Is This The Everything Rally?

Summary

• Avi’s frame: this is a lockout rally — bears sold the Iran-war dip for a “fundamental” story (escalation, Strait of Hormuz, global recession), and when the story died they doubled down instead of re-entering; now they face all-time highs nobody likes to buy. Buying strength at highs is “one of the best most repeatable strategies” precisely because it’s psychologically hard. NASDAQ +1.43%, S&P +1%, Bitcoin back above 81k, his own book +10% on the week.
• The core call: a tiny bubble, fundamentals ahead — bubbles form “when perception outpaces fundamentals,” and today perception lags reality: only ~1-1.5% of the population pays $20/month for AI while every chip name (Nvidia, Intel, AMD, SanDisk) crushes earnings. Stay long AI, energy, and crypto until one of three things breaks: inflation running from 3.3% toward 4% (“we’re f*ed”), a collapse in top-10% consumer spending, or genuine overinvestment.
• The tradeable tripwire, offered for notebooks: “the first time that there is an earnings miss from one of these companies will mark a massive downturn in this sector” — and since analysts (“Group Think Inc.”) get fired for being wrong as outliers but not wrong with the pack, estimates stay too low and the herd will never call the top first. Sell early: “it’s always staircase up, elevator down.”
• Jonah’s framework: shocks you fade, trends you ride. Iran was a supply shock like Ukraine; AI is a structural trend like 2014 shale or China’s 2000s demand wave. The only bearish mega-trend he can construct is a global political mandate for redistribution — hence his standing call, “fade everything except socialism.”
• Memory supercycle versus Korean froth: Jonah’s rule of thumb says 25-30% undersupply quadruples memory prices, and the forward 12 months are undersupplied 45-50% — so Samsung and SK Hynix are “ridiculously underpriced,” Hynix “an easy two bagger” even after +500% off the lows. Avi’s caution: Korean margin loans have doubled in a year, and a 15-20% down day in Korea might be a useful proxy for US positioning being over its skis.
• Alt season with fundamentals attached: Zcash +80% in days on whale rotation (the privacy-versus-memes call intact — Pepe “basically at zero”), TON doubled as Pavel Durov reclaimed it from a foundation Avi calls “a relic of the past” with $7 in sight, VVV pitched as “a private ChatGPT.” Jonah’s brake: “buy it for a trade, but don’t set it and forget it” — and the a16z crypto $2B and Haun $1B raises happened “because they can,” not because altcoins get a bid.
• Avi’s Intel confession: bought at a $42 average, sold half at 70 on a target set before AI agents exploded — and agents run on CPUs, Intel’s core business. “I’m just sharing an L with you guys”: rewrite your position as the thesis changes.
• Saylor selling is bullish: funding Stretch with modest Bitcoin sales “inoculates” the market against its biggest fear — that Saylor is the only marginal buyer. As long as BTC holds 78 in the coming days, “we can see 90-95 very quickly.”

Deep dive

1. The lockout rally: bears sold a story, not a price

• Avi’s mechanism: sell on price action alone and you can get back in easily; sell on a “fundamental” story — the Iran war escalates, Hormuz closes, global recession follows — and when the story doesn’t play out, “most people just double down.” Now they carry a double whammy: the thesis died and the tape sits at all-time highs, which nobody likes to buy.
• His counter-discipline: buying all-time highs “is one of the best most repeatable strategies that you can make in the markets” precisely “because it’s so psychologically difficult” — edge lives in betting against psychological bias. The daily practice: forget what you did yesterday and “make a decision based on the current data.” Druckenmiller can’t do it consistently; neither, he admits, can he.
• The tape backing it: NASDAQ +1.43%, S&P +1%, Bitcoin finally back above 81k, TON +24% on the day, AMD crushing earnings, SanDisk down 10% post-print then +35% in a week. Avi’s portfolio is up 10% on the week — evidence “people are still massively sidelined.”

