October 2025 Random Market Ramblings
Summary
Andrew Walker says Berkshire returned roughly 11% versus 10% for the S&P 500 over the past 20 or 30 years while investors bore the tail risk of Buffett aging. Buffett was still writing multibillion-dollar checks through the financial crisis at 75 and bought Apple at 85, which Walker calls his most profitable investment on a dollar basis. The fact that the tail risk never materialized makes “late-stage Buffett” more impressive, not less.
A strong realized return does not prove the original underwriting was sound. Earning 20% annually for three years can resemble selling hurricane insurance before the hurricane arrives; similarly, big tech’s extraordinary compounding also carried antitrust scenarios that never occurred. A great outcome in hindsight does not settle whether the risk was priced correctly.
Lavish investor relations may be proprietary evidence that management treats shareholder capital as “funny money.” Walker questions a $750 million company CEO spending two hours courting someone who might buy $7,500 of stock, or an analyst day distributing roughly $30,000–$35,000 of solar-powered rechargeable chargers before dinner and event costs bring the total to roughly $200,000 or more. He admits the signal is unresolved, but his bias is negative.
The behavioral case for averaging up is that investors who know a stock best may resist buying precisely when favorable news removes the downside branch. A stock rising from $10 to $13 may be more attractive if the probability of a zero falls from 33% to 2%: a tweet Walker may be paraphrasing said, “Everybody wants to average down, but no one wants to average up, and that’s why there’s alpha in averaging up.” Concentrated portfolios and fund-level risk limits make acting on that logic harder.
A rising stock can become mathematically cheaper, but whether it is truly cheaper depends on what drove the earnings revision. If price moves from $10 to $11 while expected EPS doubles from $1 to $2, the P/E falls from 10x to 5.5x. Walker distinguishes one-time cost cutting from 40% growth, excess demand, and a successful new product; those are very different changes to the outlook.
The counterweight to averaging up is the ease with which price action corrupts underwriting. A move from $10 to $20 can validate the thesis—or merely be a short squeeze before a fall to $10 or $8. Walker’s warning is against “letting loose the dogs of Excel” and underwriting 18% perpetual growth because a winning position feels good.
The podcast will slow after Walker’s second baby arrives, due in mid-November. Random Ramblings will likely pause for roughly two months, with the broader feed and newsletter potentially quiet for a couple of weeks, though guests are already lined up. He expects to return with “the rambliest rambling of all time.”
Deep dive
1. Buffett’s aging was a tail risk, yet his late-career record was extraordinary
Walker’s “risk writing” framework separates outcomes from underwriting: a position compounding at 20% for three years may still have carried catastrophic tail risk, like hurricane insurance that looked profitable only because no hurricane arrived. A great outcome in hindsight does not establish that the risk was calibrated correctly.
Big tech supplies the market example. Google, Facebook/Meta, and Amazon became exceptional compounders through strong businesses and successful pivots, yet investors also bore scenarios in which European-style regulation reached the U.S. earlier, Facebook was blocked from buying WhatsApp or forced to divest it, or Amazon faced tougher retail antitrust enforcement.
Walker says Berkshire returned something like 11% versus 10% for the S&P 500 over the past 20 or 30 years, probably with less risk given its underlying asset base. But Berkshire’s insurance business also writes disaster risk, including hurricanes, a Los Angeles earthquake, and even a nuclear-weapons attack.
The overlooked tail risk was Buffett himself: “How are you ever going to know that he’s lost it until he writes a really bad investment?” Walker compares Buffett with John Malone, whose roughly age-75-to-85 decade and lieutenants did not cover them in glory, and Carl Icahn, whose Icahn Enterprises fell roughly 50% over 10 years while the S&P rose roughly 450%. He cautions that IEP’s starting premium to NAV matters and also points to its weaker five- and three-year records.
The counterfactual is razor-thin. Buffett bought IBM at around 80, exited relatively quickly, then bought Apple at 85; had he remained slightly less sharp, held IBM another decade, and missed Apple, Berkshire’s concentrated portfolio might have underperformed. Instead, he made crisis-era investments—including Goldman Sachs and Bank of America preferred investments—and bought BNSF after 75. Walker calls Apple his most profitable investment on a dollar basis: “The man is just a one of one.”
