Summary

Walker sees a market split between AI-linked “golden gods” and businesses reporting recession-like results. He guesses the S&P 500 was up double digits and the Russell low single digits by mid-August, yet Deere fell 10% on earnings, Crocs dropped 20%, telecom names imploded, and Chipotle posted roughly negative 4% comps. With the Mag 7 around 35% of the S&P 500 and largely tied to AI, active managers in old-economy stocks face an unusually punishing benchmark problem: “I don’t know how to explain this economy.”
Companies buying stock immediately before an earnings miss offer a signal whose direction remains deeply ambiguous. Crocs repurchased about $100 million near $100 per share before falling toward $80-$85, which might reflect genuine conviction; Vivid Seats instead “repurchased themselves into financial distress” while its business deteriorated. Walker’s screen historically produced three to five such situations but now shows roughly 15, raising doubts about whether tariffs and other “one-time” items explain everything.
Celebrity-led “hot girl marketing” can manufacture attention without proving incremental demand. Sydney Sweeney was named head of creative for Crocs-owned Hey Dude without, in Walker’s view, moving the needle; her American Eagle campaign really took off, while reports he saw suggested foot traffic remained down year over year. His unanswered investor question is how management distinguishes real ROI from access, buzz, and “stroking the egos of the management teams.”
Paramount’s UFC rights deal looks to Walker like the same tension between financial logic and owner prestige. He cites what he thinks was a roughly $7.7 billion agreement that takes UFC’s annual deal from about $550 million with ESPN to $1.1 billion with CBS—an enormous price for a property he considers below the NFL or NBA. Sports rights have value, he stresses, but the deal might also be an “ego indulgence” that buys ringside access and celebrity proximity.
When differentiated information changes a thesis, Walker believes position size should be re-underwritten rather than left on autopilot. He recounts receiving news he thought was almost certainly positive for an already-large holding, declining to add, and later being challenged: “You had real edge… Why didn’t you increase it?” Investing involves “unknown probabilities,” but real edge is rare enough that investors should explicitly decide whether to add, reduce, short a related basket, or buy puts.
John Paulson’s purchase of Carl Icahn’s roughly 10% BHC stake is a rare, unusually visible challenge trade involving two sophisticated, highly informed insiders. Both held board seats, yet Paulson paid more than $300 million at $9 per share while BHC traded around $7-$8—a premium where a normal block sale might clear near $4.50-$5. Icahn could have been selling for reasons unrelated to BHC’s value, but Walker notes he would not sell at $9 if an obvious BLCO windfall were imminent, while Paulson would not willingly pay that price if everything looked disastrous. Walker has no BHC position but calls the transaction a “very interesting signal.”

Deep dive

1. AI’s golden gods are masking a recessionary-feeling market

Walker’s starting point is confusion, not a macro forecast: around mid-August, he guessed the S&P 500 was up double digits for the year and the Russell up low single digits, despite April’s volatility. Unlike the COVID-era panic market, this is “such a tale of two cities,” and he repeatedly concedes, “I don’t know how to explain this economy.”
Anything touching AI has become “a golden god”: earnings beats lift expectations, misses receive the benefit of the doubt, and backlogs look extraordinary. With the Mag 7 roughly 35% of the S&P 500 and largely tied to AI, an active manager holding old-economy businesses is structurally underweight the index’s dominant engine and finds it “almost impossible to keep up.”
Outside AI, the tape looks brutal. Deere fell about 10% after earnings; Crocs dropped 20% toward $80-$85; Lululemon’s chart makes Walker say, “You’re going to barf on yourself”; and supposedly stable telecom and cable companies have been “absolutely imploding.” Retail weakness no longer looks confined to traditionally cyclical metals, mining, autos, or trucking.
Restaurants deepen the puzzle: Chipotle reported roughly negative 4% comps, while Cava and Sweetgreen were also being hammered. Walker characterizes many underlying reports as “deep recessionary numbers,” even as broad economic statistics appear roughly flat and AI-linked companies remain “up to the right” and unable to stop. His conclusion stays deliberately unresolved: “Weird market, weird everything.”

2. Pre-miss buybacks may reveal conviction—or failed forecasting

Walker’s custom screen looks for companies falling roughly 20% on earnings after repurchasing shares, with management attributing the miss to temporary factors. The central question is whether executives saw exceptional value despite a short-lived setback—or bought aggressively without understanding the deterioration directly ahead of them: “Does this company know how to forecast?”
Crocs is his constructive specimen. A company worth roughly $4.5-$5 billion repurchased about $100 million during the quarter near $100 per share, before the stock fell toward $85. Based on its history and earnings-call language, Walker says he “would be shocked” if Crocs stopped buying, though he owns no shares and has no plan to initiate a position.
Vivid Seats illustrates the opposite outcome. Its private-equity controllers appeared sophisticated, yet the company bought stock “hand over fist” as the business weakened, then reported second-quarter results far below expectations. It has now “repurchased themselves into financial distress,” leaving so little capacity that analysts were not even asking whether buybacks would continue.
Historically, Walker found only three to five companies matching the screen; now it returns about 15, many retailers blaming tariffs and other one-time effects. He says the tariff explanation makes intuitive sense but wonders whether “there’s other stuff there,” even speculating—while acknowledging he cannot quantify it—that deportation could shift demographics and marginally reduce restaurant demand.

