Larry McDonald — research hub

Stock & name index

▲ Positive

► Neutral / referenced

▼ Negative

Overall thesis

• NEW (Sep-08) — the contrarian long-duration trade. The first genuine change of view since 2022: unconnected fixed-income PMs in the chat and at the dinners "starting to buy duration." The screen is sentiment, not forecast — record bears-vs-bulls, CFTC positioning inverted versus the 2017–21 max-long extreme, "everyone knows the bear case, nobody's thinking about the bull case" — so the bear case is "80 to 90% priced in" and the last leg down is the entry. The mechanism is the supernova: a commodity-led inflation spike (diesel, corn, wheat; the strait closed ~200 days; "the next three CPI prints are very treacherous") pushes yields up one more time, breaks a bottom-70% consumer already decimated, and pulls recession forward — the 2021→2022 template that ended with the Nasdaq −35%. Instruments: TLT and zero-coupon ZROZ for duration, IVOL for the steepener ("up 15, 20% total return" on a good steepening). The convexity is the argument: a 6⅛% Google issue at 88 goes to "120 or 130" if the 30-year returns to 3–3.5%. Language is deliberately staged — "start at least thinking about buying long-term bonds."
• NEW (Sep-08) — short the lenders, not the borrowers; hedge, don't sell, the winners. Record price-to-book on JPM and BAC, priced for two-year-old Bowman/deregulation news while carrying bond mark-to-market losses and the off-balance-sheet AI credit; the chat's two-year financials bull flipped bearish on BofA and Goldman. Sizing rule: 1–3% of the year's gains into protection — one-year XLF/BAC puts (vol is cheap), VIXY rented for Sep–Oct, FAZ only with the levered-decay caveat ("rent, don't marry"). Hard assets stay on: still max long SLB/oil services, coal (CNR +32% y/y), copper (+82%), gold miners (+50%) — "but now you want to protect those gains."
• Private-credit "credit crisis" — a 2008 echo. Wall Street's "idiosyncratic" spin is breaking — two names became ~17 credit events; BlackRock gated its $26B HPS fund (met only 54% of redemptions); quarterly liquidity on the world's most illiquid asset class turns "mark-to-myth" into forced selling. The 21 Lehman-era systemic indicators jumped to "intermediate-high"; bank equities (BofA) and BDCs (KKR) down 30–40% — the worst financials start since 2008.
• AI "under the valley" first — and the Mag 7 hollowed out. The trillions of AI capex are "disruptive capital" gutting software firms and labor before any productivity boom (Block −40% staff; Adobe/Salesforce/IBM); 200–300k job losses feared by summer. China's IP theft (platforms like Kimi "distilling" Claude into open source) dilutes US return-on-capital — "hard assets are destroying the Mag 7," which is ~unchanged since October (Microsoft/Meta/Oracle stalled). Another "DeepSeek moment" could drop the Mag 7 ~50%.
• Semis are a hyper-bubble. SOXX +280% in 56 weeks vs the Nasdaq's +130% into the 2000 top ("double"); SOXX ~54% above its 200-DMA, Korea's EWY ~156% above its 200-week MA. The chip/IPO crowding "has to crash" to make room for the SpaceX/Anthropic/OpenAI mega-IPOs being force-fed into passive indices.
• The mega-IPO supply wall (Jun 11). The Friday sell-off is "indigestion," not Hormuz: SpaceX (~$2T ≈ 6% of GDP) + Anthropic + OpenAI ≈ $200–250B of immediate raises, Google's $80B secondary, surging convertible issuance (CFOs — "the smartest sellers in the world" — dumping equity like Q4-2021), and ~$3T of restricted shares unlocking 6–12 months post-IPO. Liquid Mag-7 stock is what gets sold to make room (S&P vs equal-weight divergence); NASDAQ-100 went $30T→$41T in <50 trading days and the 2021–22 analog ($19T→$12T in four quarters) "is an example of what's about to happen."
• New inflation-shock regime → steepener + healthcare (Jun 11). Supercore CPI (which "you can't fake") annualizes to 5.2% by year end → 6–8% headline a year out; with $1.1T of debt interest the hike "muscle memory" is "a mirage," so he's long the 2s30s steepener (via IVOL). The crowded momentum factor (long semis / short healthcare, now 8% of the S&P vs 16%) turns at quarter-end: healthcare "a big winner the second half" — ISRG/BAX as the AI-medical data plays.
• Stagflation trap + financial repression. Deficits ran $250B (2006) → ~$1.9T (2026); interest on the debt is now ~$1.1T. He puts ~80% odds on 4–7% CPI and 20–25% on double-digits. The escape: hold rates below inflation (financial repression / proto-yield-curve-control), a new ~3% target, recalculated ("trimmed") inflation — "the only way out of a $39T debt hole" short of a soft default. Against this much fiscal, the Fed is "a squirt gun at a fire."
• The dark side of passive investing. Top-2 S&P names ~16% (Mag 7 + Oracle ~50% of the index) vs financials ~27% before 2007; Microsoft ~14× sales, Nvidia ~31× sales. Shares locked at BlackRock/State Street/Vanguard, "nobody on the conference calls" — the market "can't turn" and index rules are being waived so billionaires can dump IPOs into it. Jun 16: the 401(k) has been "hijacked" — a $1M passive portfolio is ~$450k of tech at record valuations, the same alarm as financials at ~30% of the index in 2007-08; passive is 60–65% of the market, so "at some point passive becomes evil because it gets too big and too easy to game." SpaceX ($30B in 2019 → ~$2T, 2× Berkshire, 14× Facebook's IPO size) is being fed in on a ~4% float and an unusually aggressive unlock, with Anthropic and OpenAI behind it. The inclusion mechanic is the tell: CoreWeave added to the Nasdaq while "financially a mess," Lululemon added at all-time highs and now −60% — insiders front-run the add and "the little guy gets holding the bag." Consequence: the S&P likely unchanged for 5–10 years (the 2000–2010 precedent) while energy/materials/industrials go 14–15% → 30–40% of the index.
• Own the hard assets & global value; play the AI power, not the chips. The most underappreciated AI trade is the power bottleneck — natural gas (FCG), coal, nuclear (NUKZ), uranium (owns the SRUUF commodity over miners like CCJ/URNM; likes Denison's team). Plus gold/silver miners (bought GDX/SLV into the sentiment washout; gold dip ~$4,100–4,800 → ~$6,000), oil services (Schlumberger "best AI play", Weatherford, OIH), and the "1968–81" basket (BHP, Rio Tinto, Chevron, Vale). Hard assets are still ~1–1.5% of household wealth vs 3–3.5% in the early '80s.
• Behavioral edge — capitulation creates bargains. Democratized info + 0DTE options mean everyone hits stops at once, so sell-offs over-shoot — year-end tax-loss bargains in beaten-down quality (Chipotle, Lululemon, Target) and the gold-miner washout. Crypto/leverage gamification (Robinhood; 8 tertiary coins −$210B) flushes retail at the lows; he went long Bitcoin for the first time at the March-30 liquidity low.
• Regime + plumbing. Post-2022, lower stock prices now come with higher bond yields — no more bond hedge; long-duration US/Japan/UK/France bonds have bled since 2022. "Stealth QE" (banks force-fed Treasuries via SLR as China buys fewer; JPM reserves −$200B) held rates down; Bessent & Warsh now "coordinating openly." Bearish dollar, bullish hard assets; gold is the #1 US export (Tether buying tons). Sep-08 revision: the long-end move is now read as a term-premium / supply problem rather than an inflation-expectations one — hyperscalers issuing "an extra 5, 600, 700 billion" over 12–24 months on top of Treasury supply, plus French/UK/DSA political risk, while inflation expectations were falling. Bessent is the offset ("much more like Trichet" than the Yellen "faculty lounge"): threatening to issue fewer 10s and 20s and hinting at buying them back — an operation twist — because a global yield breakout would hammer the banks holding $30T of 2018–21 paper (a Google issue par→88, an Apple bond par→49, an Oracle bond at 8%). Not another SVB, though: "everyone kind of knows not to trust risk parity" — the 60/40 ETF is unchanged since 2021 — so the danger is a breakout nobody is positioned for.

