David Woo — Fed To Trigger '1987' Market Crash This Week?
"Oil prices are going to keep going higher until the stock market goes down, because Trump is not going to taco until the stock market goes down."
One-line take: On the day the US 10-year hit 5% (Brent ~$108, WTI ~$103, a Fed hike ~90% priced for Wednesday), Woo is short stocks and long oil — and surprised the market isn't lower. His loop is endogenous: oil keeps rising until stocks fall ~10% and Trump "tacos," and yields keep rising until stocks fall because a rising market eases financial conditions and lets FCF-negative hyperscalers keep issuing debt for AI capex. Oil is the bigger driver (a synchronized global tightening is being priced off it); the Houthis joining the war lets Iran throttle both Hormuz and the Red Sea (Forties–Brent at its widest since April, Oman–Brent premium rising, Asian inventories depleted, the US SPR near its operational floor). Bessent is trying to cap long rates to save the AI bubble — "a joke" against Chinese selling, Japan urging pensions home to JGBs and Norges Bank cutting $80B of Treasuries — so 5.25–5.30% on the 10-year is where the AI trade breaks. The non-rates trigger: a US ban on Chinese AI models (Chinese labs caught up on inferior chips; the FBI/CIA "distillation" warning is the excuse) → Chinese rare-earth retaliation, with the Xi summit and the $2T Anthropic IPO ("a referendum of the entire AI trade"; secondaries at ~$1.3T) as the dated catalysts. Warsh's Jackson Hole hawkishness echoes 1987: a new chair building credibility, one hike, then the crash when the market expected a second. Positions: Dec WTI 95/105 call spread and QQQ put spreads / OTM index puts, all expiring around the midterms.
1. Stocks & names mentioned
Woo is a top-down macro strategist; the only positions he states are options on the WTI future and on the Nasdaq 100 / S&P 500, shown here on their proxy ETF rows (USO, QQQ, SPY). The other names are evidence for the rates/AI/oil loop. Stance reflects how each was framed in this interview. Sponsor Kalshi is omitted.
| Ticker | Name | Research | View | What Woo said | At |
|---|
| USO | United States Oil Fund (WTI crude — proxy row) | QT · SA · STK | Positive | Long oil via a December WTI futures call spread (95/105) expiring the day after the midterm election — a tactical trade into the midterms: at $105 it pays ~10× the premium. Iran has the most leverage before the vote, and the Houthis now let it throttle Hormuz and the Red Sea. (Position is on the WTI future, not USO.) | 06:04 |
| NVDA | NVIDIA | QT · SA · STK · FA | Neutral | Financing the AI party: $125B into OpenAI's $500B data-center project and reportedly a $10B lead anchor for the Anthropic IPO — "they have to keep this party going." Jensen Huang may shrug off the China question; Wall Street roadshows don't. | 28:14 |
| MSFT | Microsoft | QT · SA · STK · FA | Neutral | Hyperscalers are mostly FCF-negative with capex outgrowing earnings, so every capex dollar is new debt; if stocks don't fall, Microsoft and Amazon keep issuing and rates keep rising "until the stock market pops." | 21:51 |
| AMZN | Amazon | QT · SA · STK · FA | Neutral | Named with Microsoft as the debt issuers funding AI capex — part of $3T+ of off-balance-sheet commitments "the bond market is rebelling against." | 21:51 |
| SpaceX | SpaceX | — | Neutral | The contrast case: six weeks before the SpaceX IPO "nobody was selling" in the private secondary market because everyone was sure they would make money — unlike Anthropic today. | 35:41 |
| DeepSeek | DeepSeek (private, China) | — | Neutral | The benchmark for China's catch-up: 18 months after DeepSeek, five or six Chinese labs match US models while the chip gap has widened (no Blackwell, let alone Rubin) — "the cat is out of the bag." Chinese AI has 25% US share, 50% in six months. | 33:48 |
| QQQ | Invesco QQQ (Nasdaq 100) | QT · SA · STK · FA | Negative | Long QQQ put spreads expiring before the midterms — a tactical trade, traded in and out rather than a standing short. Higher oil + a 5% 10-year are "not good for the stock market"; in July the Nasdaq fell as yields rose. The 10-year breaks the AI trade around 5.25–5.30%. | 20:41 |
| SPY | SPDR S&P 500 ETF (index proxy) | QT · SA · STK | Negative | Long out-of-the-money puts on "spot" (probably SPX/SPY — transcript garble) expiring before the midterms, a tactical trade: stocks must fall ~10% before Trump tacos on oil; a Chinese-AI ban around the Xi summit would bring "a much bigger sell-off." | 20:41 |
| Anthropic | Anthropic (private, IPO pending) | — | Negative | Wouldn't buy the IPO "not at two trillion dollars" — secondary shares trade nearer $1.3T, so insiders sell six weeks out; Amodei's slow-down essay reads as roadshow damage control; the listing is "a referendum of the entire AI trade." | 35:22 |
| OpenAI | OpenAI (private) | — | Negative | Same answer on an OpenAI IPO — "there has to be a price for everything." Paused "Astra" over security risk, then launched "Astron" as the start of AGI: the national-security race with China now overrides safety; its $500B data-center plan feeds the AI debt wave; falling token prices squeeze margins. | 35:01 |
| INDA | iShares MSCI India ETF | QT · SA · STK | Negative | India, a major oil importer, has one of the worst-performing stock markets this year — high oil hits inflation and the current account, and the rupee has been very weak. | 15:12 |
| FXY | Invesco CurrencyShares Japanese Yen Trust | SA · STK | Negative | The yen is "already the weakest currency in the world," and higher oil makes it harder to defend for big-importer Japan; Bessent fears a yen run sparks "the mother of all sell-offs in US treasuries." | 14:16 |
2. Talking points
0:00 The setup — 10-year at 5%, oil above $100
- Monday, September 14: the US 10-year hits 5% (highest since 2023), Brent ~$108, WTI ~$103, and a Fed hike this week is ~90% priced. Trump posts "whoever wins AI wins."
