy-the-dip behavior. * Most market participants under ~41 have never seen a real bear market; recent declines (tariff day 2025, 2020) were violent but brief, conditioning aggressive dip-buying. * First regime where AI capex is effectively the entire driver of GDP growth, now equity- and debt-fueled. A fall from $1.6 trillion (2027) to ~$1 trillion (2029) could trigger a real recession given the economy-wide multiplier. * Scale illustration: SpaceX estimated ~$150 billion of 2027 cash burn to be debt financed, versus the panic a couple of billion of Tesla burn caused years ago. If that credit does not come through, the trickle effect is enormous. * Brad’s framing: not a bear, just highlighting the risk. Either the $1.6 trillion of capex produces returns or it does not; both paths are volatile, and he would rather buy quality at depressed multiples with macro-agnostic returns. Housing, Pools, and Consumer Staples (Minute 36) * Not dangerously early on housing-derivative names: many down 50%+ from peaks with good balance sheets; favor remodel exposure over new-build sensitivity. Historical pattern: prolonged underproduction is eventually followed by overshoot to the upside. * Staples: GLP-1s, private label penetration, and input inflation have crushed the sector, creating chances to buy under-earning, recurring businesses at 6 to 8x cash flow (not the XLP megacap staples). * Example long: a UK prepared-foods company at ~10x earnings growing earnings 15 to 20%, roughly 25% salad-exposed, gaining share as supermarkets reallocate shelf space to prepared foods for higher sales per square foot. * Anecdote: a merger with the number two competitor made reported guidance look like a miss (consensus reflected both businesses; guidance reflected partial-year consolidation). Stock dropped ~20% on a non-fundamental optics issue, they added, and guidance was raised the following two quarters. * Small/mid cap lesson: 20%+ moves on non-fundamental noise are common and represent gifts if you can identify them. M&A as a Value Signal (Minute 44) * Takeout premiums are not a major driver of returns, but attempted or failed M&A is a strong signal of hidden value and a prompt for time allocation. * L’Occitane example: orphaned Hong Kong listing at half the valuation of global skincare peers; the controlling shareholder’s (later withdrawn) take-private bid flagged value, and the buried asset turned out to be… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.bennysquirrel.tv/subscribe

    Volatility is a Gift!
  2. 5d ago

    AI Post-‘Jexit’, Pre-$ANT

    Recorded at 6.30pm EST on Sunday 15th November. Another unmissable Sunday Show’ after a big week for bonds, AI, NVIDIA, the US dollar and crypto.We start with the violent reversal in the bond rally and why positioning, CPI and hawkish Fed commentary combined to produce one of the largest moves in two-year yields in more than two years.Then we turn to NVIDIA: the political fallout over advanced chips, Jensen Huang, earnings risk, customer disclosures, receivables and whether the market is already pricing in sell-side downgrades. Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. LLM Monitors Show Notes Bond selloff and rates positioning * The move in two-year yields is described as one of the largest in more than two years. Substantial remaining short interest suggests crowded positioning may have amplified the repricing. * The sell-off was initially attributed to hotter CPI and hawkish comments from Federal Reserve governor Christopher Waller. * Is it correct to say energy impacting CPI was a surprise? On top of that, energy prices and energy equities had recently been falling. Just markets fitting a narrative to price action? * Waller is influential but is not the Fed chair, so his comments alone should not have caused such a large move. Positioning is viewed as the more important driver. * Similar reaction to Warsh’s first major speech, when rates moved sharply despite the comments not being fundamentally important. One-sided positioning can make ordinary news produce outsized market moves. * The bond sell-off is characterized as a potential buying opportunity. The expectation is that both rates and previously crowded AI trades will retest their former levels. SOFR curve and asymmetric trades (Minute 5) * The chart shows implied federal funds rates, term premium and forward rates. Term premium is estimated at roughly two to three basis points per quarter, meaning the break-even rate does not exactly equal expected federal funds. * SOFR futures are difficult to interpret because the contract already represents a forward rate, while the implied break-even is effectively a forward calculated from another forward. * Six-month call butterflies show that rate-hike expectations are concentrated over the following year rather than evenly distributed along the curve. * Peak policy expectations are around the summer or fall of 2027, followed by market pricing for rates to move lower. * A CME FedWatch chart for the November 2027 meeting reinforces that substantial tightening is already priced in. * The “action in the butterfly” refers to six-month call butterflies on individual SOFR contracts, with activity concentrated inside roughly one year and diminishing beyond September. * The hosts