Berkshire Hathaway Part III: The Modern Empire and Succession (1992–2021)

In one sentence: The trilogy's finale — the most admired investor on earth runs headlong into an Internet age he can't parse: Coke stalls, Gen Re blows up, an entire technology generation is missed — and yet, in 2008, he executes a textbook crisis deployment with "money nobody else had," then, at 86, redeems nearly every late-career error with an Apple position he didn't even come up with ($89B of paper gains), before Charlie accidentally leaks the successor live at the 2021 annual meeting with a single line, "Greg will preserve the culture." Acquired closes two threads with this episode: "Compounding large numbers, David, sure is hard," and the series' ultimate takeaway — never bet against the Internet.

The Company on One Page

Founder Profile: The Late-Career Buffett and Munger

The framing lens: Andrew Marks (TQ Ventures) hands Acquired the key to the whole episode — "Warren was the world's greatest status quo investor." So long as the future mostly resembles the present, he is a genius at pricing certainty: he could price New Coke's recovery, and see Coke's international runway. His valuation math is "modest value × very high probability," and it structurally rejects the "very high value × low probability" distribution of tech — a genius move on the essentially static chessboard of the 1950s–early 90s, and a losing one once the rate of change begins to compound (the PC, the Internet, several news cycles a day). Coke's 10x first decade and 3.5x next 25 years is the cleanest sample of that curve. David rewrites an 1848 line about German history into the episode's thesis: "Investing in corporate history reached its turning point and Warren failed to turn."

The ham-sandwich philosophy (Alice Schroeder, The Snowball): "Buffett avoided technology stocks partly because these fast-moving businesses could never be run by a ham sandwich. He thought it no shame to have a business that could be run by a ham sandwich; he wanted to get Berkshire Hathaway to the point where it could be run by a ham sandwich, too." David's reflection: those businesses no longer exist — the most valuable companies of the future will be built by the companies, leaders, and entrepreneurs able to navigate change. "This is not Coke."

Psychological portrait: scoreboard + being loved: Warren simply wants the net-worth number as high as possible by the day he dies — money is a scoreboard, not for spending; and at the same time he craves adoration. Nothing in securities law requires a shareholder meeting to be a festival — that pilgrimage is a Warren Buffett creation. Ben: "He wants to be a beloved figure, and teacher, on top of being the wealthiest person on earth." David's verdict: "You could accuse him of not being innovative in his investment philosophies. Never accuse him of not being innovative in finding ways to get what he wants."

A tech stack frozen at the fax machine: a flip phone, a dependence on faxes. Summer 2009 at Sun Valley, his daughter Susie looks at his phone: "Dad, there's a text message on your phone." Buffett: "What's a text message?" — the text was from Lehman. His standing verdict on crypto is "rat poison squared"; David doubts either man ever seriously studied what Bitcoin or Ethereum actually are. His stock answer on retirement: "about five years after I'm dead."

Charlie Munger: author of the raisins-and-turds line; the bluntest partner in history on the Gen Re deal ("very late in the game on this one"); and at the 2021 meeting, the accidental leaker of Greg. David riffs off it: "He's got that line about Charlie and I have never had an argument. Bullshit — I bet you had one after that." Charlie's ~1950 logic for choosing Los Angeles — a place "large enough to have an impact, but still small and growing enough that he could become somebody there" — is the exact frame David borrows in the finale to define today's Internet.

Apple at the 2018 annual meeting (the most important late-career self-correction): "I didn't go into Apple because it was a tech stock... I think it's much more the nature of consumer behavior." Ben and David pile on: "Yeah, he's in it for the M1." "He's locked in to iMessage." "He feels that this integrated strategy is the right one."

The Playbook

Each entry: origin story → insight → effect.