Buffett is stepping down as CEO this year to become chairman emeritus, with a new CEO and chairman. Walker sees additional tail risk in the transition from a founder-led conglomerate, including the shift from Buffett’s roughly 20% ownership to younger people with significant but non-founder stakes.
2. Lavish investor relations may expose a culture of “funny money”
Walker questions both dollars and executive attention. If a $750 million company’s CEO spends two hours across two weeks courting a shareholder who may invest $7,500, the gesture is generous—but executive time has value, and the allocation appears difficult to defend.
His sharpest specimen is an energy-company analyst day that gave roughly 200 attendees solar-powered rechargeable batteries retailing for $135. Walker estimates $30,000–$35,000 of gifts and roughly $200,000 or more for the analyst day once the dinner and networking cocktail event are included: traditional IR, certainly, but possibly also “schmooze and booze” funded by shareholders.
The same discomfort applies when companies distribute $50, $100, or $150 consumer products “like candy” to analysts, shareholders, and prospective investors. Walker cannot prove these expenses predict poor stewardship and concedes they may be immaterial or produce a return; his bias remains that they reveal management treating shareholder capital as “funny money.”
That ambiguity may itself create an edge. Free treadmills would technically affect the numbers but, Walker says, would not be visible as a distinct COGS item to an outside observer or quant fund. A headquarters visit can uncover behavior that requires doing the work to find out, even though Walker has not determined how heavily to weight it.
3. Averaging up can improve the odds even as it raises the price
A tweet Walker may be paraphrasing lodged in his mind: “Everybody wants to average down, but no one wants to average up, and that’s why there’s alpha in averaging up.” His value-investor reflex is the opposite—buy more from $10 to $9 to $8, then sell rather than buy after excellent earnings lift the stock 20%.
John Hemp’s cautionary sequence captures the danger: begin with a 2% position at $50, double down at $25, then again at $12.50, $6, and $3; bankruptcy turns the original idea into a 10% fund loss. Loving a business more at $30 than at $50 is sensible only if the facts have not deteriorated—and Walker says he must adjust his mind if they have.
Price alone also misses the changed probability tree. At $10, a stock might have branches toward $30, $20, and zero; after good news sends it to $13, the zero branch might fall from a 33% probability to 2%. It is more expensive in dollars yet potentially much better on risk-adjusted odds.
Portfolio construction complicates the lesson. Value managers often run concentrated books because they can do only so much work on so many ideas; a position rising from $10 to $13 is already worth roughly 30% more before another share is purchased. Adding may breach fund limits, risk limits, or simply the manager’s “sleep-at-night limit.”
4. “Cheaper today” requires understanding what changed
Walker’s clean arithmetic: expected EPS doubles from $1 to $2 while the stock rises from $10 to $11, taking the P/E from 10x to 5.5x. The stock is therefore “cheaper today than it was yesterday,” although buying it now can feel like chasing.
The mechanism behind the earnings revision matters. Cost cutting pulls forward the time value of the savings and may continue unless management loses discipline and lets expenses rebuild, but it is a one-time lever that investors may already have modeled.
A jump from 10% expected growth to 40%, inability to satisfy demand, and an unexpectedly successful new product are very different circumstances. Walker notes that businesses may still have the cost-cutting lever available, whereas the demand and product developments change the growth outlook in another way.
Walker’s unresolved problem is separating improved fundamentals from self-confirmation. He has watched stocks run from $10 to $20, felt the thesis working, and then seen them fall to $10 or $8 after a short squeeze. The danger is to “let loose the dogs of Excel” and convert excitement into 18% growth in perpetuity.
5. A second baby will temporarily interrupt the ramblings
Walker’s second baby is due in mid-November, though “it could come tomorrow.” He expects to keep working but pause Random Ramblings for roughly two months; the wider podcast schedule will slow, and the feed and newsletter may go quiet for a couple of weeks amid late-night wakeups.
Guests are already lined up for the interim. After accumulating roughly three months of material—and losing sleep—Walker promises to return with “the rambliest rambling of all time.”