consider how to trade a scenario in which AI stocks roll over, the economy weakens and rates fall materially. * Possible structures include a 97/98 December SOFR call spread or a ratio backspread, such as selling the 95 calls while buying multiple 96.50 or 97 calls. * The bearish macro premise is that AI investment has become disproportionately important to the aggregate economy, while the median household and business remain under pressure. * If AI investment or equity performance disappoints, the broader economy could “catch down,” creating scope for a substantial rally in rates. * December 2027 is preferred for option structures because it offers more convexity for each dollar of premium. December 2028 is preferred for a straight futures position. NVIDIA, earnings and factor divergence (Minute 13) * Shareholder sentiment may not determine the outcome because NVIDIA’s most important audience is now Congress rather than its cap table. * Political scrutiny is expected to intensify over NVIDIA’s possible involvement in advanced chips reaching China. * Attention turns to earnings after Tuesday’s close. The hosts question who can credibly sign the Sarbanes-Oxley certifications given uncertainty around customer segmentation and historical revenue. * A possible restatement of historical earnings is raised if geographic or customer classifications were misrepresented. * Drawing on Paul Sankey’s analysis, the hosts suggest the buy side generally anticipates earnings revisions before the sell side. NVIDIA’s price action may therefore signal that consensus estimates will be reduced. * Semiconductors begin diverging from the broader momentum factor. The hosts question whether this is merely a positioning unwind or a sign that the semiconductor trade has already been washed out. * Momentum stocks are still being liquidated, but buyers appear to be returning selectively to semiconductors. Dollar, crypto and AI financing (Minute 19) * The focus shifts to the US dollar, which has broken out of a well-defined range extending back to “Liberation Day.” DXY is estimated to have room toward 105 before meeting meaningful resistance. * Interest-rate differentials appear to be driving the dollar again, making it a potentially useful risk indicator. * Strong-dollar conditions are connected to recent weakness in Bitcoin and, to a lesser extent, gold. * Bitcoin perpetual-futures funding turned negative over the weekend. This means shorts must pay funding to remain short and signals pronounced downside pressure. * One explanation for Bitcoin weakness is that AI is “eating crypto.” Bitcoin miners and early holders are reportedly selling coins to finance computing infrastructure. * Are they selling Bitcoin and other assets to make room for the forthcoming Anthropic IPO? * The chart compares an Anthropic perpetual contract on Polymarket with NASDAQ futures. Unlike the NASDAQ, the Anthropic contract failed to bounce during the week. * Despite that weakness, the IPO is reportedly about six times covered. 🐿️ is skeptical of the valuation and suggests using listed options as “schmuck insurance” as and when contracts become available after listing. AIDC Politics (Minute 24) * Political resistance to data centers is then discussed. Although data centers became contentious during the midterms, one host argues that they ranked relatively low among voters’ main concerns - see charts below from Bruce Mehlman. * Broader AI investment is viewed as too important to national economic performance for political opposition to stop the build-out in the near term. * More significant risks include fallout from the NVIDIA chip controversy and the debt market’s capacity to finance the remaining AI infrastructure. * Financing is expected to remain manageable next year, but questions are likely to intensify about the following year. Marginal AI infrastructure companies are most exposed to tightening credit. Consumer staples (Minute 28) * Recent rotation into consumer staples does not automatically make them defensive. Some trade near six times forward sales and carry substantial valuation and factor risk. * The counterargument is that institutional investors remain underweight staples and are returning to cash-flow stability. * Stable businesses can trade like bond proxies because their cash flows are more predictable. * Examples include Clorox, Hefty bags and McDonald’s, where demand remains relatively resilient even when economic conditions deteriorate. * Most institutional buyers own staples because they require equity-market exposure combined with defensive characteristics. Discretionary investors do not necessarily need to chase those names outside a tactical trade. * Consumer staples had a strong tactical run after rate-sensitive and growth stocks sold off in September. * XLP rose about 12% during October, making a short-QQQ and long-XLP relative-value trade particularly successful for that month - most probably a tactical rotation rather than the fact that staples offer attractive long-term returns at current valuations. Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. If you act on anything provided in this newsletter, you agree to the terms in this disclaimer. Everythi… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.bennysquirrel.tv/subscribe