1. Different scales demand a completely different playbook (the through-line)

• Story: the trilogy's spine — the cigar-butt playbook stops working first; then the "buy a great hundred-million-dollar family business, keep the managers, and leave them alone" playbook works until capital swells past the point where there simply aren't enough furniture stores and family jewelry chains in America to buy. Warren has been warning "we're too big to move the needle" for 25 straight years.
• Insight: every leap in scale requires a completely, completely different playbook; once capital outgrows its target universe, returns get dragged toward the S&P 500 by a "law of gravity."
• Effect: from the mid-2000s the strategy is forced toward elephant-gun acquisitions and crisis lending; Berkshire has made every switch so far — the episode's suspense is who makes the next one. That is the essence of the succession problem.

2. The valuation switch on acquisition currency, and the anti-speculation design of the B shares

• Story: Buffett is famous for buying only with cash — the businesses Berkshire already owns are better than almost anything he could buy, so trading treasure for "who knows what" is a bad deal. Ben pulls up the numbers live: Berkshire has traded meaningfully above 2x book (normally 1–1.5x) only twice — ~1996 and ~1998 — and he acted both times: the B shares in 1996, the all-stock Gen Re deal in 1998. The B shares were triggered by the late-90s "shadow Berkshire" phenomenon: people packaging Berkshire stock into public trusts and selling fractions, and others copying the 13F/10-Q/10-K homework line by line (lagged, and without the float) — which Warren sees as retail getting swindled, even as he refuses year after year to split the stock ("I love our current investor base — why would I change this great set of shareholders?").
• Insight: when the market prices your stock "richly," the stock becomes phenomenal currency; the B-share design is mechanism engineering — 1/30 the A-share value, gutted voting rights, and an open-ended offering (fill all demand at a fixed price) that kills scarcity and hoarding arbitrage from the supply side, and picks up cash for free: "This is even better than float. You never have to give the cash back." (Ben: "They're raising their series-C.")
• Effect: shadow trusts eliminated, no split, and free cash — three birds. Buffett's own wording is the valuation signal: "He didn't say it's overpriced, he said it's not underpriced." Ben, via Ben Thompson: this is a strategy credit — do the thing you wanted to do anyway, and get applauded for it.

3. Gen Re: the price of overthinking, and the limits of the White Knight playbook

• Story: in 1998 Berkshire buys one of the world's largest reinsurers for $22B in all stock. The deeper motive (The Snowball): Warren believes the tech bubble will burst and wants to reduce Berkshire's equity exposure, but "Warren Buffett is selling stocks" would itself crater the market — so the all-stock deal brings in ~$20B of Gen Re bonds, cutting exposure without selling a single share. Ben: "That is some financial jiu-jitsu engineering." The result: within the first week, the Unicover insurance-fraud scheme costs $300M (against Buffett's rule number one, "never lose money," and rule number two, "see rule one" — down $300M in week one); Hollywood box-office underwriting loses ~$1B; 9/11 costs nearly $2B through coverage holes; and Gen Re is dragged into propping up AIG's balance sheet. Meanwhile Warren's literal promise to the old management was "strictly hands off." David, one word: "Yikes."
• Insight: buying a bad company to hedge a signaling risk is the cost of overthinking (David: "He definitely overthought this one because Gen Re sucked"); the "keep management, hands off" White Knight playbook only holds for good businesses run by founders or great operators — copy it onto a non-founder-run company that's mispricing risk and you've waived inspection on an out-of-control system. The playbook is a conditional strategy, not dogma. And: Charlie's value is in the ex-ante veto, not the ex-post commentary.
• Effect: Warren ousts the old management, sends in Joe Brandon and Tad Montross to clean up, and only hands the whole thing to Ajit in 2016; "call Charlie early, not late (see Salomon Brothers)" becomes a series-level lesson.