2. A tiny bubble: perception is still running behind fundamentals

• The definition Avi trades on: bubbles form “when the perception of reality is far ahead of what’s actually happening” — and today the dreams lag. Only ~1-1.5% of the population actively pays $20/month for an AI subscription, against Jonah’s “everybody I know pays for an AI subscription or three or five”: “we’re in a tiny little bubble in a tiny little world and there’s a big world out there.”
• Every memory, CPU, and GPU stock — Nvidia, Intel, AMD, SanDisk — is “crushing earnings” because “the street doesn’t understand… people are underestimating the impact of AI on the economy.” He invokes Andrew Kang’s “exponential horizon” and the book Technological Revolutions and Financial Capital: financial capital initially underinvests in a technological revolution. “In 2 to 3 years the world is going to look very different than it looks today.”

3. Jonah’s commodities lens: shocks you fade, trends you ride

• The reduction: ask whether the cause of a selloff is temporary or structural. Shocks — Ukraine, Iran — are temporary disappearances you fade; trends — 2014 shale taking Brent from $110 to $27-30 by 2016, China lifting a billion people out of poverty — you ride. But “you can get absolutely carted out on a stretcher fading a shock” too early or with the wrong trade expression.
• His Iran expression: not selling $90 crude calls but buying assets he liked — SPY, computer-chip stocks — now “blasting through all-time highs.” The victory lap, at the risk of sounding obnoxious: “fade the blips, ride the trend.”
• The only bearish mega-trend he can construct is a global sweep of political mandates for redistribution — hence “fade everything except socialism.” The bullish trend on the other side: 1% of the world pays for “a constant never-resting PhD intern… what if that number goes to 5%, 10%? What does that do to SPY? What does that do to Micron… to Nvidia?” Surfer test: Iran was a wave you duck-dive; AI is one you paddle into and stand up on.
• Avi’s addendum: most trading disagreements are really horizon disagreements — they had publicly said the Iran war would resolve in 4-6 weeks, “believing Trump probably had a plan,” and it did.

4. Three rally-killers, and why you sell early

• Avi’s checklist of what ends this: (1) inflation comes roaring back — “we are at 3.3%… if we go to 4% we’re f*ed”; (2) top-10% consumer spending collapses — not happening while stocks rise, and the notion the rich save at much higher rates than the public “was debunked”; (3) genuine overinvestment — SanDisk grows 10% when the street bet on 20%.
• The concrete tripwire: “The first time that there is an earnings miss from one of these companies will mark a massive downturn in this sector.” And you probably can’t wait for the miss itself — start lightening up when the beats get thin.
• Why early beats late: “it’s always staircase up, elevator down.” Jonah’s objection — “this feels like an elevator up” — draws the retort: “when you see what happens on the way down, you’re going to realize that we’re on the stairs… all I can tell you is it’s not going to be pretty.” Until then: long AI, long energy, long crypto.

5. Group Think Inc.: earnings beats measure analyst cowardice, not corporate magic

• Jonah’s reframe of the beat parade: a “beat” is against “the published prediction of CJ Muse at Cantor Fitzgerald, Timothy Arcuri at UBS, Blaine Curtis at Jefferies… who are these people? Why do their opinions matter?”
• The incentive structure: wrong as an outlier gets you fired; wrong with the pack is safe — same for the trillions in passive money keyed off the estimates. So “the estimates should be a lot higher, but nobody wants to be an outlier” — a “mega pocket of opportunity” for retail investors with risk tolerance willing to hold outlier projections.
• The corollary both hosts sign: “Group Think Inc.” will keep projecting beats long after they shouldn’t and “is not going to be the first to get bearish” — which is exactly why Avi can use the first miss as his bubble barometer.

6. Korea: froth barometer meets memory-supercycle math

• The move: Samsung +14%, SK Hynix +10.6% — tradeable now through Interactive Brokers, where Korean tickers are literal number strings (“you’re buying stock 1834267”). SK Hynix says chip demand exceeds manufacturing capacity; Q1 revenue rose 198% to 52.6 trillion won with 37.6 won operating profit, and ~44% of the KOSPI’s value is this AI complex — memory, packaging, racks, “the unglamorous infrastructure.” One month ago consensus was bearish Korea on Hormuz energy costs.
• Avi’s first froth sighting: Korean margin loans have doubled over the last year. Like an open-interest-driven Bitcoin move: “Is there real buying or is it just gambling?… Is it a shock or a trend?” His watchpoint: a 15-20% down day in Korea could be a good proxy for US positioning being over its skis — not an automatic exit signal, but the market he’s watching.
• Jonah’s counter is the memory rule of thumb: forward-12-month undersupply of 5-10% doubles prices, 15-20% triples, 25-30% quadruples. The coming 12 months are undersupplied 45-50% — implying 400-700% profit growth. Samsung and Hynix are “ridiculously underpriced”; Hynix, already +500% off last year’s lows, “will rally another 500%. Or it could” — “an easy two bagger,” safe at least for the first 50-100%.
• Avi’s meta-observation: CT has traded Korea since January, and “when CT gets into something it’s over” is no longer true — crypto traders are just faster at narrative-chasing, and the whole market now trades their way: “reading headlines, reacting to headlines, and allocating capital based on short-term trends.”