2:25 Short stocks, long oil — surprised stocks aren't lower
- Stocks are selling off on both oil and yields. Woo is short stocks and long oil and is surprised the market hasn't fallen much more over two weeks of surging yields and oil — are stocks "the new safe haven"?
2:49 The endogenous loop — nothing stops until stocks fall
- Oil keeps rising until stocks fall because Trump won't taco until they do; yields keep rising until stocks fall because central banks see a rising market as easing financial conditions. The market is "not out of the woods by any stretch."
4:02 Oil is the chicken — synchronized global tightening
- Lagarde warned of second-round inflation effects; over six weeks the Fed, BoE, BoC and RBA have all repriced to more than two hikes, and the ECB already hiked. Oil drives it — if oil collapsed tomorrow, rates would tumble "at least for a while."
6:04 The oil trade — a 95/105 Dec WTI call spread
- Expires the day after the midterms. Oil could reach 150, but "I've been there before and then Trump is going to taco" — so the target is realistic: at 105 the spread pays ~10× the premium.
- Iran has the most leverage before the midterm, when Trump is politically constrained, to force him back to July's deal terms.
7:22 The Houthis change the map — Forties–Brent blows out
- Absent March–July, the Houthis joined in August (partly because the Saudis overplayed their hand). Iran can now throttle Hormuz and the Red Sea; the Forties–Brent spread, a gauge of European physical tightness, is at its highest since April as Red Sea tanker traffic collapses.
8:50 Reserves won't be spent — Asia is scrambling
- Chinese imports are up two months running, and teapot refiners are shopping worldwide; the Oman–Brent premium (Asian vs European price) is rising. The US SPR is near its operational floor (~150–200M bbl vs 280M now); Japan told refiners no strategic releases in September–October. Governments won't underwrite the shortage indefinitely — bullish oil.
11:02 The story everyone missed — shuttle transfers were Iran's gift
- August's $80s oil rested on Saudi/UAE tankers shuttling crude out of Hormuz. Woo's view: that only happened because Iran allowed it while negotiating, via Oman, a deal giving it control of the strait. With the talks collapsing (yesterday's meeting cancelled) and Trump sabotaging the deal, Iran could let no tanker through — the path to $120.
14:16 Not an Asian crisis — but Japan and India are exposed
- The 1990s crisis was leverage, fixed exchange rates and mismatches, not oil. But higher oil hurts the yen (already the weakest currency), which feeds straight into Treasury yields; India's market is among the year's worst on oil, the current account and a weak rupee.
15:55 Bessent's cap on long rates — protecting the AI bubble
- Bessent moved to cap long-term rates the same day news broke of Anthropic's $2T IPO plan and OpenAI's $500B data-center project with $125B from Nvidia. Surging yields are largely AI capex: hyperscalers are FCF-negative, capex outruns earnings, and $3T+ of off-balance-sheet commitments means debt is coming.
18:08 "A joke" — the foreign buyers are leaving
- An $8B buyback can't fix it. China sells Treasuries every month (accelerating); Japan, now the largest holder, told pension funds to consider selling Treasuries to buy JGBs; Norges Bank will cut US bonds by $80B. They'll use financial repression to save AI, but "they're fighting a losing war."
19:38 Crowding out
- Savings are flat to lower (household saving ~3%, government saving worse) while investment surges — long rates rise and crowd out mortgages, home construction "and possibly even AI." In July the Nasdaq moved inversely to bond yields.
20:41 The equity short — puts expiring before the midterms
- Long QQQ put spreads and out-of-the-money index puts, both expiring before the midterms, when "things are going to get most interesting." Trump tacos on oil only after a ~10% fall; until then oil and rates keep squeezing.