4. The reverse of adverse selection: the risks nobody dares insure are the fattest margins

• Story: Ajit Jain, who joined in 1985, builds a reinsurance business from scratch inside Berkshire — "the great entrepreneurial story within Berkshire." He opens by running a full-page ad in Business Insurance: "We are looking for more casualty risks where the premium exceeds $1 million." — "This is crazy. Nobody does this." After 9/11, when the whole world panic-buys terrorism insurance, he judges the actual risk overrated and writes policies with premiums in the tens of millions: "I'm going to make a killing."
• Insight: if your pricing ability is the best in the industry, the weird risks everyone else declines are the fattest margins (super cats — "a really cool name for a pretty boring thing"); when panic inflates the premium rather than the risk, selling insurance is buying cheap.
• Effect: factory reinsurers like Gen Re and Swiss Re have capital but no nerve; small players have nerve but no capital — only Ajit + the Berkshire balance sheet can stand at the steepest point of the supply curve. David: "If Ajit had decided to be a venture capitalist, he would have been like Bill Gurley times 10."

5. Innovation ≠ capturable value

• Story: in a 1999 Fortune piece warning about the bubble (without calling a top — he just says he isn't interested), Buffett notes that the early auto industry had hundreds of manufacturers and an explosion of innovation, and ended with only a few; he closes on the airlines — as of 1999, the industry's cumulative profits minus losses since inception were negative.
• Insight: the investor's question isn't "is there innovation" but "is there a moat that keeps the profits from being arbitraged away"; his late-90s proof point: no Internet company has ever made $100M in a year in profit, and he has no proof it can exist.
• Effect: he dodges the bust — but the same man who wrote that he'd like to go back to Kitty Hawk in 1903 and "do capitalists a favor and shoot the Wright brothers down" later buys every airline stock in the industry, twice (see Deep Cuts).

6. Raisins and turds: a truth that holds only in the narrow window at the bubble's peak

• Story: at the 2000 annual meeting, Warren and Charlie get pummeled in the arena — "everyone else is getting these five-x's in a year, and you're trying to make me 15%?" Warren: "I don't want to speculate about high-tech," then compares the whole thing to a Ponzi scheme. Charlie's coda: "The reason we use the phrase wretched excess is because it produces wretched consequences, it's irrational. If you mix raisins with turds, they're still turds."
• Insight: the two positions differ subtly — Warren is "I don't understand it, I refuse to engage, it's a Ponzi"; Charlie concedes there are real companies in tech (Microsoft = raisins), but when the bubble pops the splatter from the turds drags the good ones down too.
• Effect: David's backtest (to May 2021): a dollar put into each at the exact bubble peak in 2000 leaves Berkshire ahead of the NASDAQ, well ahead of the S&P, and slightly ahead of Microsoft — Charlie was right at that precise instant; but at almost any other entry point (1997's Sun Valley panel, or just a year later in 2001), Microsoft crushes Berkshire. Right on the timing; long-run growth still wins on the stock.

7. Crisis lending and term design: you're not selling money, you're selling your name

• Story: in 2008 the government opens the spigots ("bazookas," proto-QE), capital is nominally free, but counterparties inside the financial system don't trust each other, and taking government money means wearing the bailout stigma. Berkshire is sitting on $37B in cash — "back then nobody else had that kind of cash anywhere, the only people who have it are governments" (the most valuable company on earth was only ~$200–300B) — and is the one private actor with cash, a trusted brand, and a single decision-maker (no LPs, no committee). Chastened by Salomon (don't be the big equity holder in a crisis; don't get dragged before Congress again), he does it all as high-coupon debt/preferred + warrants: Wrigley 11.45%, Goldman 10%, GE 10%, Swiss Re 12%, Harley-Davidson 15%, Tiffany's 10%, Sealed Air 12%.
• Insight: Goldman is buying five words — "Buffett believes in me" — the anti-run signal the government can't give (its money = shame) and other funds can't give (no brand); the preferred/debt locks in ~10% downside protection while the warrants keep equity-like upside — "rule number one, don't lose money," in its crisis form, is fixed income.
• Effect: $18B deployed, ~$25B back over ~five years, not a single loss ("good, not amazing" — a VC would call it a top-1% vintage); the real money is in the warrants: Goldman makes ~$3B (invest $5B, ~$8B back in two years; strike $115, GS at $364 at recording), while GE is far worse (strike $22.25, GE only $13 at recording); the 2011 Bank of America rerun ($5B preferred @5% + $7.14 warrants, BAC at $42 at recording, ~6x) is up ~$26B to date — more than every other crisis deal combined ($7B), possibly more than any single investment in Buffett's prior career. Contrast 2021: Amazon/Apple issue debt at ~0.3% — same companies, different year, a chasm in the price of capital.

