David Woo Unbound · ex-BofA Head of Global Rates, FX & EM Strategy — a top-down macro/rates strategist. Running synthesis of his video appearances, with per-transcript breakdowns and a stock index. Names are usually cited as evidence for the macro view, not single-stock calls.
Wouldn't buy its IPO at a $2T-type price — safety pause then "AGI" launch shows the China race overriding risk; $500B data-center plan feeds the AI debt wave.
In one line: Short stocks, long oil. Oil and long yields are an endogenous squeeze on a market that is really one AI trade: neither stops until equities fall (~10% before Trump "tacos"). Washington will do anything to protect the trade (cap long rates, ban Chinese AI), but each defence costs something, so he holds catalyst-dated options into the midterms: a Dec WTI 95/105 call spread, QQQ put spreads and OTM index puts. Earlier (June) he framed the bond sell-off as a real-yield story with two legs, oil and AI, and liked a steepener plus long-dated gold calls.
The loop that forces a sell-off (Sep 2026). With the 10-year at 5% and oil above $100, a rising market keeps feeding both pressures. Trump won't back down on Iran until stocks fall, central banks treat a rally as easy financial conditions, and FCF-negative hyperscalers (MSFT/AMZN) keep issuing debt for AI capex. Oil is the driver of a synchronized global tightening; if it collapsed, rates would follow.
Oil: Iran's leverage peaks before the midterms. The Houthis' August entry lets Iran throttle Hormuz and the Red Sea. August's shuttle transfers only happened because Iran allowed them while negotiating, and those talks are collapsing. Physical gauges agree: Forties–Brent at its widest since April, the Oman–Brent premium rising, Chinese imports recovering, the US SPR near its operational floor, Japan freezing releases. Trade: a Dec WTI 95/105 call spread, ~10× premium at $105.
Rates: Bessent is defending the AI bubble and losing. The long-rate cap coincided with the $2T Anthropic IPO plan and OpenAI's $500B data-center deal; an $8B buyback can't offset Chinese selling, Japan urging pensions home to JGBs, or Norges Bank's $80B cut. Crowding out follows. The AI trade breaks near 5.25–5.30% on the 10-year. Warsh's rigid Jackson Hole framework echoes 1987: a new chair, one hike, then a crash on the second.
AI: the China race overrides everything. Chinese labs caught up on inferior chips (half the top six models; 25%→50% US share). Expect a US ban on Chinese models, justified by the FBI/CIA distillation warning, and Chinese rare-earth retaliation. That is the non-rates trigger, dated by the Xi summit and the Anthropic IPO, which he calls "a referendum of the entire AI trade" and wouldn't buy at $2T (secondaries ~$1.3T). "The AI trade is the US economy."
Process: in and out, benchmarked to T-bills. Bearish all year yet up ~3% vs 3-month bills because long oil pays for the shorts. Never short and stay short: he caught July's drop, exited, and re-shorted in August.
It's real yields, not inflation (Jun 2026). Of the 70bp rise in the 5-year since Feb 27, ~60bp was real yields and only ~10bp break-evens; the 30-year's rise was almost entirely real. If oil were the driver, break-evens would have jumped — they didn't.
Two legs: oil + AI. AI's productivity promise lifted potential-growth expectations and real yields; the AI wealth effect let the bond market discount oil's upside-inflation risk over downside-growth risk; plus expected issuance to fund AI CapEx. Neither leg alone explains the repricing — only the combination. To extend, bonds "need two legs to stand on."
Oil leg intact — heading higher. Hormuz stays closed (no tankers June 3rd), inventories are depleted (an Exxon exec flags $150–160 risk), and Woo thinks Iran is walking away from talks — so Trump can't credibly promise a deal. Status quo or a broken ceasefire both push oil up.
AI leg looks tired. Q1 hyperscaler CapEx (MSFT/AMZN/GOOGL/META/ORCL) fell vs Q4 — even more in volume terms given memory-chip costs. Capabilities are "plateauing," competition is closing in on Claude Code (Cursor/Copilot/Codex; MSFT canceled most licenses), and "token maxing" is giving way to budget-constrained usage (Uber blew its annual budget in 4 months). Chips (INTC/AMD/AVGO) have lost momentum; the rally is narrowing and crowded.
The June trades. Too early to buy bonds outright, but a steepener makes sense again. He likes long-dated gold calls — if the AI bubble pops, the wealth effect reverses, real yields fall, and investors rotate back to safe havens.
The product
Grounded in Woo's own statements in the archived appearances (2026-SEP-14 David Lin interview; 2026-JUN-04 channel video).
What it is:David Woo Unbound. Woo runs "a big institutional business," advising "some of the biggest investors in the world, sovereign wealth funds, hedge funds," and manages a total-return portfolio benchmarked to 3-month T-bills. For retail there is a free YouTube channel (David Woo Unbound), a paid retail subscription service at davidwoounbound.com for "my investment strategy," and his book (as spoken, Merry-Go-Round Broke Down, on Amazon; a TV adaptation is in development).
Offering
What it is
How he runs it
Seen in the index
YouTube channel (free)
Short macro monologues on rates, oil, FX and the AI trade
Tracks stories ahead of the market ("if you've been watching my channel, you'll know," e.g. the Hormuz shuttle transfers)
Macro advice to sovereign wealth funds and hedge funds, plus his own total-return book
Benchmark = 3-month T-bills; shorts traded in and out on technicals, paired with a paying long
2026-SEP-14 (up ~3% YTD while bearish stocks all year)
Book
Merry-Go-Round Broke Down (title as spoken)
On Amazon; screenplay for a potential TV adaptation underway
—
Catalyst-dated, defined-risk trades. He shows retail how to express a macro view without unlimited risk: spreads sized to a realistic target (~10× premium) and expiring around the event (the midterms).
A process for being bearish without blowing up. His own record, bearish all year yet up against T-bills, comes from going in and out rather than staying short, and from pairing the short with a long that pays in the same regime (oil).
Cross-asset causality. He links oil, yen/JGBs, Treasury supply and AI capex into one chain, so a subscriber can see which variable to watch (e.g. the 10-year at 5.25–5.30%).
Transcripts
One dated page per appearance — each has its talking points and the saved transcript. Newest first.
David Woo appearances discovered via search (David Woo Unbound), not yet processed — verify publish dates & channels, newest first. Limited to the last ~2 years. None queued yet.