3. Celebrity attention can conceal weak marketing economics

Sydney Sweeney’s campaign prompted Walker’s broader question about “hot girl marketing.” She was named head of creative for Crocs-owned Hey Dude, and to his knowledge “has not budged the needle one iota.” American Eagle’s campaign generated vastly more publicity than it likely expected, but reports he saw still showed year-over-year foot traffic down after launch. Virality might eventually drive sales; he simply cannot see how management measures the increment.
The concern extends from influencers promoting aesthetics, fitness products, energy drinks, and protein powders to wealthy owners buying movie studios. Walker suspects some smaller influencer programs have ROI that is “absolutely zero or negative,” while studio acquisitions may rationalize an owner’s ability to approve projects and meet stars. The financial case is always presented, but the underlying purchase may be an “ego-stroking thing.”
He applies the same suspicion to David Ellison’s Larry Ellison-backed control of Paramount and its announced UFC deal. Walker says the agreement was roughly $7.7 billion, taking the UFC from about $550 million a year with ESPN to $1.1 billion with CBS. He stresses that sports rights have value, but wonders whether that scale also reflects an owner’s desire to flex, sit ringside, and hobnob with celebrities, given that he considers UFC a less valuable property than the NFL or NBA.

4. Real edge demands a fresh decision about position size

Walker has considered asking podcast guests whether a pitched stock is a monthly idea, an annual idea, a five-year idea, or a Buffett-style “punch card idea.” He avoids it because portfolio structures differ and no guest wants to pitch something as merely average, but the question usefully forces investors to quantify how exceptional their claimed edge really is.
His own uncomfortable example came when new information made him believe good news was almost certain for one of his largest positions. He did not add because he was “already huge,” only to hear the challenge later: “You had a differentiated view on that news… Why didn’t you increase it?” A Kelly-criterion model, he acknowledges, would have pointed toward a larger position after the probability changed.
Blackjack supplies the contrast: with 11 against a dealer’s six, the update is mathematically clear. Investing offers probabilities without known values—“What does this 8-K mean? What is that press release for you?”—so re-underwriting involves judgment and can benefit from another person willing to ask whether the information constitutes genuine edge.
The same discipline applies when the update is negative. An investor who spots a cyclical downturn before analysts but refuses to alter a large long-term holding might still be right on intrinsic value; Walker nevertheless argues the new information should change the trade somehow—through a smaller position, a short basket with similar beta, or puts. “Real edge comes along rarely.”

5. Paulson’s premium-priced BHC purchase is a visible challenge trade

Walker defines a challenge trade through fantasy football: two managers exchange running backs because each thinks the other owns the better player. Unlike a complementary running-back-for-receiver trade, one side must be relatively wrong. Public-market transactions are normally the same contest, but anonymization hides the seller, motives, and competing thesis.
BHC, the former Valeant, makes that contest unusually observable. Its remaining acquisition-era debt turns the equity into a financial-engineering thesis: shift valuable assets from the debt into equity holders’ hands, and even a modest transfer can create enormous equity value. The pivotal asset is BLCO, Bausch + Lomb; about 19% already trades publicly, most of the remaining stake sits in an unrestricted subsidiary, and distributing it to BHC shareholders could unlock substantial value.
John Paulson and Carl Icahn each owned roughly 10% and held BHC board seats, giving both access to internal negotiations, constraints, and opportunities around that financial engineering. Yet Paulson bought Icahn’s entire stake for more than $300 million at $9 per share while the market price had recently been near $6-$8, and Paulson was also buying in the open market.
A normal 10% block from a sophisticated exiting shareholder might require a steep discount—Walker imagines bidding $4 and settling near $5, or an investment bank clearing it around $4.50-$5. The $9 premium therefore fascinates him, but he stresses that Icahn’s motive is unknown: Icahn could be selling because he is at the end of a fund life or for another reason. Conversely, if both knew BLCO would imminently be sold for $100 per share with the proceeds going to BHC shareholders, making BHC worth $50, Icahn would not sell at $9, while Paulson would not buy if the structure looked disastrous. Walker has no BHC position and views the trade as unusually informative because both sides had substantial access.

Himalaya: Factory to Global Brand and Succession, with 张琪 and 张栀君 The CEO Who Cheated Death, Slept an Hour a Night, and Built a Healthcare Empire