  3. Nov 12

    The Convexity Maven

    Recorded at 5.30pm EST on Wednesday 11th November. For Episode 103, we are joined by fixed income legend Harley Bassman, creator of the MOVE index and the man behind The Convexity Maven. Harley explains why this rate cycle is about fiscal malfeasance rather than inflation, why TIPS at 3% real are the wrong number, why re-IPOing Fannie and Freddie is a moral hazard disaster, and why things get “very jangly” if Treasury yields approach 6% again. Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. Show Notes How to Read The MOVE (index) * Harley created the MOVE index (the “VIX for bonds”); first published in 1994, with internal data extending back to 1988/89 * Historical MOVE range was 80 to 120; more recently 60 to 100, or 70 to 110 * The MOVE is a real number, not a rebased index: a reading of 100 implies a one standard deviation (68%) chance rates are plus or minus 100bp a year from now * Divide by sixteen (technically by 15.9, the square root of 252 trading days) to get the daily implied move: a MOVE of 100 means roughly 6bp per day; a VIX of 16 implies a 1% daily S&P move * “All of vol trading is ninth grade statistics dressed up with Greek letters to justify higher fees” Why convexity is hard (Minute 4) * Volatility confuses most participants because it is non-linear and path dependent; a bond that travels 100 to 110 to 90 and ends at 101 made one point, but an option’s outcome depends on how you get there * Humans think in two dimensions; convexity is a three dimensional problem, and most people are neither paid nor trained to think that way * Leveraged ETFs paying 2x or 3x the daily move are a massive short convexity position few holders realize: up 10% then down 10% leaves you at 99, and repeated round trips grind the asset toward zero * This year has been a live lesson in path dependency, with rates walked up and down repeatedly What is really driving rates (Minute 7) * Harley argues the rate rise is not about inflation: ten year breakevens have sat near 2.35% for three years straight * The real driver is the fiscal situation: a roughly 6% deficit with 4.1% unemployment and solid GDP, which historically only happens in war or recession; a real recession could push the deficit toward 10% * The Treasury will borrow about two trillion dollars net this year; AI capex adds another 700 to 800 billion, meaning a handful of tech companies will borrow roughly a third of the federal deficit * That supply dynamic explains rising rates across the West, the US in particular * On Michael Green’s passive thesis: Harley is agnostic, noting active bond managers still beat the index, though he concedes supply one day equals demand * As a University of Chicago-trained monetarist, he holds the old rule that ten year yields should sit near nominal GDP: with nominal growth around 6%, a ten year at 5% is not crazy and 3% was the wrong number * He is a self-described inflationist: expects inflation near 3% for a long time, driven partly by millennial household formation, an echo (smaller in percentage terms) of the boomer “pig in the python” Fiscal politics and the inflation exit (Minute 14) * On the three variable problem (rates, currency, fiscal): you can only control two; Japan controlled rates and the yen went from 80 to 160; the US faces the same arithmetic * MMT was theoretically right (borrow until inflation appears, then tighten) but politically impossible: politicians are paid to get reelected on two year cycles, not to cut spending or raise taxes * The silent default will be inflation, a stealth tax on the middle class; run nominal growth above the debt rate and debt to GDP burns down, as after World War II The mean vs median economy (Minute 17) * Stocks are fine as long as unemployment stays below 5%, because millennial 401k money recycles into passive indices regardless of price * His advice to his own kids: “max the 401k, take the match, 70 to 80% in an S&P index fund, the rest in a five year IG fund, reinvest, call him in thirty years” * The post pandemic “jaws” between the S&P and consumer sentiment is simple mean versus median: in a village of ten where one person earns 500k and nine earn 10k, the average is 50k and the median is 10k; nationally the economy looks great, individually it feels terrible The Fed cycle and rate levels (Minute 22) * The Fed will keep hiking as a reults of hubris; Powell and Kevin Warsh both want to be Volcker, not Burns, so they will keep tightening until inflation is resolved. Harley thinks the Fed has a couple more hikes in it. * His levels: hard to see two year notes above 5%; ten year fair value is 4.35%, but he has no problem with tens trading up through 5% before falling back * At some point a government guaranteed yield beats stocks at a giant PE; a 30 year TIP at 3% real makes him ask why a pension or insurer owns equities at all * A 20% equity drawdown would scare people but not change the economy, since the top 10% own most of the market and their spending barely moves; hosts counter that RSU driven layoffs and upper quintile consumption are real transmission channels, which he partly concedes Mortgages and MBS (Minute 32) * He likes mortgage bonds now: think of a Fannie, Freddie, or Ginnie bond as akin to being long a ten year bond at 100 and short a three year call struck at 105 * Current pickup is an additional 100 to 110bp of spread versus roughly 50bp for Investment Grade credit * The bigger driver of MBS spreads is the yield curve, not rate or vol level: the flat, inverted curve (two year up far more than ten year over recent months) explains spreads near 110 over versus a 75 long run average GSE re-IPO and moral hazard (Minute 35) * He is firmly against re-IPO-ing Fannie and Freddie: privatizing gains for shareholders while socializing losses is classic moral hazard * Housing is 16 to 17% of the US economy; the government will always backstop the GSEs * He hints the investor base in GSE common stock explains the political push (“you kind of know what is going on here”) * What killed the GSEs pre GFC was running as government sponsored hedge funds, buying their own MBS, hedging with futures, swaps, and options to earn a spread, with management cashing options before the blowup. To recreate those incentives would be “nuts” * Why mortgage spreads move with the MOVE: MBS is short a call option, and option prices rise with implied vol; if the short call goes from four points to five, the bond drops from 96 to 95 * The MOVE is an imperfect hedge vector for MBS: it measures one month vol while mortgages price off three to four year expiry options * Treasury vol no longer trades; swaption vol is the market since the early 2000s (Fannie and Freddie drove the shift): global liquidity, no repo or borrow costs, effectively infinite issue size The Treasury basis trade and CTD risk (Minute 44) * The basis trade (long cash bonds, short futures) is now carrying water for the US government, absorbing supply; pension funds are the natural futures buyers, wanting thirty year duration off balance sheet while earning credit spread from five year corporates * The fragility: in volatile markets, cash and futures decouple, mark to market losses force risk manager stop outs, as seen when Treasury injected liquidity a few years back * CTD (cheapest to deliver) risk: each futures contract has a basket of deliverable bonds with conversion factors, like grades of wheat in Chicago; a rate shift can flip the deliverable from an eleven year to a nineteen year bond within weeks * If rates approach 6% again, “it will get very jangly” as futures holders stop knowing what they own; no defaults, just more volatility * Basis trade capacity is vol governed: risk models cut the trade exactly when the market needs absorption most; last cycle the funds actually won because the future extended duration faster than the cash bond * On persistently negative swap spreads since the GFC: Bassman says they reflect balance sheet costs and accounting constraints, not stupidity, and pushes back on th… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.bennysquirrel.tv/subscribe

  4. Nov 9

    What Now?

