[BidClub_]
1000x · · 60 min

What The Anthropic IPO Means For The Market — Will We See A Rebound?

Avi FelmanJonah Van Bourg

EquitiesCryptoAI & SoftwareSemisInvesting
YouTube ↗
TL;DR
  • Anthropic’s IPO delivered a debut pop without reversing the AI selloff. Avi cites shares at $98 against what he thinks was an $83 issue price, while the NASDAQ reversed from roughly 2% higher to down 75 pips. He worries buyers positioned for the listing will exit without an end-of-day bounce. Jonah warns that proceeds diverted into real estate rather than recycled into AI names could become a “liquidity suck for the overall environment.”
  • The central disagreement is whether AI faces impaired demand or “white hot bull market volatility.” Avi sees three connected risks: insufficient revenue growth, stalled U.S. data-center construction and tighter chip exports to China. Jonah questions the revenue-plateau premise and argues that even cheaper Chinese models can lift American companies’ productivity; Avi’s rebuttal is that “the future can be here and companies can still go bankrupt.”
  • Both see a potential rebound, but their willingness to buy differs sharply. Avi thinks one is close “if not in terms of price, in terms of time,” and says Intel at $42 would be a good place to buy; he cites SanDisk at $577 as its Iran-war low. Jonah says an S&P 500 at 6,350 would trigger selling his T-bills and buying indices with leverage, although he doubts the market reaches it.
  • Micron at $700 is Jonah’s clearest near-term buy. He cites memory shortages through 2030 and relative strength during the selloff, calling MU his favorite opportunity over the next couple of months. His final positioning is restrained despite broader AI optimism: “I’m not selling my longs, but I wouldn’t add” except Micron.
  • Jonah proposes long exposure to Hyperliquid and businesses making money against shorts in those that do not. He raises shorting Trump; Avi says Trump can no longer move markets and identifies Worldcoin rallies as short opportunities. Hyperliquid and Lighter gaining market share support Avi’s conclusion that “reality now matters in crypto,” while Jonah notes Solana could bounce after criticizing its chart.
  • Bitcoin divides Avi’s credit-shock case from Jonah’s lifetime debasement hedge. Avi’s conditional opportunity is AI bankruptcies forcing government intervention, potentially with the Qs 30% lower and BTC at $40,000–$45,000; otherwise, he prefers Micron. Jonah would neither buy fresh BTC here nor sell remaining holdings to time that decline, still expecting $1 million per coin within his lifetime: “That’s my horse, and I’m going to keep riding it.”
  • Cybersecurity is their strongest candidate for rotation beyond crowded AI trades. Jonah favors CrowdStrike and Palo Alto Networks as increasingly capable, unrestricted models expose unpatched systems, sketching security longs two to four times the size of shorts in vulnerable companies. Avi cites a report describing exploitation “with no human in the loop” and questions whether crypto’s irreversible, instantaneous settlement becomes a liability under that threat.
Digest · the substance, structured for research

1. Anthropic’s IPO creates a liquidity test

  • Opening on the IPO, Avi cites Anthropic at $98 versus what he thinks was an $83 issue price, yet the NASDAQ has reversed from roughly 2% higher to down 75 pips. Intel’s move toward $60 has faded to $56. The debut delivered a shareholder pop without delivering the broader-market relief AI bulls wanted.
  • Avi thinks buyers positioned for the listing — Intel was up 7% and SanDisk 6% that morning — may exit if there is no bounce by the close. The failed catalyst could prompt another flush before prices become “genuinely cheap,” even if a rebound is close in time.
  • Jonah says the next test is where IPO liquidity goes. If Anthropic sellers buy California or Texas real estate instead of recycling proceeds into AI equities, he sees a “liquidity suck for the overall environment.” He says continued revenue growth from both Anthropic and OpenAI is needed; the listing alone cannot reinvigorate the trade.

