AI Cracks: Where Do You Hide?

Summary

• Avi describes the trade’s second leg as hyperscaler spending: demand → hyperscalers spending $600–700B on data centers (Google’s $80B equity raise included) → that value is now basically baked in and cracking. “It’s all about what are these companies spending their money on — that’s what’s driving the market forward.” The next hiding place is what AI actually changes in the real economy: short Accenture, long biotech (XBI, ARKG, BLLN, Cure). “If the market’s going to crack… you kind of have to hide there.” He’s explicitly “not a doomer” — late innings, but “it’s the late innings of the trade that often provide the most insane returns.”
• Avi’s saturation tell, as told: at a Tablet Magazine awards event he overheard a 13-year-old whose portfolio held Google, Uber, and the hyperscalers — “if 13-year-olds are buying single-name stocks in the hyperscalers, maybe the trade’s over.” But the Hollywood guy hadn’t taken part in the memory rally: “your taxi drivers are not buying memory stocks yet,” and peak bubble is “every 13-year-old buying SanDisk.” He told the kid to buy Intel.
• “Bitcoin’s cooked, Ethereum’s cooked. Everything that doesn’t make revenue right now is cooked” — Avi’s thesis is vol competition: “as long as there is a trade that has higher vol… you just can’t buy Bitcoin,” and SanDisk going +40% in days is that trade. Bitcoin broke 60 live on air (“we’re probably heading down to the 50s”); his range-re-entry short from ~65 targets a 49–53 cover. Buy list for the collapse: Cards (likely Cardano) under 10 cents, Hype under 45 maybe 40, Zcash ~280; Solana to 46, Worldcoin short, “Fartcoin is quite literally going to zero.”
• A Saylor blow-up “feels inevitable” — STRC is depegging because paying the dividend requires selling STRC, MSTR, or Bitcoin, and each path is self-defeating; the depeg prices the probability he pauses the dividend and kills the product. Jonah’s kicker: when the blow-up finishes, “the story for Bitcoin will never have been better” and the next trough is “probably the best trade of my lifetime” — his GBTC-after-FTX playbook repeated. Million-dollar target intact, but he’s not rebuying: “it’s not like Bitcoin is going to hedge you. It’s going to go down twice as fast as the Nasdaq.”
• The biotech disagreement is the episode’s best exchange: Avi owns XBI/ARKG 70/30 as a bet the whole industry re-rates on AI drug discovery; Jonah won’t touch spaces “where fundamentals matter and you have no edge” and poses the test — “ask yourself how you’ll feel if you’re down 50%.” Avi concedes it’s “a brilliant question” for every long-horizon trade.
• Politics as capital flows: the New York election cycle is socialism arriving via a wedge issue, and a 2028 blue wave would matter for assets. Jonah’s tradeable framing: “sell what the socialists are going to redistribute and buy into whatever monopolies they’re going to create.”

Deep dive

1. Read volatility to know what regime you’re in

• Avi’s defense of technical analysis: it works “when there’s a lot of retail in an asset and there aren’t many fundamentals” — which is why Bitcoin “traded basically technically perfectly for many years until the quants came in.” In crypto, buying breakouts above all-time highs is a high-hit-rate strategy because crypto is a momentum asset, and high-vol regimes there are momentum regimes — the opposite of equities, where high vol means chop, like Nasdaq going sideways since May 15th.
• Avi, an options trader, maps vol by asset class: equity vol rises when prices fall (that’s where the nest eggs are); crypto vol is “a barbell” — extreme on euphoria and on 2022-style unwinds; and oil ruined market makers for decades because it’s stable in dollars per barrel, not percentage — a barrel moves about as many dollars a day at $30 as at $100.
• The payoff: “you’re not going to get heightened upside volatility on mild greed.” With charts showing hyperscaler free cash flow heading toward zero after a two-decade uptrend, Jonah reads the memory-stock action as “white hot upside volatility… the behavior of the Nasdaq in 1999” — room to run, but “starting to get into the scary zone, the later innings”: buy the dips only with tremendous conviction, otherwise sidelines.