    Recorded at 6.10pm EST on Sunday 8th November. The dust is still settling after Thursday’s washout. NVIDIA has sliced through its 200-day moving average and is getting marked down in the perps market going into the weekend. The Anthropic IPO price talk is drifting toward our number (sort of). Chinese stocks have ripped 12% against the S&P in seven days on no news, right as the earnings revision streak finally rolls over. So what now? Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. Show Notes NVIDIA, The Highlander, has a tough week ahead * NVIDIA is being marked down roughly 2.5 percent in the perps market (Hyperliquid) on weekend China Honeypot spy scandal news. * The stock sliced through its 200-day moving average at the end of the prior week and barely recovered on Friday, setting up a potentially tough week for the market leader. * Core question: can the broader market rally properly if NVDA is not participating? “What Now” After the Puke (Minute 3) * The 🐿️ took his hedges off into Thursday’s washout and is now thinking about what conditions warrant putting them back on. * Revisiting the CoreWeave CDS blowout (”the snap of the jaws”) as the tell that blew out the “situational awareness” fund manager for a second time in a year. * Looking back to the late May and June period when CoreWeave CDS was at the tights and the Nasdaq at the highs; the “echo boom” gave a false sense of confidence before everything caught down. Anthropic IPO Repricing (Minute 5) * The market needs to be stronger for the Anthropic IPO to get done convincingly; hedges likely go back on once Anthropic’s pricing is in the rearview mirror. * Price talk has moved toward his estimate, but a 30 to 40 percent range signals the underwriters have no real idea where it prices, and they have had pushback. * Precedent-setting moment: a pre-IPO trillion dollar repricing is unprecedented; his constant view was it was always going out at a trillion (even at the June peak of AI euphoria). * Valuation caveat: fundamental valuation work is deeply out of vogue, so all of this is guesswork. Does the Index Level Matter? (Minute 6) * Debate over Mike Green’s view that index price has no macro value and is “the dumbest macro product on earth.” * Counterpoint: in a hyper-financialized economy, consumption is a derivative of where the Nasdaq and S&P trade, so the index is a valid macro indicator via the wealth effect. * The index level drives human behavior and matters for macro, but day-to-day index readings contain little information; the violence beneath the surface is the true harbinger. Historical markers cited: late January 2018, September 2018, February 18 2020, late 2021, and right before Liberation Day (February 18 to 19). Positioning: GS Prime Slides (Minute 9) * Gross leverage is still high (around 310%) while net exposure is down; the net decline is not genuine deleveraging. * Thursday was clearly a machine selling day, with many humans already out of the way. * CTA equity positioning remains elevated; by no means has all that systematic selling happened yet. Earnings Season Reality Check (Minute 10) * The bull narrative of relentlessly rising forward earnings has been “shot in the head”: consensus S&P EPS missed at the index level, while numerous stocks with good prints fell 5 to 15 percent. * Aggregate year-over-year growth (roughly 30 percent) hides a weak median participant; the average company is effectively in a recession and a bear market. * The market is repricing how it discounts 2028 to 2030 AI economics while 2027 capex probably remains safe. * Rate-sensitive orphaned assets have rallied since early October, but that does not make put them in a bull market; the only way forward if the AI trade fails is much lower rates. China: Confucian Confusion (Minute 15) * New piece published over weekend covering whether another bull leg in Chinese equities is possible. * FXI vs SPY has snapped back roughly 12 percent in seven days on no news * China proxy pairs (for example Li Ning as “China’s Nike”) have outperformed their US equivalents * Critically, this happened against aggressive negative earnings revisions: the streak of upward CSI 300 revisions running since summer 2024 has now turned down abruptly. * The summer 2024 positive inflection in earnings was the original entry signal. * Core framework: stocks rise where governments are spending and earnings are being revised upward. * US angle: fiscal support is effectively underwritten by the AI capex schedule; a slowing rate of change in that capex is as impactful as a government cutting fiscal spend. * Sector-level revision breakdown for CSI 300 and MSCI China: analysts are fading the consumption recovery hard and pushing it out, yet the market is fading those cuts in price. * Tech hardware and materials show the inverse: stocks getting drilled while analysts still raise numbers, a massive multiple compression dynamic. * Credit impulse has turned up again and looks like a buy. China Positioning and FXI Shorts (Minute 21) * FXI (4 billion AUM, roughly 800 million daily volume) is the standard tool allocators use to hedge China within broader EM mandates without selling the underlying. * Historically, peaks in FXI short interest have preceded rallies in Chinese stocks; shorts look crowded. * China remains a stock pickers market, not a place to grab beta indiscriminately, and not until actual stimulative policy arrives. Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. If you act on anything provided in this newsletter, you agree to the terms in this disclaimer. Everything in this newsletter is for educational and entertainment purposes only and NOT investment advice. Nothing in this newsletter is an offer to sell or to buy any security. The author is not responsible for any financial loss you may incur by acting on any information provided in this newsletter. Before making any investment decisions, talk to a financial advisor. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.bennysquirrel.tv/subscribe