The product — The Bear Traps Report

The trade types below are from his own descriptions across the appearances archived here (esp. 2026-MAR-09, 2026-MAY-09 and 2026-JUN-18).

How it serves retail investors:

• Institutional intelligence down-market. The pitch is explicitly that flow, positioning and credit-stress data — normally visible only inside bank/fund Bloomberg chats — reaches advisors, family offices and individuals at newsletter prices.
• Concrete, dated alerts. Entries and exits with tickers and levels (GDX/SLV buys, the SLV 106→80 round trip), not just macro narrative — so a subscriber can replicate the trade.
• Risk dashboard a retail investor can't build. The 21 systemic-risk indicators and the liquidity model condense credit-market plumbing (CDS, loan-market stress, reserve flows) into a single "how much risk to carry" signal.
• Behavioral discipline. The recurring edge sold is capitulation math — buying washouts and trimming euphoria — which is exactly where retail typically does the opposite.

Transcripts

One dated page per appearance — each has its full stock table, talking points, and the saved transcript. Newest first.

To process — backlog

Larry McDonald (Bear Traps) appearances discovered via YouTube search (Larry McDonald / Lawrence McDonald), not yet processed — publish dates & channels verified, newest first. Limited to the last ~2 years; excludes "MS NOW" / "The Last Word" (that's Lawrence O'Donnell). "(clip)" = short TV segment.

For personal study — not investment advice. Source material © the respective shows / Bear Traps Report.