2. Ask how you made your money — then follow the spend downstream

• Avi’s anatomy of the trade: AI demand exploded, hyperscalers committed $600–700B to data centers (Google raised $80B of equity for capex), and that value got “baked into the price… you saw them shoot up in a straight line. And now you’re seeing cracks.” What holds the market up now isn’t Google/Meta/Amazon price action — “it’s all about what are these companies spending their money on.”
• The second leg is where the hyperscalers are spending; the third is fiber-optic cables and downstream beneficiaries (optoelectronics like AAOI, energy — “basically everything has gone downstream”). Avi then looks for what AI changes in the real economy: short Accenture (“consulting firms are probably going to be less in demand”), and buy biotech — more XBI, more ARKG, plus BLLN (Billion to One, via a friend who’s “smart on this”) and Cure.
• The stance is explicitly not bearish: “I’m not a doomer… I think that we have time” — pending Micron earnings that day — but the mentality must shift from “demand for AI going higher” to industries AI actually revolutionizes. “If the market’s going to crack… hyperscalers are cooked, memory is cooked… it’s what is AI actually going to affect in the real economy. And then you kind of have to hide there.”
• Jonah’s caveat on playing the spend — worth keeping: “Google, Microsoft, Meta, they’re not buying Micron’s stock. They’re buying Micron’s products.” Buying rallied, hyper-volatile proxies for high-bandwidth memory is not the same trade, “so tread lightly.”

3. The 13-year-old portfolio test: hyperscalers saturated, memory not yet

• Avi’s specimen anecdote: at the Tablet Magazine awards he overheard Alan Loeb — writer of Wall Street 2: Money Never Sleeps — advising a 13-year-old on his portfolio, which held Google, Uber, Nasdaq, and the hyperscalers. “If 13-year-olds are buying single-name stocks in the hyperscalers, maybe the trade’s over.”
• But when Avi suggested memory stocks, Loeb pushed back — “don’t you think they’ve run up far and far too fast?” — while admitting he hadn’t taken part in the rally. Avi’s read: the hyperscaler trade is saturated but the memory trade is “still maybe not fully penetrated” — “your taxi drivers are not buying memory stocks yet.” Peak bubble looks like “every 13-year-old is buying SanDisk in their account,” and this kid owned none. He told him to buy Intel (“really hope that works out — not financial advice”).
• Jonah agrees from the crypto playbook — the bubble anatomy is your Uber driver asking “should I buy XRP here?” — and on price-to-forward-earnings for the component makers, “I don’t think we’re at peak euphoria yet.”

4. Crypto is cooked until it’s forgotten — then buy the list

• Avi, unhedged: “Bitcoin’s cooked, Ethereum’s cooked. Everything that doesn’t make revenue right now is cooked.” The mechanism is vol competition — “as long as there is a trade that has higher vol, that is more interesting than crypto, you just can’t buy Bitcoin. And that’s AI right now… SanDisk can go up 40% in a few days. Nobody’s buying Bitcoin.”
• Live on air Bitcoin broke 60 — “we’re probably heading down to the 50s.” His short thesis from ~65: trade a range (Feb 2–Apr 6 on the weekly), break out, then re-enter and retest — “this is a phenomenal short,” cover 49–53. At current levels he’d rather short Worldcoin (oddly near recent highs), sees Solana to 46, and “Fartcoin is quite literally going to zero.” The accumulation list: “wait for the crash… buy Cards [likely Cardano] under 10 cents, buy Hype under 45, maybe 40,” Zcash around 280 — “that’s how you get really, really easy two-to-three-x’s,” because crypto is “so bipolar… when they hate it, at some point they’re going to love it again.”
• The asymmetry trade for the timid: long Hood, short crypto — if crypto rips, Hood does extremely well; if crypto bleeds, Hood can still outperform because revenues are expanding beyond crypto.
• The process lesson Avi wants remembered: get every thesis “to a yes or a no,” set alerts at your level, and when it hits, “don’t change your mind. Some of the best trades I’ve ever taken are just where it looks the worst.”

