Daily Briefing

TL;DR: The optimists and the pessimists on AI now agree about the next six months. Even the most bullish voices we track — Raoul Pal and Jordi Visser, who argue the AI boom is a durable "supercycle," not a bubble — now expect a 3-6 month digestion period while roughly $4 trillion of new stock offerings (SpaceX, OpenAI, Anthropic and others coming to market) absorbs investors' cash. We moved two probabilities today: the odds of an oil price spike by September were cut from 45% to 40%, and the odds that this wave of new stock supply weighs on the broad market were raised from 55% to 60%. 4 new interviews processed, 1 failed (retries tomorrow).

Each standing view carries a probability — the odds we currently assign to it being right. Moves in those numbers are the day's real news.

• New stock supply (thesis 6): 55% → 60%. The analysts at Forward Guidance (a macro research show we track daily) counted the new offerings: SpaceX $75B + Google $80B + Oracle $40B + CoreWeave $3.5B — companies are now issuing more new stock than they retire through buybacks for the first time in 10-15 years. This matters because every dollar of newly issued shares must be bought with cash that would otherwise be supporting existing prices. And since the AI bulls themselves expect months of digestion, both sides of the debate now point the same direction for the near term.
• Oil spike (thesis 3): 45% → 40%. A third independent analyst team now argues the administration is deliberately engineering de-escalation with Iran — to protect the SpaceX stock offering and avoid +80% gas-pump prices going into the midterm elections. A peace deal means cheaper oil and less chance of our spike scenario. The levels that would change our mind are unchanged: US oil (WTI) closing back above $89 starts re-confirming the spike risk; above $107 confirms it.
• AI (thesis 1): 65% unchanged, but the counter-argument got stronger. Our standing view is "a real bubble, but with time left." Pal and Visser made the best case yet that it is not a bubble at all: price-to-earnings ratios have actually *fallen* because earnings kept pace with prices, and data centers are only ~30% built versus plans. One warning both sides agree on: if AI software becomes efficient enough that big tech cuts its hardware spending, every AI-adjacent investment — power, memory chips, networking — cracks at the same time.
• Bitcoin (thesis 7): 65% unchanged. A new source of selling pressure was identified: bitcoin miners are selling their coins to fund a pivot into AI computing — a steady seller that has nothing to do with the usual market cycle. Visser, for his part, stays out of bitcoin until the price breaks back above its long-term trend line and the AI-spending rotation is visible.
• Fed (thesis 4): 55% unchanged. Markets currently price only about 2 more rate hikes through mid-2027. One analyst argues that is the most hawkish it can get — meaning if even those hikes fail to materialize, interest rates fall and bonds and gold win. That is an asymmetric bet: small downside if wrong, large upside if right.

When the most bullish and the most bearish people we track agree on what happens *next* — a multi-month digestion while $4 trillion of new stock lands — and disagree only about what comes after, the near-term action carries unusually high confidence. That action: trim the most crowded AI-infrastructure positions.

"vs 200-day avg" compares each price to its own 200-day moving average — a standard trend line: above it is an uptrend, below it is a warning.

Interest rates and the Fed — from Forward Guidance (hosts Tyler, Felix and Quinn, macro analysts):

• Quinn: betting that short-term interest rates fall (via SOFR futures — contracts that pay off if rates drop) is "the most asymmetric trade" available: markets price ~2 hikes by mid-2027 as the ceiling, so the bet wins if rates are cut *or* if the economy merely softens. Gold is the companion position.
• Quinn: market-implied inflation expectations are falling while bond yields hold steady — that combination means *real* (after-inflation) borrowing costs are rising on their own. The economy is being tightened without the Fed lifting a finger, and real wages are already negative.
• Felix: incoming Fed chair Warsh uses an inflation measure that strips out energy — giving him cover *not* to raise rates at his first meeting. Anyone positioned for a hawkish debut may be surprised.
• Tyler: the administration's repeated pattern — manufacture a crisis, spike fear, then retreat and let markets squeeze higher. When the options market's fear gauge inverted recently it was, in his words, "one of the greatest contra indicators ever."

AI — the case it is NOT a bubble — from Raoul Pal (macro investor, Real Vision founder) and Jordi Visser (40-year markets veteran, former hedge fund chief):

• Pal: demand for AI computing still exceeds supply, so today's bottlenecks *raise* the required buildout spending rather than ending it; he is "struggling to see how this is not a supercycle."
• Visser: data centers are only ~30% built versus announced plans. For scale: Google's $85B capital raise exceeds the combined market value of the bottom 360 companies in the S&P 500. He sees traditional software companies as the structural losers, and drugmaker Eli Lilly's weight-loss-drug cash flows as what finances the next generation of biology-focused AI.
• Visser: silver is the scarce input (solid-state batteries); copper, the consensus play, is now crowded — Tesla's switch from 12-volt to 24-volt wiring cut its copper use 70%. Pal's gloss: efficiency finds a way around every shortage.
• Bloomberg reporting: London has become Europe's AI hub ($18B raised in 2025, more than half of Europe's Q1 AI funding); energy costs and labor rules are the binding constraints.