  5. Nov 6

    The Ownership Dividend

    For Episode 101 of ⁨the show, we were joined by Daniel Peris, head of the Income and Value group at Federated Hermes. Dan explains why cash in hand beats contingent gains, why total return is a broken measurement, how the S&P 500 stopped measuring the market and started driving it, and what a historian and former Sovietologist sees in the long decline of Russia. Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. Show Summary Dan’s dividend philosophy and its origins * He came to dividend investing through a historian’s critique of total return as the dominant, quasi-statutory measurement framework. * Transitioned from academia (Sovietology , later modern Russian history) into financial research in the early 1990s, then migrated toward dividend-oriented portfolio management. * His core lens is agency costs: as a minority shareholder with no control over CEOs or boards, the dividend is the tangible manifestation of business ownership. * Focus on cash flows from large companies he does not control is an almost natural historical framing, given recurring human behavioral patterns in markets. Historical basis: dividends as the norm (Minute 4) * Until the 1980s, serious issuers came to market with a cash flow attached; stocks without dividends were considered speculative or defective. * The capitalization ladder was bonds (fixed coupon), preferreds (fixed coupon with priority), and common equity (riskiest, but with a growing dividend). There was effectively no “no dividend” option for credible companies. * Nineteenth-century railroad stocks are the archetype: entire capital structures were built around the commitment to pay. The business ownership relationship is thousands of years old, and stock-market dividends roughly 200 to 400 years old. * Dividend investing has fallen out of fashion in the U.S. over about forty years for reasons he acknowledges, but he keeps a plain vanilla business-ownership approach for conservative, often business-owning clients. Buybacks versus dividends (Minute 7) * Peris reframes the Kenneth French quip: buybacks divide people who understand finance from people who understand business ownership through the stock market. * A buyback is a liquidity event for share sellers; a dividend is a business outcome for shareholders. Fundamentally different phenomena that have been “smushed together” for convenience. * He objects to companies presenting “dividends plus buybacks” as a combined cash return (the “buyback yield” or “shareholder yield” framing). * Dan cites analysis showing announced buybacks are often far less effective than claimed: investment bank frictions, poor execution prices, and share counts that shrink much less than advertised. * Exceptions exist: large dividend payers like big energy and phone companies that genuinely and persistently reduce share count, allowing per-share dividend growth without runaway aggregate payouts. Tax treatment debate (Minute 12) * The elephant in the room: dividends create taxable leakage versus buybacks. Peris pleads “guilty as charged” but refuses to subordinate investment policy to tax minimization. * Taxes are a sign of success; the current regime is historically benign since qualified dividends and long-term capital gains have had equalized rates for about two decades, whereas parts of the twentieth century saw dividends taxed far more heavily. * Some investors organize their entire financial lives around tax avoidance out of anger; that is a business model for some advisers, but not his approach. June 2026 selloff days like this one show that capital-gains-only exposures carry their own risk. Critique of total return as a measurement (Minute 14) * His method: ask where the rules came from, whether they were fit for purpose then, and whether they still are. Much of modern investment practice (MPT, the S&P 500, total return) is a product of the 1950s to 1970s. * History of total return: academic research in 1924, referenced by Ben Graham (1934), T. Rowe Price and Babson’s, Lorie and Fisher (1964), became standard on brokerage materials in the 1960s through 1980s, and was only permitted on documents by the SEC in 1988. It is roughly fifty years old, not a timeless principle. * Total return is fundamentally a product of daily pricing technology. Without daily prices, the concept would not exist; in the nineteenth century, income streams were the only realistic information investors had. * The S&P 500’s introduction in 1957 as a daily index was a genuine technological marvel, and combining daily price change with dividend yield was an innovative, useful shortcut at the time. * The problem: the shortcut conjoins two unrelated phenomena, a market phenomenon (daily price change) and a business phenomenon (a company distributing a share of profits). Cash returns versus contingent returns (Minute 18) * Dividends and realized gains are real cash returns; unrealized gains are contingent returns you cannot consume, pay rent with, or reliably borrow much against. * Economics holds that the ultimate point of investment growth is funding consumption, and only realized cash can do that. * Acknowledges harvested capital gains count as “real” once sold; the issue is mixing unrealized price moves with actual cash in a single number, which was defensible in the dividend-rich 1960s and 1970s but has become less helpful. * Brokerage statements routinely show unrealized gains and losses prominently while hiding aggregate income behind toggles, and yield reporting varies. * He cites the 2022 paper “Reconsidering Returns” by Hartzmark (University of Chicago) and Solomon (Boston College), which found brokerage statements handle this badly and that few even attempt to show total return properly. * True portfolio total return is genuinely hard to compute because of unrelated cash flows (401(k) contributions, withdrawals, fees, taxes) and the arithmetic versus geometric linking problem; he jokes the word “geometric” kills podcast audiences. * His reform proposal: statements should distinguish cash returns from contingent returns for the current period and over time. Siegel’s Stocks for the Long Run (Minute 24) * A cottage industry of high-end academic debunking of Siegel’s long-run return data now exists (Peris notes Edward McQuarrie at Santa Cruz and financial historians reworking nineteenth-century data), and Siegel has responded responsibly. * Key challenge: the roughly 10 percent historical equity return implicitly assumes free, frictionless dividend reinvestment, which drives about 90 percent of actual account returns, yet reinvestment was historically hard and expensive, so realized returns were likely lower. * Modern improvements (no broker charges, near-instant reinvestment, fractional shares) changed this. Any reinvestment delay creates drag, so frequency matters. The Global Context (Minute 27) * Large, mature, highly profitable companies paying no or de minimis dividends is a uniquely U.S. phenomenon; European and emerging markets, and historically dominant markets like pre-1917 Russia and pre-1940s Germany, were “all about the coupon.” * The current U.S. dividend-free culture of the last thirty to forty years is anomalous both against U.S. history and against other developed markets. He concedes the era’s productivity success is real, but his research zigs where the market zags. * On dual-class and non-voting share structures: his fund avoids them; agency costs become too high given passive investors elect boards and “imperial CEOs.” Missing Tesla, Broadcom, or similar names is an acceptable cost; there are other fish in the sea. Then versus now: the 1960s profit pool (Minute 31) * In the early 1960s, despite being an industrial, high-capex economy, companies reliably shared profits with minority shareholders via dividends. * Today capex as a share of sales is lower yet profitability is far higher, and there is minimal inclination to pay dividends; the money goes to buybacks and acquisitions. Two very different market worlds, and few divi… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.bennysquirrel.tv/subscribe