8. Extreme preparation → decisions in minutes

• Story: after Bear collapses in March 2008, Lehman CEO Dick Fuld calls Buffett for a capital injection; Buffett studies Lehman's 2007 10-K, marks the printout up by hand (he later produces the physical evidence in a crisis-retrospective video), and passes on the risk.
• Insight: Ben's canonical Buffett: "I'm going to wait until prices are rational again. I'm going to do all my research. I'm going to be so prepared that when the moment presents itself, I can act in mere minutes." — the 1962–63 AmEx move, replayed in 2008.
• Effect: Goldman, GE, and Wrigley all close within days or a weekend; the passes on Lehman and AIG are equally products of preparation — the note "too much risk" was already written. Ben: since buying all of GEICO in 1995, the 2008 sweep and the later Apple buy are the two or three most important moves of 25 years.

9. Opportunity cost as first principle: sins of omission and commission are equally guilty

• Story: 2011–2016 is what David calls "probably some of the worst decisions Buffett ever made": IBM $10.7B (built in November 2011 — the App Store was already four years old and Bezos had been pitching AWS at Startup School for three; sold in 2018 for a ~$2B total loss); Kraft-Heinz $10B (roughly zero return over eight years; David: "you're partnering with a private equity firm — they're your enemy, you know what they do"); Precision Castparts $37B (the largest acquisition ever, a $10B write-down in 2020 — "That's a dog"); airlines (bought 2016, dumped at the pandemic bottom in 2020); plus a mediocre J&J in 2008 and the ~$1B GE rail-car fleet taken on via Marmon around 2015. Worse are the passes: a man who understands the AmEx interchange business watches the "criminally undervalued" IPOs of MasterCard (2006) and Visa (2008) go right by, along with Google, Facebook, and Amazon.
• Insight: the true cost of capital allocation is opportunity cost — "everything is only worth talking about when you compare it to its next best option"; that $10.7B into IBM, put into any large tech company, would have won big. The circle of competence is double-edged: it keeps you out of what you don't understand, and lets you walk past what you should have.
• Effect: over the trailing five years (from 2016), ex-Apple, Berkshire underperforms the S&P by about half a turn on a multiple basis; David: "Warren himself has genuinely underperformed the market over the last five years."

10. Outsourcing philanthropy = an extension of the circle-of-competence principle

• Story: the plan had always been that Susie would outlive Warren and the Susan Buffett Foundation would give away the then-$40B+ family fortune after his death — voided when Susie died first in July 2004. Warren had thought since his teens about "what do I do when I'm really rich" (worrying over overpopulation, energy, famine), but his obsession with measured performance made giving — where he couldn't see a return, pick a manager, or compound — deeply frustrating ("that's a Susie problem"). In 2006 he calls a press conference to give away 85% of his Berkshire stock (then $37B): 5/6 to the Gates Foundation, 1/6 to his children's foundations and the Susan Buffett Foundation — no Warren Buffett Foundation, no personal giving decisions.
• Insight: he admits that spending and earning are two different skills, so he handles philanthropy the way he handles investing — find the best manager, then let go. David calls it a win-win-win: the world calls him the most generous person in history; the Giving Pledge makes "give it all away" the ultimate billionaire status symbol, using social competition to drive giving commitments; and Warren never has to run any of it — "Warren is getting exactly what he wants."
• Effect: the Giving Pledge launches with the Gateses in 2010 (the 2006–2010 lag blamed on nobody wanting to make big gifts during the financial crisis; Ben's joke: it should be the "Buffett Giving Pledge"). The cost (Ben): the Gates Foundation will define the Gateses' legacy longer than Microsoft will — for an ego as large as Warren's, giving up a giant namesake foundation must have been a real trade-off.

