Summary

Andrew Walker sees a “casino market” in which crypto-treasury, AI, and newly public growth stocks are behaving like the dot-com bubble or late 2020 and early 2021. Bitcoin-treasury companies can jump 100%-300% and trade at 1.5-3x NAV despite investors being able to own Bitcoin directly; Circle IPO’d at $31 and Walker said it was at $223 by the recording, while CoreWeave became roughly a four- or five-bagger within months. His tentative, explicitly non-prognostic view is that this may be the “middle innings,” with room for irrationality to intensify.
A broad selloff during the Elon Musk–Donald Trump rupture suggested that speculative assets may share more hidden leverage than their distinct stories imply. Tesla fell roughly 20%, Bitcoin dropped about 3%, Walker thinks Coinbase fell 7%, and Palantir fell 10%. His concern is that leverage can drive prices much higher before causing them to “fall apart a lot faster once it stops.”
SPACs near trust value are Walker’s preferred asymmetric way to participate if the mania continues. A SPAC bubble could make their embedded call options valuable, while a collapse should return the trust value and leave holders with something resembling a cash equivalent: “heads I win, tails I don’t lose.” He immediately hedges the timing—after publishing the idea, a lot of these names weakened on July 18, making him wonder whether he had “personally marked the top.”
Pattern recognition is an investor’s accumulated advantage, but the same instinct can quietly become stubbornness. Buffett’s experience during the savings-and-loan crisis and at Salomon Brothers may have helped him understand the GFC’s cascading bank-run dynamics; conversely, Walker’s memories of recurring independent-power-producer bankruptcies caused him to pass on Talen Energy (TLN) before the stock became roughly a 10-bagger. His unresolved question: “Are you staying within your circle? Are you refusing to expand your circle? Are you being lazy?”
Rules that once protected capital can become destructive when the underlying market changes. “I only buy things at three times price earnings” has been a poor standalone strategy once companies are picked over by computers, while Walker’s own valuation rigidity contributed to missing Facebook despite recognizing Instagram’s potential after its $1 billion, zero-revenue acquisition. Buffett’s Apple investment in 2017 or 2018 illustrates the alternative: retain a coherent discipline while allowing the circle to evolve.
Management’s claim to be “the father of this industry” is a potential governance warning, not simply an expertise credential. Walker worries that the subtext can be contempt for outside scrutiny, resistance to capital-allocation feedback, or a belief that management deserves all the value the company or industry creates. He stops short of a rule—sometimes the authority is real—but treats possible arrogance as evidence requiring examination.

Deep dive

1. The casino market is manufacturing premiums from ticker symbols

Walker’s starting observation: every day seems to bring another crypto-treasury announcement followed by a 100%, 200%, or 300% stock surge. Echoing Matt Levine’s framing, one Bitcoin is worth one Bitcoin directly but somehow becomes “somewhere between 1.5 and three times Bitcoin” inside a public-company wrapper.
The contradiction matters because direct access is no longer scarce. Closed-end funds holding ordinary securities often trade below NAV because of illiquidity, fees, and management risk; publicly traded uranium ETFs do not tend to command giant premiums despite the difficulty of owning physical uranium. A readily purchasable asset “should not trade at a premium to NAV just because it’s put into a stock ticker.”
That premium creates an apparent “infinite money hack”: announce a Bitcoin treasury, reach 2x NAV, issue richly valued stock, and buy more Bitcoin. Walker calls the structure “very bubbly” and thinks historical parallels exist, though if forced to guess he would call it the “middle innings”; he is not a market prognosticator.
Circle is his sharpest specimen. It IPO’d at $31 in early June, reached $83 by month-end, and Walker said it was at $223 by the recording, while noting uncertainty about where it was trading. On June 23 alone, it opened around $240, peaked at $300, and finished near $260 before ending the week around $180. CoreWeave likewise cut its IPO price and share count, needed NVIDIA to anchor the offering, and became roughly a four- or five-bagger within three or four months.

