SpaceX Over, Fed Meeting Done, What’s Next?

Summary

• Goldilocks through the summer: Kevin Warsh’s presser sounded hawkish but the dot plot’s one hike through year-end was “kind of what we expected,” and the big fears — Iran war, oil-driven inflation, the SpaceX IPO as a liquidity moment for the markets — “none of which really ended up materializing.” The host expects continued performance from what’s already working — “your Intels of the world, your SanDisks of the world, your Microns” — at least until August–September: “I’m fairly certain that we’re going to all be pretty happy by the end of the summer.”
• The Mag 7 trade in one move: they’re selling equity and raising debt to fund data-center capex, so their stock struggles relative to the semis they’re purchasing. Buy what they’re buying, not what they’re selling — if a company raises $100B to invest in data centers, own the chips.
• Hood over Hyperliquid: Robinhood crypto revenue fell 47% in Q1 but platform assets are up 50% year-over-year to roughly $400B, Gold subscribers up 36%, and agent trading is rolling out to all users. Even the 10% workforce cut is framed as bullish — “money doesn’t just disappear, money gets reallocated” to marketing, buybacks, and comp. Hyperliquid (which the host still says he’s a bull on) is being “held back by the broader crypto complex.”
• Saylor’s STRC is in a spiral with no exit: the 11% dividend is funded by selling equity, which compresses the MicroStrategy premium; the options are pause the dividend (sending STRC down further from 82.61), cut it, or sell “an exceptional amount of BTC” against debt coming due in 2027. “If I’m you, I’m probably just not touching BTC.”
• The emerging-markets thesis for 2026 was “completely incorrect” — only Korea worked, purely via semiconductors. All real innovation is US AI, which strengthens the dollar and is “bad for gold, bad for Bitcoin, bad for commodities in general.” Intel just ripped to all-time highs on a deal to manufacture chips for Apple in the US.
• The Iran memorandum was “phenomenal for Iran” and advances neither US nor Israeli interests — an “untenable situation” that leads to future conflict, likely not in the next one to three months but potentially September–October or after the midterms, which would spike oil and send markets tumbling. “Until then, I think we’re good to go.”
• Next megatrends: biotech and nuclear. AI is finally hitting physical products — Midjourney’s scan device (claimed to perform more scans than every MRI in the US), an FDA more lax courtesy of RFK that sent a stock called Unicure Huntington up 80% on a June 17 U-turn — so the host put 3% of the portfolio into ARKG and plans to research further. On nuclear: bet on companies, not the commodity, and URA is “probably going to get a 3 to 5x” over 3–5 years.

Deep dive

1. One hike was already priced — welcome back to Goldilocks

• The host’s frame (solo on the stream this week): the market is a weighing tool that prices the future, not the present. The forward curve had begun pricing rate hikes off rising CPI, causing the recent sell-offs — then the dot plot landed at one hike through year-end and the market shrugged: “that’s kind of what we expected.” Warsh sounded hawkish in his presser; “the market just doesn’t care.”
• What we’re seeing is “a return to normalcy”: Iran war, rising oil, weak earnings, and the SpaceX IPO as a liquidity moment were all feared, and “none of which really ended up materializing.” With summer slowing news flow — maybe an Anthropic or OpenAI IPO ahead — he expects continued performance through August–September from Intel, SanDisk, and Micron.
• The Mag 7 mechanic worth keeping: they’re underperforming because they’re “selling their equity, raising debt to fund CapEx build-out.” If someone raises $100 billion to invest in data centers, “you want to go buy the things that they’re buying, not what they’re selling” — and what they’re buying is semiconductors.
• The 2026 “year of the emerging market” thesis (Brazil, Korea) “ended up being completely incorrect” — Korea only worked because it’s exclusively driven by semis. True innovation is all US AI, which strengthens the dollar and hurts gold, Bitcoin, and commodities. “If you’re still exposed to that EM thesis, that’s definitely not where you should be.”

2. Hood over Hyperliquid — and layoffs are bullish

• Robinhood’s crypto revenue was down 47% in Q1, “but everything else is doing ridiculously well”: platform assets up 50% year-over-year to ~$400 billion, Gold subscribers up 36%, agent trading rolling out to all users. Hyperliquid — “obviously I’m a Hyperliquid bull” — is getting “dragged down by the broader crypto complex” via cross-asset holders, making Hood the better play.
• The reallocation argument, stated as a law: workforce cuts are coming everywhere — finance, software, “once the humanoid robots come” — and “that is not bearish, that’s bullish. Because money doesn’t just disappear, money gets reallocated”: every dollar saved on a fired engineer goes to marketing, buybacks, or C-suite comp, which ends up as more spending.