  6. Nov 2

    Make or Break Markets into Mid Terms

    Episode 100! Benny & The Squirrel hit the century mark!! Markets into the midterms, the Anthropic IPO, why Google and Amazon got hammered while Oracle ripped, and the one trade that has held the book together since June. Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. The LLM Monitor Show Summary Anthropic IPO Note * The podcast kicked off with a 5-minute run through of the 🐿️’s weekend note on the Anthropic IPO. Find it (no paywall) by clicking on the image below. Midterm Elections (Minute 5) * Brazil last week was a salutary lesson against treating political outcomes as foregone conclusions. * Pollsters have had a terrible decade; Democrats reportedly ahead in the Kansas Senate race, GOP more or less stopped funding North Carolina. * Structural wrinkle: Democrats need one extra seat because J.D. Vance holds the tie breaking vote as VP, and even then a one seat majority is fragile (the “Fetterman effect,” swing senators needing to be whipped or bought). * Polls cannot capture the sheer money in specific races: hundreds of millions from Musk on the Paxton versus Talarico Texas Senate Race; Musk’s total cycle spend is around half a billion dollars, making him the largest political donor in history, ahead of Soros. Market Internals: Hidden Bear Market (Minute 7) * Recap of the structural bear market call: roughly 60% of individual stocks were already in bear markets (down 20% from highs). * Two distinct drawdowns: September hammered out of favour, rate sensitive and GDP sensitive names; October broadened the selling. * September weakness partly mechanical: many mutual funds and the government have a fiscal year end on 30 September, so tax loss selling of losers is traditional, with those names then resold into October. * Triple C spreads show the biggest divergence “of the century” versus BB (the best of junk) and the high yield index average, though triple Cs are only a small sliver of the index, so no panic yet. * Is the credit story simply the continued selling of the most toxic remnants? * Ties into Mike Green’s passive dynamics: money comes in, spreads tighten, issuers keep hitting the bid; high yield still has capacity on a normalized basis. * IG is where hyperscaler TMT issuance has blown out spreads since last September; sector risk limits then force mechanical reallocation into other, sometimes terrible, credits at their best ever spreads relative to fundamentals. Machines vs Humans (Minute 10) * The S&P index is only about 3% off its high, but the trading underneath has been violent: rate sensitives smoked in September then rebounded; tech held in then dumped. * Nasdaq futures sit near the 200 day moving average for the first time in a while. * COT data (large and small spec length) shows the October drawdown barely dented positioning: the humans sold, the machines did not. * Sell side says CTAs have sold heavily, but the 🐿️, with 15% of NAV in trend strategies, sees no major equity index reduction in his own CTA holdings; vol target funds also remain very long because realized vol has remained depressed. * Thesis of Benny’s upcoming “casual barbell” essay: fewer humans matter at the margin than ever, so discretionary deleveraging is largely done and the systematic, mechanical players are the latent bag holders; if realized vol picks up, those are trapdoor sellers with no discretion. Primary Supply (Minute 13) * Anyone under 41 on the Street has never seen a genuine multi year bear market (2022 probably does not count). * New and untested dynamic: the net supply of primary equity paper. Giant caps doing $90 billion equity raises and straight to the top of the index IPOs were flagged in May and June when they turned bearish on the AI trade. * The question to ask: why are the richest companies and executives in the world taking your savings, and why move down the capital stack to equity this early? Implication: the IG market was tapped out unless capital ratios improved first. Hyperscaler Capex Reckoning (Minute 16) * The narrative has shifted to free cash flow, which we flagged (way too early) back in April and May. * Oracle, “the dog of the hyperscalers,” rallied massively when Ellison merely hinted at spending a little less; Google and Amazon got hammered on their prints. * Cloud growth is no longer the driver; investors now watch capex and cash burn; relief rallies now come from spending restraint, and the revenue side of the AI equation remains unanswered. Google: Search Erosion Arrives (18:30 to 22:10) * The “AI chatbots kill search” narrative from mid 2024 finally showed up in the numbers last quarter; those who bought that panic low did incredibly well. * Search erosion is coming home to roost alongside unyielding capex. * Both hosts pay for Gemini Pro and rate it highly (Nano Banana image generation is best in class), yet Google now front runs its own search links with AI summaries. * Framing from the NYC dinner with Mike and Hari: “search” is retrieval, “ask” is intelligence, and “ask” is coming for the search business. * Defence of Google: to stay in the retail LLM game it had no choice but to put its advertising cash cow at risk; unlike Kodak or newspapers, it is at least being proactive, but the fundamentals of its old utopia have changed. Politics Returns (Minute 22) * Blue wave and blue tsunami talk is everywhere; remember Brazil and wait for all the dice to roll before pre positioning for a Democratic sweep. * The September setup (long GDP sensitives, housing, bonds) to front run a sweep already worked; the October bond rally has made the trade much more coin flip like now. * The AI trade is such a large share of the market that political flow can dominate price action regardless of AI headlines. * Data centres are the sleeper political issue: Democrat opposition to manufacturing tax breaks traces to data centre backlash; many projects are being overrun and it has become bipartisanly unpopular remarkably fast. * Driver is the visible wealth divide: trillionaires, $20 to $25 cocktails, offensive New York rents; if you are not a winner in Menlo Park or Manhattan, why support what is creating this. The Equal Weight Trade and Correlation (Minute 25) * Anchor for the 🐿️’s BUSHY™ core beta book since early June: long equal weight S&P (RSP) versus the Nasdaq 100 (QQQ); ripped initially, pulled back hard in September, then almost made new highs and has held the book together. * One month implied correlation (black) and realized correlation (red) collapsing S&P High Beta / Low Vol mirrors RSP/QQQ; it is the same trade as pod shops grind correlation down. * Mechanism: whatever works keeps getting bid, defensives and orphans get no bid, then reprice violently; since bonds bottomed around 8 October, high beta versus low vol repriced hard * Caveat: this is not a new bull market in mediocre mid curve businesses; many were simply funding shorts being unwound, driven by market mechanics. Benny & The Squirrel is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. If you act on anything provided in this newsletter, you agree to the terms in this disclaimer. Everything in this newsletter is for educational and entertainment purposes only and NOT investment advice. Nothing in this newsletter is an offer to sell or to buy any security. The author is not responsible for any financial loss you may incur by acting on any information provided in this newsletter. Before making any investment decisions, talk to a financial advisor. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.bennysquirrel.tv/subscribe