Bitcoin:

• Quinn: miners selling coins to fund AI pivots is a structural seller the usual cycle math ignores, and the professional arbitrage trades that normally absorb supply are "falling apart."
• Visser: staying out until price reclaims its long-term trend line AND the capital-rotation is visible — he does not endorse the October-low timing call others make.
• Pal: small investors capitulating is the standard sign a bottom is forming.

Oil — from Forward Guidance:

• All three hosts lean bearish near-term: a deal is the administration's preferred outcome, demand destruction from $4+ gas is underappreciated, and a reopened Strait of Hormuz would push inflation down on its own. The opposing view (analyst Gordon Johnson: no durable deal is likely, the physical risks remain) stays on the books as the dissent.

We group holdings by what actually drives them, not by sector label — positions that crash together count as one exposure.

• Exposure groups: digital assets 37.7% of the account (MicroStrategy + the bitcoin ETF + the STRC preferred), AI 23.7%, oil & gas 11.1%, biotech 7.9%, cash 19.6%. In the worst combined stress scenario the account would be down roughly a third — that is the cost of the concentration.
• Sharpest conflict (third day running): Tesla is 15.4% of the account, sits below its trend line, and the model ranks it bottom of its peer group — and both AI camps now expect the air pocket its sleeve sits in.
• Biggest gap vs plan: biotech at 7.9% versus a 12% target (~$134k to add) — and a third independent healthcare argument landed today (Visser's Eli Lilly case above). The healthcare sector is already leading the market.
• Model verdicts: Bitcoin is in the model's buy zone (price ~half of fair value, crowd at extreme fear) — the play is steady scheduled buying from the cash sleeve, not a lump sum, with the new miner-selling pressure as the watch item.
• Dry powder: cash is 19.6% against a 1% target, and T-bills are 0% against 8%. The roster still expects one more washout (gold toward $3,000-3,400, bitcoin into the fall); parking cash in T-bills earns ~4% while waiting and costs nothing.

7. Experience becomes a liability when the rules change. — Jordi Visser. "The smarter you are, the more you know, the worse you are during this time... your brain is taking this pattern and connecting it to too many other patterns." Deep familiarity with 1929 and the dot-com era causes seasoned investors to exit early on false alarms precisely when the economy's underlying machinery has changed. Apply: when the regime itself is in question, trust current primary data over historical rhymes.

8. The cure for the big spenders' pain is the kill switch for everyone downstream. — Tyler, Forward Guidance. The only thing that fixes big tech's underperformance is cutting the AI spending the market is punishing them for — and that same cut breaks every business feeding on the spending (power, memory, networking). Apply: for every second-order investment, define in advance the primary-spender signal that ends it.

• SpaceX's first trading day (today): how the market absorbs it; insider shares unlock ~Dec 2026, with ~$3T of similar unlocks industry-wide by end-2027.
• The Fed's rate meeting — Warsh's debut: does he validate the ~2 priced hikes or use his energy-stripped inflation measure as cover to hold.
• US oil (WTI) $89 and $107 — the two levels that resolve the oil thesis; Iran deal finalization vs the unresolved red lines.
• Bitcoin: miner selling flows, and whether the $63k zone holds above the February lows.
• Memory-chip spot prices (TrendForce): the daily tell for whether the AI-hardware boom is cooling.

Pulled 4 | blocked 0 | errored 1 | unavailable 0. Errored: Mario Nawfal — "Trump Wants Peace or Kharg Island" (video unplayable, likely a live stream still processing; retries next run). Proxy path clean otherwise. Caliban reachable — 1 new chart pulled (the giants-vs-average-stock migration).

• Forward Guidance — Policy Intervention Is Keeping The Bull Market Alive: https://www.youtube.com/watch?v=9z4YYo9BQ08
• Raoul Pal x Jordi Visser — The AI Supercycle: https://www.youtube.com/watch?v=jGNmWFeCIWE
• Raoul Pal x Jordi Visser — The Scarcity vs Abundance Trade (backlog fold-in): https://www.youtube.com/watch?v=v-001JkOa6M
• Bloomberg Tech — AI Boom in London: https://www.youtube.com/watch?v=AFe8WpulQ0s
• Bloomberg Tech — Osborne/OpenAI on Governments Adopting AI: https://www.youtube.com/watch?v=TwycqbMBq_I