11. The sequenced choreography of institutionalized succession

• Story: the 2006 annual report runs an open "help wanted" ad for a CIO (the transcript misstates it as 2016): "I intend to hire a younger man or woman with the potential to manage a very large portfolio, who we hope will succeed me as Berkshire's chief investment officer... as part of the selection process, we may in fact take on several candidates." He then splits his job into operating and investing tracks; on the investing side he starts candidates at ~$1B each on trial and expands with performance ($13B → nearly $20B each at recording); the CEO name lives in an "envelope," with a character veto — and when Sokol offends in 2011, the mechanism proves real and the envelope swaps seamlessly to Greg Abel; the January 2018 Abel/Jain Vice Chairman promotions make it public.
• Insight: succession isn't finding "the next Buffett," it's a sequence: open call → split the roles → small-position trial → expanding-rope curve → character veto. The comp design is just as intricate: pay on excess return over the S&P, settle on a rolling 3-year basis, with a claw-back — reshaping hedge-fund incentives into Berkshire-style long-cycle incentives. Top independent managers (Li Lu, rumored Einhorn) keeping 2-and-20 won't come work for a salary, so Berkshire can only attract people on the way up rather than at their peak (Todd running $100M, Ted running $2B) — a self-selection mechanism in itself.
• Effect: over 2011–2016 Buffett turns in the worst stretch of his career while the new system delivers Apple — organizational redundancy hedges the founder's decline. In 2019 he owns it publicly: "yeah, they both beat the S&P by a little bit but they smoked me."

12. Berkshire's real shared culture is only three rules

• Story: outsiders keep calling Berkshire's culture "sacred." Listener research (relayed by Ben) punctures it: a truly sacred culture would be codifiable and protectable; in reality each operating company's culture is independent (a Borsheims employee doesn't think about Lubrizol's culture) — they've taken decentralization to the extreme. The only head-office rules that actually run through hundreds of thousands of employees are three: (1) never put Berkshire's reputation at risk; (2) bend over backwards to avoid/defer taxes, because taxes pull money out of the business — keep it all inside compounding; (3) funnel all cash back to headquarters for reallocation. "That's the culture. That's what the head office really requires of its subsidiary managers."
• Insight: pushed to the extreme of decentralization, "culture" becomes the minimal executable rule set — reputation, taxes, cash flow, and everything else devolved. Those three rules are the daily form of Berkshire's structural genius: it's an operating company, not a fund — infinite horizon, never-sell goal, no fee drag ("the sticker performance is actually the performance you get"), and managers who suffer alongside you when there's nothing to buy (because they believe cash is the best current option).
• Effect: even as Warren can't parse the Internet and the world shifts under his feet, the one big thing he got right was designing the structure flexibly enough for successors to face the future in a non-Warren way — in 2020 Todd (per Ben) buys Amazon and Snowflake, "a non-Warren-approved strategy already running inside the company" (Ben's caveat: culture and internal politics are open questions).