  7. Oct 31

    AI Outcomes Have FAT Tails

    In Episode 99, Benny & The Squirrel are joined by Brandon Carl. Is AI headed for its shale boom moment? Brandon is a rare operator - with a decade on Wall Street in fixed income derivatives followed by a decade building in AI. He is now the founder of Soundcast, and he argues that the loudest voices on BOTH sides of the AI trade are wrong. We really enjoyed this conversation - it’s a banger! The Slides The Finance/AI Divide * Playing both sides of finance and AI reveals two smart camps that completely miss each other: * Finance people understand compounding but not compounding innovation. * AI people understand Jevons-style efficiency but not macro constraints like savings rates and financing equations. * Mispricings emerge where the two frameworks fail to meet. * Tech guys assume funding is “always there” after a 17-year bull market, while finance guys ride AI narratives without understanding the underlying mechanics. Token Deflation and the Supply Glut Case (Minute 3) * Brandon frames the bull/bear split: * Bears: Michael Burry applying a 2008 playbook to “circular financing,” with the thesis resting heavily on how fast GPUs depreciate; Brandon argues GPUs last longer than bears assume, especially in a supply-constrained market. * Bulls: Jevons paradox means demand grows forever (Bill Ackman’s “infinity ROI” line); Brandon rejects this too. * Brandon’s base case: a traditional commodity-like cycle of shortage then glut, which is actually an optimistic view, not pessimistic. Analogy: British railways were overbuilt, capital holders got smoked, but society benefited for a century. * The key puzzle addressed: how token prices can fall while GPU rental rates rise. * Game-theory: if recursive self-improvement (RSI) is winner-take-all, you buy compute to deny rivals and slash prices to block new entrants; but this alone cannot explain Anthropic’s observed gross margins. * The fundamental answer is token yield: Algorithmic improvements (i.e., DeepSeek’s inference-serving updates) deliver 2 to 3 times more tokens from the same GPU and data center. Prices fell slower than yield rose, creating an arbitrage that justifies buying more hardware, and the end result of that is a supply glut. * The parallel the shale boom, horizontal drilling unlocking inaccessible reserves; same capital-heavy dynamics, same glut outcome. * The headline numbers: * A fixed unit of machine intelligence (”one unit of 100 IQ”) is deflating roughly 13x per year; nothing in history has deflated at that rate. * Extrapolating 13x annual efficiency gains, 25% annual data-center capacity growth, and a 25% shift of training to inference implies roughly 8,000x today’s token supply in three years. * Demand must still be financed out of roughly $115T of global GDP; AI takes share rather than conjuring new income. * Labor risk: challenges the lump-of-labor fallacy. Past technologies created new jobs because people learned faster than machines; this is the first time machines learn faster. Cannibalizing part of a roughly $40T global white-collar wage bill in 3 to 5 years guts the households with the highest propensity to consume, breaking the circular flow that funds Meta, Google, and Amazon free cash flow. RSI vs AGI, Model Tiers (Minute 17) * RSI is not a walk-back from AGI; they are different axes: AGI is the level of intelligence, RSI is the rate of change. * Karpathy’s “jagged intelligence” idea: models are superhuman in some domains, subhuman in others, so AGI has become partly a marketing term. * Chollet and others argue LLMs are only quasi-symbolic reasoners; true symbolic reasoning could unlock up to six more orders of magnitude of efficiency, which dwarfs today’s price declines. * Enterprise behavior: the “token maxing” era is over; big banks and firms like Uber now issue token budgets. ChatGPT’s default “auto” mode is the answer to wasteful model selection, though routing is genuinely hard. * The “Jev” model launch: a deliberately non-general model, trained on reinforcement learning optimized for task completion rather than human-pleasing behavior, can be about 100x cheaper and faster. Extends the toolkit beyond the “nuclear sledgehammer” to screwdrivers and saws. * The profitability paradox: as efficiency compounds while prices lag, inference economics become wildly profitable; DeepSeek’s paper implied roughly $38 per hour per NVIDIA rack. The market is inflecting from supply-constrained to demand-testing, hence the scramble for new use cases (3D modeling, cyber, robotics) to fill an 8,000x supply curve. Personal Agents & Data Security (Minute 33) * Personal-agent reality check: Sora/Muse-style products show weak retention (about 22% after two weeks) because agents require habit change, not entertainment. Brandon’s end state: personal agents end up largely free, bundled to sell devices or intercept shopping. * Data security and $AAPL * Enterprise buyers prioritize data security over price (JP Morgan containerization example); open-source models on personal machines have likely leaked 10x more data than any prior year. * Apple at 39x forward earnings is partly a privacy and on-device AI bet; notable as the rare mega-cap tech not spending down free cash flow on CapEx, even as Intel reports awful consumer PC results. * The “third leg” framework: alongside the real economy and the financial economy sits a new AI economy. Whether or not LLMs are a good business, a massive AI labor force gets built anyway, with major implications for labor’s wage-bargaining power. Mike Green’s framing: the Roman Empire ran on slaves; AI labor has no leisure, rights, or illness, a fundamentally different economic model. * Macro arithmetic: even the bullish 15 to 20% AI-driven GDP growth forecasts require a matching surge in money supply or velocity; velocity is the wrong candidate since US money velocity peaked in 1999 and has fallen since, like Japan. Soundcast and the Reckoning Ahead (Minute 43) * Podcasts are the most information-dense medium (where Huang, Musk, Nadella actually share their thinking), yet there is no real search engine for them. Brandon applies his old “Palantir for sensitive data” playbook (he previously processed bank surveillance data for market-misconduct detection) to structure, attribute, and search podcast content. * Working with roughly 100 sub-agents daily Brandon’s usage ladder runs from level 0 (do it yourself) to level 4 (multiple concurrent projects); his stack is automated research, development and hardening, plus Cloudflare’s serverless platform to skip heavy engineering. * Vibe coding versus commercial software: roughly a 40:1 time ratio; two hours to prototype, two weeks to properly engineer, and the ratio is hard to compress further. * Coding craft shift: the industry is moving from artisan to industrialized phase, but AI is weak precisely at the foundational layers (architecture and database decisions). “Radioactive weeds”: about 10% of AI-generated code is quietly wrong, and models replicate those errors from your codebase over time. Printer-press analogy: what becomes scarce is editorial judgment, not output. * The labor cycle: first-order panic layoffs, second-order hiring of engineers to build automation, then 6 to 18 months of governance and RIF structure before labor data actually weakens, which is what we are now seeing. * $XPO XPO Logistics example: an 18-month internal build, now generating large returns; ROI is real but slow, like any capital project. * The market timing question: when does the hated, puked-out 2002-style entry point arrive? * Goldman’s May forecast of 120 quadrillion tokens per month by 2030 was already outmatched: with current DeepSeek-class serving efficiency, roughly 10,000 quadrillion tokens per month is producible today. * Lab CEOs (Dario) must commit to compute purchases 24-plus months out while model efficiency compounds and demand is unknowable; odds of nailing supply are under 10%. * Because this buildout is debt-financed, overbuilding is not a soft equity loss but “balance sheet Russian roulette.” Even if CapEx stopped today, algorithmic gains would keep compounding o… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.bennysquirrel.tv/subscribe