2. AI’s financing chain faces three simultaneous threats

  • Avi’s financing concern starts with indebted neoclouds that borrowed heavily to fund construction. He says revenue is flatlining and reads Anthropic’s S-1 as showing falling amounts paid for compute, partly because each unit of intelligence is getting cheaper and more efficient. There are not enough new corporate deals or token-volume growth to compensate. Building a website is cheaper than three months earlier. His caveat matters: he thinks recent model-capability gains have been limited, and another step change could alter demand.
  • The other tests are access to capital and chip demand as U.S. data-center projects face delays and exports to China face tighter controls. Jonah’s “self-licking ice cream cone” describes Nvidia investing in startups that return the cash through chip or compute purchases. The hosts worry that the CFO might curtail that financing loop.
  • Jonah says he thinks 20% of Nvidia’s revenue was going to Singapore and that a substantial amount of those chips probably reaches China. He describes a report he believes is from Hunterbrook alleging AMD customers circumvent export controls; AMD fell 10% before recovering somewhat. Weaker revenue, stalled domestic buildout and restricted exports together make the case for continued declines.
  • Jonah’s pushback — worth keeping: he does not know whether Anthropic’s revenue is plateauing and calls forecasting its cash flows a “fool’s errand.” Even if cheaper Chinese models displace U.S. suppliers, efficiencies can reach ordinary American businesses; the expression might shift from NASDAQ to the S&P 500. Avi counters: “the future can be here and companies can still go bankrupt,” as railroad overbuilding showed.

3. A rebound can coexist with a damaged AI trade

  • Jonah calls the selloff “white hot bull market volatility,” comparing it with NASDAQ pullbacks of 20–30% in 1999 and Bitcoin’s 2021 reversals. He thinks forced selling by Leopold Aschenbrenner and leveraged trading pods amplified the decline. He expects the market to bottom and rip back, while warning that prices could go lower first. Avi, meanwhile, says he has “never been this confused.”
  • Avi agrees a bounce is near, “if not in terms of price, in terms of time,” but finds the market “fundamentally scary” to buy. Jonah notes Nvidia is off 40% from its highs and Intel fell 30% in a day. Jonah’s counterweight to bearish fundamentals is that sophisticated investors may have sold weeks before the public learned the reasons, leaving the immediate trade dependent on who remains to buy or sell.
  • Avi’s concrete reference points are Intel at $42, which he calls a very good place to buy, and SanDisk at $577, its Iran-war low. He says fundamentals are better than in April even if worse than two months earlier. Jonah later says he is looking to buy March lows in specific assets and would like SpaceX around $100. For investors already holding shares, Avi suggests selling calls against them while implied volatility is “through the roof.”
  • Jonah’s extreme downside plan is explicit: S&P 500 at 6,350, around the April lows and roughly unchanged since mid-2025, would make him sell all his T-bills and buy NASDAQ and the S&P with leverage. “I just don’t think it’ll get there.” His bullishness still carries an AI cyber-catastrophe risk he cannot dismiss; he accepts it because “you have to underwrite something.”

4. Micron becomes Jonah’s strongest near-term conviction

  • Avi’s retrospective warning is Nvidia’s stalled leadership: it struggled around $230, briefly hit $241, and traded sideways from May through October while downstream AI trades accelerated. “This is all hindsight,” he admits; he did not call it then. His analogy is Bitcoin going sideways while altcoins surge — broader participation can mean liquidity has dispersed into riskier bets, increasing volatility as a stage of the bubble nears its end.
  • Jonah’s preferred expression is MU at $700: he favors Micron over Intel and SanDisk, citing shortages in its memory modules through 2030 and observations from Electronica in Munich. Some B300s are floating around, and server racks can be obtained at a very high price, he says, but most of the components needed to “make sand think” remain unavailable. Micron’s strength against otherwise weak AI shares reinforces his decision to keep scaling in.
  • Jonah’s final positioning: “I’m not selling my longs, but I wouldn’t add to them here, other than Micron, of all things.” He expects Google, Meta, Microsoft and SpaceX to do well from these levels, while risky neoclouds such as Galaxy can keep falling. Separately, Avi says that over the next couple of years, BTC on dips could offer the best risk-reward; over the next couple of months, he prefers MU.