7. The Intel confession: agents are a CPU trade

• Avi bought Intel at a $42 average and sold half at 70 — his original target, set before agents blew up. The mechanics: models live on GPUs, but agents — formatting output, making API calls, sending messages through Telegram — sit in the middle and run most cost-efficiently on CPUs, a major Intel line of business.
• The lesson, told against himself: he never updated his priors, so he sold on a stale target. “I’m just sharing an L with you guys. Not everybody trades perfectly.” The retained half has since grown back to the size of the whole position at 70. Jonah’s grade: “You called it dead right. And you called it publicly” — flowers, not an L.
• One taxonomy correction from Jonah worth keeping: GPUs are generalized for parallel compute; the truly specialized chips are ASICs and Google’s TPUs.

8. Alt season with fundamentals: Zcash, TON, VVV

• Zcash is up 80% in a few days with whale rotation behind it — the payoff of their start-of-year “privacy versus memes” call: Pepe is “basically at zero” since, while ZEC sits only 20-30% off its highs (Monero likewise). Extended here — Avi wouldn’t go all-in but would be “a big buyer” on a reversal toward 400.
• TON doubled in 2-3 days on Pavel Durov returning and shunting the Ton Foundation aside. Avi’s structural read, crediting Mike Dudas: “foundations are a relic of the past” — legal cover from the 2019 fundraising mess — and “if Ton can trade to $7 when Pavel is not even able to directly work on it, it can probably trade to $7 again.”
• VVV is “effectively a private ChatGPT… ridiculously important infrastructure” — his screen being “good companies building important products for the world that just happen to be associated with crypto,” a club he says includes Hyperliquid and now TON.
• Jonah’s brake: these pops won’t create sustainable investment vehicles — “buy it for a trade, but don’t set it and forget it.” Bitcoin is exogenous (“still so bullish over the long haul I can’t see straight” — a phrase Avi needles him for using before a 50% drawdown), but with rates high and a Western affordability crisis he doubts durable speculation returns. Avi’s twist: violent alt moves themselves bring attention back, and stock-market money could leak in “if the markets stop ripping so much.”

9. The $3B VC puzzle: raised “because they can”

• a16z crypto raised $2 billion, Haun Ventures $1 billion, and neither host can locate the target — “billions have been poured into crypto and it hasn’t produced a single useful product.” Avi, sponsor-risk acknowledged: “the vast majority of crypto products are completely useless,” and the industry’s biggest exit — Bridge’s $1B sale to Stripe — “doesn’t even cover 20% of their fund.”
• Verdict: the money will probably go mostly to growth rounds (a Ripple-scale deal, perhaps), probably not seed deals, and probably isn’t accretive to altcoins — though it might pump public companies that embed stablecoins. Dixon’s posted rationale (likely Chris Dixon) — better infrastructure, the historically interesting point in tech cycles — “doesn’t explain why you would want to deploy $2 billion into this market with literally zero apps and zero uptake.” They plan to book an a16z crypto partner “and ask them what the hell they’re doing with that money.”

10. Saylor’s sale as vaccination — and the levels

• Avi’s closing solo thesis: Saylor floating Bitcoin sales to fund Stretch is bullish because concentration risk was the standing objection — allocators feared Saylor was the only marginal buyer and the market would collapse if he ever sold. A small, survivable sale is inoculation: “the market generates an antibody defense to the concept of me selling and Bitcoin can go much higher than it would have gone previously.”
• The line in the sand: if Bitcoin doesn’t go below 78 in the next few days, “I’m very bullish. I think we can see 90-95 very quickly” — and BTC’s refusal to sell off on the Saylor news is already part of the evidence.