2. Correlated selling hints at leverage beneath the narratives

During the Musk–Trump fallout, a roughly 20% Tesla decline made company-specific sense because government contracts could be threatened. What caught Walker’s attention was the sympathy move across speculative assets: Bitcoin down about 3%, what he thinks was a 7% Coinbase decline, and Palantir down 10%.
He concedes that a shared “meme factor” could explain the correlation, but keeps a darker possibility in mind: “maybe these things are more interconnected than you think.” Hidden leverage can enlarge a boom well beyond reasonable levels and then accelerate the unwind.
His analogy is Archegos, the firm led by Bill Hwang, which was concentrated in ViacomCBS/Paramount and a handful of other companies. Continued buying lifted the interconnected positions until the structure broke and the stocks collapsed. The analogy is not a prediction; it is the risk pattern that one unusually synchronized day brought into view.

3. SPACs offer a protected seat at the mania

If markets become fully manic, Walker thinks there could be a SPAC bubble. Early evidence includes buzzy deals rising severalfold, notably CEP, Cantor’s first SPAC, run by Howard Lutnick, now Secretary of Commerce. It announced a Bitcoin-treasury transaction and tripled.
The attraction is structural rather than a claim that any operating company is cheap. Buying near trust value provides an embedded call option on a speculative SPAC surge; if the market instead breaks, Walker says that buying at trust should get the trust back, leaving the investor with “basically a cash equivalent.”
That payoff could help value investors resist capitulating at the worst moment as stocks like CoreWeave run from $40 to $120 to $480. Still, Walker preserves the timing uncertainty: the day after his July 18 post, the air came out of many of these names, so “maybe I personally marked the top.”

4. Experience becomes dangerous when analogy replaces analysis

Pattern recognition explains part of investing’s age advantage. Walker argues that Buffett entered the GFC with experience younger managers lacked: the savings-and-loan crisis plus an inside view of how a run could cascade while chairing Salomon Brothers.
Walker’s own exclusions reveal the ambiguity. Something that trades in China is an immediate pass because he remembers frauds in China in the early-to-mid-2000s and 2010s and sees serious political, structural, and other tail risks involving assets there; Alibaba is his example. Yet he cannot decide whether that is sound boundary-setting, stubbornness, or “laziness on my end.”
Talen Energy, ticker TLN, makes the cost concrete. After it emerged from bankruptcy, friends highlighted strong backers, good assets, and the prospect of rising power demand. Walker instead remembered deregulated independent-power producers as complex commodity businesses where “one small part…breaks the whole thing” and every such producer ends up in bankruptcy roughly every five to seven years. He passed, and the stock became roughly a 10-bagger.
The opposite failure is forcing every new situation into an old victory. Some long-tenured investors with poor recent returns can invoke a winner bought in 1999 and sold for a five-bagger; similarly, “I only buy things at three times price earnings” mistakes a screen for a thesis when computers have picked over the universe and most extreme cheapness has a reason.

5. Adaptation and arrogance are both capital-allocation questions

Buffett’s dot-com-era refusal to abandon an understood strategy that had made money for one he did not understand was disciplined—but stopping there would miss his later evolution. His Apple purchase in 2017 or 2018 followed years of saying he did not do technology, showing that conviction need not mean permanent category blindness.
Walker applies the lesson to himself: he saw Instagram becoming a category killer when Facebook paid $1 billion for a business with zero revenue, yet never meaningfully bought Facebook. Being “too married” to buying at roughly 10x earnings kept the valuation rule intact while the opportunity escaped.
Newspapers show how inherited wisdom can become fatal: owning the local paper was a multigenerational “license to print money” until the internet changed the economics. Families that recognized the break could sell and probably preserve their fortunes; those that waited to “weather the storm” probably lost most of them. The same tension appears in the oilman’s maxim—“Whatever you do, don’t sell the oil rights”—and in Tom Thibodeau succeeding with a system until staying stuck in it helped get him fired.
Management arrogance raises the parallel governance problem. Sometimes a CEO’s claim to be “the father of this industry” is accurate, as with Edison or Bell; sometimes the manager was not even present for the founding. Walker worries that the subtext can be that outsiders cannot understand the business, capital-allocation criticism can be dismissed, and management deserves all the value the company or industry creates. His conclusion remains a question, not a screen—authority can be an asset, but inflexibility and entitlement may travel with it.
Walker also says he continues these ramblings not merely because he enjoys them, but because listener feedback and the discussions they generate help him improve as an investor.

AI is Eating Search The U.S. Can’t Build AI Without These Materials