3. Saylor’s STRC has no exit

• STRC — “the Michael Saylor Bitcoin savings account,” down to 82.61 — pays an 11% dividend funded by selling MSTR equity, which compresses the MicroStrategy premium. The spiral: pause the dividend (sends STRC down more), reduce it, or “sell an exceptional amount of BTC” to cover debt obligations, much of which “comes due in 2027.” The host’s alternative history: the last six months of equity issuance should have refinanced that debt instead of “buy buy buy more Bitcoin” until it got untenable.
• The viral clip of Saylor saying he used ChatGPT to design STRC’s structure draws the episode’s best riff: ChatGPT is “a little bit sycophantic” — “If we’d used Grok, we would have blown up. If we’d used Claude, Bitcoin would have been at 500k a coin. Because we used ChatGPT, we’re just cooked.” Bottom line: “if I’m you, I’m probably just not touching BTC.”

4. $38 smoothies and the K-shaped risk

• Ground-truth inflation check: a sandwich and a smoothie at Joe & the Juice ran $38 — “this is not going to end well in 5 years, in 6 years, in 7 years,” more likely showing up in the presidential election than the midterms.
• The structural version: AI means “the top 10% of people are now 300 times more productive than the bottom 50%,” so they’ll suck up even more capital — Andrew Yang “was kind of on top of it from the beginning” with UBI. The realization that the bottom half is struggling and capital needs redistributing is “probably the largest risk to forward market appreciation” — though “not something that we necessarily have to worry about” right now. Meanwhile, only half tongue-in-cheek: the most patriotic thing you can do for America’s 250th is “spend frivolously.”

5. The Iran deal is untenable — flare-up risk after the midterms

• The host’s read: Trump went into Iran “riding high off of the Venezuela deal,” expecting a simple operation, and “got stuck in a quagmire.” Iran’s leadership is taken out — possibly leaving moderates in charge — but the memorandum of understanding, even if signed, advances Iran’s interests, not America’s or Israel’s: “It was phenomenal for Iran. Iran sort of ran away with that one.”
• That makes it “a situation that’s going to lead to future conflict” — probably not in the next one to three months, but come September–October, or after the midterms once the administration’s incentive to keep a lid on it fades. A Hezbollah–Israel exchange “torches the deal,” spikes oil, and sends markets tumbling. “But until then, I think we’re good to go.”

6. Next megatrends: AI hits atoms — biotech, nuclear, and the physical world

• Biotech is the new homework assignment — the host sent a note to Martin Shkreli, “the king of biotech,” asking how to get up to speed. The trigger: Midjourney’s device claimed to perform more scans than the entirety of all MRIs in the United States — “the first time that I’ve seen AI actually be applied to a physical product as opposed to just coding.” Honest hedge intact: “Will it work? I have no idea. It might work. It might not work. But at least it’s being tried” — $70 million raised, and something like Butterfly Therapeutics went up 33% on the news.
• The regulatory tailwind: an FDA “more lax courtesy of RFK than it’s ever been” — on June 17 it pulled “a full U-turn on gene therapy” and a stock called Unicure Huntington went vertical 80%. One pure play he’s researching: In Silico Medicine, Boston-based but Hong Kong-listed, using AI “specifically and only to generate new drugs.” Following the Soros/Druckenmiller rule — “if you have an idea, you got to allocate a little bit and then do the research” — he put 3% of the portfolio into ARKG and plans to research further. And pushback on a TBPN take that AI won’t solve health itself: “that’s actually the complete opposite… we have way too much data that we actually haven’t really analyzed” — it’s like the ocean.
• Nuclear, via the FT’s “Trump administration nuclear bros” piece (terrible headline, fun-sounding job): new reactor companies drive demand, but the trade is companies, not the commodity — if uranium demand goes 10x, miners can invest to produce 20x, so revenue 20x while the spot price falls. “I would never advocate for anybody to bet directly on commodity prices except for gold.” Hence URA — off a ton from highs, he’s been buying — with a call of “a 3 to 5x” as the nuclear revolution narrative hits “in the next 3 to 5 years.” Caveat as stated: Russia coming back online would be bad for uranium prices but good for US refiners.
• Closing question from chat — does capital rotate from digital to physical? “Absolutely 100% it does.” Since Facebook, capital went almost entirely to improving life digitally; the US is ~3x richer in GDP than two decades ago with little infrastructure to show for it. AI plus robotics plus “dynamism funds” means grid upgrades and beautification — his childhood Meridian Hill Park fountain restored from disrepair — while collapsed engineering costs push budgets into marketing and IRL events: “IRL is going to come back in a big way.”