22:39 Amodei's essay and the case for bombing data centers
- Amodei's Saturday essay: slow advanced AI, but not so much that China overtakes. A former Obama official (Jacob Stokes) said the US should use espionage, cyber attacks or even strikes on Chinese data centers if China nears AGI.
- OpenAI paused "Astra" over security risk, 100 companies warned of AI cyber attacks, then OpenAI launched "Astron" as the start of AGI — Woo reads it as the administration deciding China reaching AGI is the bigger risk.
25:48 The Chinese-AI ban is coming — then rare earths
- The FBI/CIA "industrial scale distillation" warning gives Trump the excuse to ban Chinese AI models, possibly before the Xi meeting (Sept 24) and the Anthropic IPO. Chinese AI holds 25% US share, 50% in six months, he says. China would likely retaliate with its rare-earth card — the non-rates trigger for a sell-off.
27:54 The Anthropic IPO — a referendum on the AI trade
- Every roadshow meeting asks about China and cyber risk; the essay answers them. If the IPO goes poorly "the whole AI trade could basically collapse," hence Nvidia's reported $10B anchor. Trump's Truth Social post attacking Amodei shows how much is at stake.
30:18 "The AI trade is the US economy"
- US ~15% of world GDP but ~60% of world stock-market cap; the Mag 7 ~30%, tech ~40%. Without the AI boom the market would already have collapsed on oil and rates. Trump has "thrown everything behind this AI trade" and has stopped listening to advisers — so Woo won't bet against him outright, but a ban opens Pandora's box.
32:46 China isn't 5–10 months behind any more
- Half of the top six ranked models are Chinese, matching US performance without Blackwell (let alone Rubin) chips. The compute gap is wider than at DeepSeek's release 18 months ago, which makes the catch-up more impressive, and five or six labs are racing. So the US can't slow down, even at the risk of a systemic cyber meltdown: "the biggest single tail risk."
35:01 Wouldn't buy Anthropic or OpenAI at $2T
- Anthropic's secondary shares last traded around a $1.3T valuation against a $2T IPO: insiders who could wait six weeks are selling. Nobody sold six weeks before SpaceX's IPO.
- Inference demand grows because prices are collapsing, which squeezes margins; many chip orders are double orders and may not be binding, so a pullback would be synchronized.
37:14 They can't have their cake and eat it too
- To keep the AI bubble going the administration must give something: live with higher rates or ban Chinese models — neither good for stocks. "That's the way I'm trading this."
38:30 The break point — 5.25–5.30% on the 10-year
- Not 6%; the AI trade blows up around 525–530bp. "So we're almost there."
38:54 Warsh's Jackson Hole mistake and the 1987 analog
- Told his credibility was the problem, Warsh went hawkish with a black-and-white rule (hike while core PCE is above ~2%) — "pouring fuel on the fire." Greenspan took over three months before the 1987 crash; the Fed hiked once, and stocks crashed when a second hike was expected. If the Fed hikes this week, rates won't go down.
41:22 How to be bearish and still make money
- He runs a total-return book benchmarked to 3-month T-bills, bearish on stocks all year but up ~3% thanks to being long oil. You can't just short and stay short — "you might just lose your house first." He caught July's sell-off, got out, re-shorted in August; the trigger he's watching is the Xi summit and a Chinese-AI ban.
43:21 Where to follow
- His YouTube channel David Woo Unbound, a retail subscription service at davidwoounbound.com, and his book (as spoken, "Merry-Go-Round Broke Down") on Amazon, with a screenplay for a possible TV adaptation in progress.
3. In plain English
A jargon-free summary of the thesis behind each name — what it actually is and why he holds that view. (Plain-language companion to the table above; renders on each ticker's consolidated page.)
USO — United States Oil Fund (WTI crude) Positive
USO is an ETF that tracks US crude oil (WTI). Woo doesn't own USO; he bet on December oil futures with a "call spread": he bought the right to buy oil at $95 and sold the right to buy it at $105. That caps his winnings at $105, but the bet was cheap, so if oil simply reaches $105 by the day after the midterm elections he makes about ten times what he paid. He could aim for $150, but thinks Trump backs down before then, so he picked a realistic target.
Why he thinks oil rises: Iran knows Trump is most vulnerable just before the midterms. Yemen's Houthis have now joined in, so Iran can choke both the Strait of Hormuz and the Red Sea shipping lane. Price gaps between local crude grades show real physical shortages in Europe and Asia, and governments are unwilling to keep releasing emergency stockpiles.
NVDA — NVIDIA Neutral
Nvidia makes the chips that power AI. Woo isn't judging the stock. He points to Nvidia as the company paying to keep the AI boom going: $125 billion toward OpenAI's giant data-center plan, and reportedly $10 billion to anchor Anthropic's stock-market listing. His point is that the AI industry is now funding its own customers to keep the party going, a sign of how much depends on the boom not stopping.