13. The fiefdom succession problem (Ben's biggest worry)

• Story: one brain used to do all capital allocation — acquisitions, the internal engine (Ajit's insurance), buying stocks — switching among the three at will. Now the board is four fiefdoms: Greg (non-insurance, CEO), Ajit (insurance), and Ted and Todd (public markets, ~$20B each), each paid mainly on his own portfolio.
• Insight: who has final say? Nominally the CEO — but when a non-founder CEO without founder authority or a whole-board view has to "decide a giant deal in an hour," decision-making gets thorny, and the incentives have to be engineered precisely to work. David piles on: Greg is a great operating executive, but he doesn't have Warren's investing mind — can he think like Warren, Todd, and Ted and act with the same speed and conviction? Ben's paradox: the right move would be to try like crazy to find the next Warren and hand him everything — but that may simply not exist. David's irony: the top stated criterion for the next CEO is "experience running a large operating company" — by that standard, Warren himself doesn't qualify.
• Effect: unresolved. Ben guesses it'll be 20 years and a book before we learn how the "new four" make contentious decisions; both guess Warren and Charlie will go at roughly the same time.

14. "Position to be right" beats being right about the thesis (the closing insight)

• Story: in May 2016, "one of the fellows in the office who manages money" (Buffett's own phrasing; never officially named, but David suspects Ted — Todd does financials, Ted does everything else) buys ~$1B of Apple, then convinces Buffett to "back up the truck" — $36B in over two years, just under the $37B of the largest-ever acquisition, Precision Castparts; Buffett is 86 as it's built. Per the latest annual report: $120B in value, $89B in gains — "he literally created a whole Zoom's market cap of gains" — which David estimates approaches or exceeds the dollar returns of Buffett's entire rest-of-career combined.
• Insight: Ben's opening maxim: "Nobody cares what your thesis is. Nobody cares if you're right or wrong... you just want to be in a position to be right." David: it doesn't matter that it was Ted's idea, or that Warren hated tech stocks — all that matters is that he got himself into a position to be right; $89B of gains later, here we are. Better still, the buy thesis is exactly the moat standard he wrote down in 1995 (brand + place in the consumer's mind + switching costs), not chasing a tech wave — form and outcome, rarely, aligned (David's father's batting advice: nobody cares how you stand if you can hit .300, but learn to hit .300 with the right swing first).
• Effect: Apple is roughly 50% of every dollar Buffett ever returned (Ben's estimate); it's also compounding math's inevitability — a system compounding for 63 years is destined to produce the bulk of its absolute gains in the final years. Apple is a miracle and it is arithmetic.

Moat Analysis (the 7 Powers framework)

7 Powers is Hamilton Helmer's strategy framework, which Acquired runs every company through — but this finale has no standalone 7 Powers segment, replaced by a Bull/Bear + Playbook + Grading structure. The table below annotates the scattered judgments against the framework; it is not the episode's native structure.

Bull case (in full)

• Market consensus is "Buffett retires (he won't — he'll work until he dies) = Berkshire is toast, the stock plummets, performance vanishes" — after this series, you may reach the opposite conclusion;
• It could be better under Greg: less conservative, less cash held (cash is a drag on returns);
• Warren has gotten gun shy after the string of sins of omission and commission, trusting only the moves that worked in the past — which may not suit this environment;
• Todd and Ted are quite good: together they run only $40B, yet produced the $89B-gain Apple.

Bear case (in full)

• Style-reversal risk (the reverse bear case): Buffett-style investing may be great again someday — this may just be a season making him "look foolish" (possibly 10–15 years of it); if Berkshire drifts too far from the long-tested Buffett strategy, that itself is dangerous;
• The halo effect disappears: the tangible (the crisis coupon terms nobody else got, the calls that came to him) and the intangible (people holding and making the pilgrimage because they believe in Warren) both go with him — "Warren's gone, then what?";
• No internal growth engine: the ideal (Bezos's Amazon) has several capital-absorbing engines inside; Berkshire's only decent one is GEICO, and it's not a real engine — internal capital deployment can't clear any exciting hurdle rate, so it has to keep "going shopping";
• The generous sum-of-parts assumption: believing Berkshire trades below intrinsic value means agreeing Apple is worth its current price (its earnings multiple has gone from ~7x to ~30x during the hold) and that BNSF is worth current railroad multiples — "if you believe Berkshire is undervalued, I think you're being pretty generous with how you value the sum of the parts."