5. Crypto’s money makers separate from narrative tokens

  • Jonah proposes going long Hyperliquid and a basket of things that make money, and short a basket of things that do not. He raises shorting Trump; Avi says Trump no longer has the ability to move markets. Avi calls Worldcoin a “screaming short” whenever it rises. Jonah describes Solana’s chart as ugly and warns it could bounce after he says so; Avi adds that Ethereum’s chart also looks “horrendous.”
  • The interesting reversal, for Avi, is that “reality now matters in crypto,” perhaps more than in a stock market still dominated by narrative and positioning. He says Hyperliquid and Lighter are going up because they are gaining market share, while tokens with adverse flows keep declining. Easier short access through Hyperliquid makes expressing that distinction more practical than when they first discussed shorting altcoins.

6. Bitcoin divides crisis timing from lifetime conviction

  • Avi lays out a Bitcoin bull case he says he has heard but does not know how much he agrees with: fiscal irresponsibility and excess government liquidity are Bitcoin’s best backdrop. AI-related bankruptcies could implode credit markets, bring government intervention and put the Qs 30% lower. In that scenario, BTC at $40,000–$45,000 would be his “buy of a lifetime.”
  • Absent that shock, Avi sees no special reason to own Bitcoin: if markets stabilize, “I’d much rather own Micron than Bitcoin.” He also says Bitcoin is a “great liquidity sponge,” that liquidity is coming, and that he does not think the government will let the American economy go to the woodshed over the AI unwind.
  • Jonah would neither buy fresh Bitcoin here nor sell his remaining holdings merely to repurchase at $40,000–$45,000: realizing a large capital gain to attempt that timing is “getting too cute.” He still expects $1 million per coin within his lifetime because he believes it is the best store of value and debasement hedge, with more capital controls and debasement ahead. “That’s my horse, and I’m going to keep riding it.”
  • Their security disagreement remains unresolved. Jonah thinks Bitcoin goes little lower unless it suffers a 51% attack and rejects fears that “Fable 6 or whatever is coming next” can simply rug holders. He trusts its architecture until quantum, which he thinks is at least 10 and probably 20 or 30 years away. Avi’s answer: “I think it’s way closer than that.”

7. Cybersecurity offers a rotation and challenges irreversible settlement

  • Avi wants to study CrowdStrike and the security landscape rather than keep debating the same AI leaders. Jonah expects cybersecurity stocks to outperform, including CrowdStrike and Palo Alto Networks. His mechanism is that second-, third- and fourth-generation trailing Chinese open-source models will soon approach Anthropic’s Mythos at hacking, without U.S. labs’ restrictions, while enterprises and governments struggle to patch a vast, complicated attack surface.
  • Jonah illustrates the gap with a hypothetical: JPMorgan might be patched while Fifth Third Bank is not. His proposed “two-legged trade” shorts vulnerable companies in a small size against two to four times as much long exposure to cybersecurity shares. Avi also mentions FAA and NSA outages, though he says their causes are unclear and an incorrectly uploaded contractor file may explain the NSA incident.
  • The concrete evidence Avi cites is Google Threat Intelligence Group’s report, “Agentic threat activity crosses a significant threshold.” He says attackers used an AI framework “with no human in the loop” to turn security patches into exploits; two months earlier, the group had not seen fully autonomous pipelines in the wild. He also cites U.S. aerospace intrusions and a suspected North Korean group planting Claude Code settings files to download a backdoor.
  • Avi draws the crypto implication: bank wires can sometimes be reversed, but mistaken or stolen crypto transfers cannot. A settlement delay may therefore be useful. The closing lament from Avi is “I’ve never been more disenchanted with crypto”: Sam Bankman-Fried torpedoed its first big opportunity, and he fears Donald Trump may have missed the second.
Full transcript