5. Saylor’s blow-up is the setup for the trade of a lifetime

• Avi’s STRC depeg mechanics: paying the dividend requires selling one of three things — STRC itself, MSTR, or Bitcoin — and each is self-defeating: selling Bitcoin risks collapsing Bitcoin, selling MSTR sinks the stock, and selling Stretch drives the depeg because “people don’t really believe he’s going to be able to continuously sell these equity products to pay the dividend.” Endgame: he pauses the dividend, which makes Stretch “kind of a useless product” — the depeg is the market pricing that probability. The jab: “Saylor designed it with ChatGPT.”
• Jonah’s version — “designed with a couple of friends over a table of God knows what at 3:00 in the morning”: “we’re going to buy Bitcoin with more money than we have… take money from bondholders and give it to equity holders, and also sell our Bitcoin when we run out of money, but Bitcoin’s going to go up forever.” Verdict: “very poorly concocted financial engineering.”
• The consequence is the bull case: “he’s going to blow up. When he finishes blowing up, the story for Bitcoin will never have been better” — best regulatory framework in history, price at “maybe not COVID lows, but like FTX lows.” Jonah’s precedent is his own best trade: GBTC at a huge discount in December 2022–January 2023 after FTX. He lightened up from the 124k highs, keeps his million-dollar target, but won’t rebuy yet: “if the broader market implodes, it’s not like Bitcoin is going to hedge you. It’s going to go down twice as fast as the Nasdaq.”

6. The biotech clash: industry re-rate vs. no-edge discipline

• Jonah’s refusal — as told, not smoothed: “biology happens to be a weak spot of mine… I don’t vibe with the mitoplasms and the chronoblasts.” His rule from oil trading: “you can’t succeed unless you know the fundamentals” — biotech has real fundamentals and PhD-holding VCs with real edge, and “I ran over people once I knew how to profitably trade oil fundamentals. Before… I was adrift.”
• His test for the AI-biotech thesis: if “number go down,” you have no framework — is it the FDA, a molecule issue in your ETF’s biggest holding? “Ask yourself how you’ll feel if you’re down 50%,” knowing this isn’t a meme coin where nobody has information — here there are real fundamentals and real information.
• Avi calls that “a brilliant question to ask of quite literally every trade,” then defends the position: he can’t pick single names either, so he owns XBI and ARKG 70/30 — an index bet that “the entire industry will re-rate because we are going to be able to produce more drugs at a cheaper cost.” ARKG is +8% since last week’s show, has “based for 4, 5 years,” and he’s applying “the Druckenmiller approach — you buy a little bit to force yourself to do research. I’m still in the middle of the research.”

7. Socialism as a capital-flows problem, not a culture war

• Avi flags his own bias up front (“I know my bias… I’m not here pretending I don’t have it”) and reads the New York election as Palestine used as a wedge issue “to shoehorn in communism” — nothing changes for Palestine, but “more communist policies are going to be passed in the city.” His theory, via Fukuyama properly read: once liberal governance is solved, “if life is too good, people will find struggle anywhere” — the watch parties are “transplants from Ohio” seeking identity.
• Jonah converts it to markets: the set-and-forget S&P works “until the socialists take over,” and 2028 is now “a clear and present danger” — Hillary Clinton and Obama “seem like real centrists” next to the new wave. In socialism, wealth flows to politicians, gangsters, and oligarchs — his specimen: the father of someone he dated held a cronyist monopoly on the buns for a Soviet Burger King equivalent. Recommended reading: The Oligarchs and Putin’s People (Jonah), The Empire of Wealth by John Steele Gordon (Avi).
• The tradeable symmetry, in Jonah’s words: just as you “sell what the hyperscalers are selling, buy what they’re buying,” if socialism creeps in, “sell what the socialists are going to redistribute and buy into whatever monopolies they’re going to create” in the name of the utopian common good.