    AI Outcomes Have FAT Tails
  8. Oct 25

    Assessing the AI Carnage

    Episode 98 and another unmissable Sunday Show! The AI trade is getting flushed. But is this the big one, or just another crowded-position unwind? We unpack the widening cracks in AI financing, why booming compute demand does not automatically mean booming revenue, and whether the S&P equal weight’s comeback signals a genuine market rotation. LLM Watch Pod Summary The AI Unwind in Numbers * The week’s equity weakness = a sharp reduction in crowded positioning, particularly around AI, rather than clear evidence of a broader market collapse. * This is probably not “the big one”. * Net leverage has been aggressively reduced, with positioning measures reaching the bottom percentiles * Similar to the deleveraging seen from late June into late July. * The damage within AI-related trades has been much more severe than the headline indices suggest * The central distinction is between a modest-looking index drawdown and a substantial collapse in individual speculative trades. AI financing: competition for capital becomes the constraint * Only a small number of players can compete effectively for credit-market funding. * Is access to financing is straightforward even for the largest companies? * Borrowers are having to pay more. * Beyond a certain point, a higher interest rate may not solve the problem because lenders hit capacity limits. * This creates a potential dividing line within AI: the ability to fund infrastructure may matter as much as the ability to generate demand for it. AI capex: compute demand is not external revenue (Minute 4) * Is the bullish argument that extraordinary demand for compute, including strong Hopper rental rates, is sufficient to justify AI investment. * Compute demand is not equivalent to revenue. * The relevant question is how much revenue must ultimately come from customers outside the AI industry to support the infrastructure spending. * Discussion of Paul Kedrosky’s AI capex revenue hurdle calculator * It translates AI capital expenditure into a required level of external revenue. * It then expresses that requirement relative to GDP, global software revenue and US white-collar salaries. * The purpose is to test whether the economic assumptions behind the investment cycle are plausible. * “Unlimited demand for compute” becomes less useful as an investment proposition unless it translates into real customer spending outside the sector. * Are credit markets are beginning to confront precisely this issue? AI usage versus spending (Minute 7) * Rapidly rising usage need not produce equally rapid growth in expenditure. * Better routing, system design and optimization can allow usage to grow geometrically while spending flattens or declines. * The key issue is whether customer spending will grow enough to justify the capital already committed. * The strategic motive for continued infrastructure investment. * Large operators may be trying to become indispensable, establish a toll-road position or price competitors out of the market. * In that scenario, the infrastructure owners could retain value even if the model developers face difficult economics. * Own the compute rather than be Anthropic or OpenAI but compute ownership only works if the owner can continue accessing capital markets. That narrows the investable universe substantially. * The hosts compare the possible outcome with a utility-like market in which a small number of dominant providers control access and capture the profits. What happened to Anthropic’s S-1? * Anthropic’s expected S-1 filing did not appear on Friday. * Delayed disclosure may help preserve the AI narrative. * Once detailed financial information becomes public, investors must assess the actual economics rather than an open-ended story. * Differentiation in private debt markets between OpenAI-backed and Anthropic-backed data-centre projects. * Some OpenAI-linked deals are reportedly not getting done. * Where financing is available, he has heard of differences of approximately 200 to 300 basis points. * He explicitly qualifies these observations as individual data points, not a comprehensive market survey. The sell-off spreads beyond direct AI beneficiaries (Minute 10) * The latest week’s weakness has extended further down the AI-related supply chain. * Copper miners fell nearly 10%. Iron ore miners were also coming under pressure. * The damage is moving from direct AI winners into second- and third-order beneficiaries. * Even traditional value investors are beginning to feel the effects. * He also sees increasing concern about slowing growth and a more stagflationary backdrop. * But do these apparently macro-driven moves simply just instead reflect the mechanics of crowded-position unwinds. Equal weight S&P versus Nasdaq (Minute 12) * The ratio has staged a striking rebound over past five weeks. * For it to extend, the semiconductors probably need to break down - but the weekly chart still in primary uptrend: * Betting against semiconductors means betting against currently booked revenue and earnings, not just a speculative narrative. * The hosts agree that “cheap” and “expensive” are difficult labels for highly cyclical businesses. * A bullish memory-stock thesis requires a clear view on: * The timing and scale of additional Chinese supply. * How that supply affects pricing. * Whether strong earnings can persist for another year or longer. * Very low multiples could produce attractive returns if current profitability lasts. * However, the thesis ultimately requires confidence in earnings further out, including 2028 and 2029. * The broader point is that the durability of earnings matters more than the apparent cheapness of a single year’s multiple. Industrials ‘20-for-20’ (Minute 18) * The traditional portfolio role for industrials is challenged in current pricing environment. * Industrials should ideally provide a steady base of compounding businesses. * Many now carry substantial premiums associated with AI infrastructure, electricity constraints or defence spending. * He revisits a hypothetical portfolio built from a 2006 starting point. * A portfolio extrapolating then-current narratives might have owned auto companies, handset manufacturers and GE. * A more successful hindsight portfolio could have included FANUC, Caterpillar and Schneider Electric. * The lesson is not simply to identify good businesses, but to distinguish business quality from the price paid. * Caterpillar is a classic case study - an “accidental bottleneck stock”. * He sees a similar pattern across industrial businesses connected with physical power constraints or the broader AI theme. * Defense stocks present a related problem. * Their thematic narratives are powerful. * That does not automatically make their starting valuations attractive for a portfolio assembled in 2026. Gold miner cash generation (Minute 23) * Popular narrative is that gold miners are “printing money” at $4,000 gold. * Some companies may be generating substantial cash. * That does not justify assuming the same economics across the sector or buying GDX indiscriminately. * Discussion of Respeculator research on ASX-listed gold miners. * The spread between gold prices and reported all-in sustaining costs can suggest margins of approximately $2,000 per ounce. * Yet those apparent margins are not consistently appearing as incremental cash on company balance sheets. * A major reason is spending classified as “growth capex”. * Some growth spending, including mine extensions needed to keep mills supplied, is economically necessary rather than genuinely optional. * He draws a direct comparison with AI accounting. * Ignoring model-training costs (‘EBTIT’) flatters the economics of AI labs. * Ignoring recurring mine-development requirements can similarly flatter the economics of gold producers. * The analytical requirement is to identify which companies actually convert high gold prices into cash after all necessary spending. If you act on anything provided in this newsletter, you agree to the terms in this disclaimer. Everything in this newsletter is for educational and entertainment purposes only and NOT investment advice. Nothing in this newsletter is an offer to sell or to buy any security. The author is not re… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.bennysquirrel.tv/subscribe

Ratings & Reviews

3.7
out of 5
3 Ratings

About

Benny & The Squirrel is a twice weekly show about finance and markets. The show notes for members are hosted on Substack. www.bennysquirrel.tv

You Might Also Like