MSFT — Microsoft Neutral
Microsoft is one of the "hyperscalers," the cloud giants building AI data centers. Woo says these companies now spend more than the cash their businesses bring in, so the extra spending is paid for with borrowed money. That flood of new corporate bonds is a big reason long-term interest rates are rising. As long as stocks keep climbing, the borrowing continues and rates keep going up, until something breaks. It's a macro point, not a verdict on Microsoft's business.
AMZN — Amazon Neutral
Amazon comes up alongside Microsoft as a big AI spender that raises money by selling bonds. Together these companies have promised more than $3 trillion of future spending that doesn't show up on their balance sheets, and Woo says the bond market is "rebelling" against the debt coming. Again it's evidence for his rates view, not a call on Amazon itself.
SpaceX — SpaceX Neutral
SpaceX is his comparison point. Before a company lists, early shareholders can sometimes sell privately. In the six weeks before SpaceX's stock-market debut, nobody wanted to sell because everyone expected a big pop. Anthropic insiders are selling well below the planned listing price, and Woo reads that difference as a warning sign.
DeepSeek — DeepSeek (private, China) Neutral
DeepSeek is the Chinese AI lab that shocked markets about 18 months ago with a cheap, capable model. Woo uses it to measure how fast China has caught up. US export bans keep China from buying Nvidia's newest chips, so China's hardware gap is actually bigger now. Even so, half of the six top-ranked AI models are Chinese, and five or six Chinese companies are racing. His conclusion is that the US can't afford to slow down, and that is why Washington may simply ban Chinese AI.
QQQ — Invesco QQQ (Nasdaq 100) Negative
QQQ tracks the Nasdaq 100, the big tech stocks, and it is the purest way to bet on or against "the AI trade." Woo owns "put spreads" on it: bets that pay if the index falls, capped at a set level to make them cheaper. They expire before the November midterms, and he trades them in and out rather than holding a permanent short.
Why he's bearish: oil above $100 and a 5% 10-year Treasury yield are bad for expensive tech stocks, and in July the Nasdaq fell whenever yields rose. He thinks the AI trade breaks if the 10-year reaches about 5.25–5.30%, or sooner if the US bans Chinese AI and China hits back by cutting off rare-earth metals.
SPY — SPDR S&P 500 ETF Negative
SPY tracks the S&P 500, the broad US stock market. Woo holds cheap "out-of-the-money" puts on the index. They pay off only if stocks drop meaningfully before the midterms. (He said "spot," which almost certainly means the S&P index.)
His reasoning is a loop: Trump won't back down on the Iran war, and oil won't stop rising, until stocks fall about 10%. Central banks keep raising rates while stocks stay high. In his view nothing relieves the pressure until the market falls, so he is positioned for that.
Anthropic — Anthropic (private, IPO pending) Negative
Anthropic, the maker of Claude, is preparing a stock-market listing at a roughly $2 trillion valuation. Woo wouldn't buy at that price. In private trading its shares were recently priced nearer $1.3 trillion, which means insiders would rather sell now at a big discount than wait six weeks for the listing.
He also sees the CEO's recent essay urging slower AI development as a response to investors on the pre-listing tour who kept asking about China and cyber-security. And he calls the listing a "referendum" on the whole AI boom: if it goes badly, the AI trade could unravel.
OpenAI — OpenAI (private) Negative
OpenAI, the maker of ChatGPT, is also a possible listing candidate, and Woo's answer is the same: "there has to be a price for everything." He points out that OpenAI paused a model over security fears, then weeks later launched one it called the start of AGI (human-level AI). He reads that as the government deciding that beating China matters more than the safety risk.
He also notes that AI usage is growing fast partly because prices are collapsing, which squeezes profit margins for companies like OpenAI.
INDA — iShares MSCI India ETF Negative
INDA is an ETF of Indian stocks. India imports most of its oil, so expensive oil pushes up inflation, widens the gap between what the country buys and sells abroad, and weakens the rupee. Woo notes that this has made India's stock market one of the worst performers of the year.
FXY — Invesco CurrencyShares Japanese Yen Trust Negative
FXY tracks the Japanese yen against the dollar. Japan imports nearly all its oil, so high oil prices weaken the yen, already the world's weakest major currency. That matters beyond Japan: Japanese investors are the biggest foreign holders of US Treasuries. If the yen keeps sliding and Tokyo pushes its pension funds to bring money home, they could sell US bonds and push US interest rates even higher, which is exactly what Treasury Secretary Bessent fears.
Summary & timestamps derived from the public YouTube video (David Lin, 2026-09-14; transcript in transcript.txt) for personal study. Not investment advice. © David Lin / David Woo Unbound for source material.