Grading (David rarely writes a grading framework in advance; here it's four tiers)

The A-vs-A+ debate: David gives an A — we'll probably see better investors than Warren in our lifetimes (the numbers grow over time, the world changes faster); Ben pushes back — it depends what "better" means: you can't replicate this curve starting today without more risk/leverage, and setting a million people loose for 70 years guarantees someone beats him, but mostly through luck plus risk. David's comparison is Sequoia Capital (founded 1972, "coming up on 50 years next year," with no public aggregate return, "which I suspect might be as good or better than Berkshire" — though a firm, not one person): Apple is 100% to Warren's credit, but he's "lost a step" in recent years, whereas Sequoia has only gotten better or held at the top. Ben's closing verdict: "A strong finish would be an A+; you could say Apple is a strong finish, but 13.5% IRR since 1993, over 28.5 years and two or three full cycles, can't all be explained by 'style headwind' — the last 30 years simply weren't remarkable enough."

The B anchor: BPL 29.5% → the heyday (late 70s to mid-90s, ending in Salomon) at 27% → 13.5% since 1993 — a marked decay, "even though they told us all along it would happen (the size of the capital), still..."; still definitively beating the S&P, just not what it once was.

The C logic: Apple is a strong contender for the best single investment in history, but "the gravity of capitalism" is too heavy and Warren's other investments too poor, so the last five years merely track the S&P. Ben: "Ted and Todd are making money, but Warren isn't getting paid any carry." The reason it's not lower: no 2-and-20 — holding Berkshire is equivalent to holding a zero-fee index fund.

The money question in detail: Ben spends 10 hours of show before admitting he's never held Berkshire (repeatedly, not investment advice); he considered buying it many times during research, and concluded: extremely conservatively managed, "it's a good widows and orphans stock" — no dividend (avoids dividend tax), continuous compounding, sell shares yourself for cash, won't string together five bad down years; whether it fits depends on where you are in your investing life cycle — not for the young, "at least the young in wealth." David is a longtime registered holder and keeps holding: his portfolio is almost all high-growth tech, digital assets, and San Francisco real estate, and Berkshire is his "safety" slot and emergency liquidity; he used to keep a big cash allocation for this, then realized "holding cash is dumb" — the real yield on 10-year Treasuries is basically negative, "every day you get a little poorer" — so he swapped his liquidity allocation into Berkshire: "at least I can be confident it won't go to zero and I'll get some return. Treating Berkshire as a cash substitute is a little sad for the stock, but that's where it is."

The series' ultimate takeaway: never bet against the Internet

David's close: Warren preached "believe in America" his whole life — America is undefeated in capital growth; whether that still holds for America today is debatable, but it absolutely holds for the Internet. Today's Internet is Charlie's 1950 Los Angeles: large enough to matter, still small and growing enough that you can become somebody there. The weekend before recording, Bitcoin and Ethereum both crashed 50% — "it's all noise. The Internet is still the future." "Never bet against the Internet. That's my takeaway." Ben's coda: someone should make a meme — Warren's slide reads "never bet against America," beside David Rosenthal's "never bet against the Internet." David then annoys both camps at once: in a change-dominated world, Sequoia-style expected-value "shots on goal" beat Buffett-style certainty — "we're going to rile up the growth people and the value people."

Deep Cuts (Berkshire itself)

• Fax vs. text (the full Lehman mix-up): on Lehman weekend in September 2008, Buffett is on vacation in Canada (with his partner, later wife, Astrid); after the rescue call he says, "Send me a fax to the hotel with the details of exactly what's going on and what you want me to do." He goes to a show, comes back — no fax. Lehman goes bankrupt on September 15 (AIG also blows up that weekend; the Fed saves AIG and not Lehman; Buffett passes on AIG too). A year later at Sun Valley, his daughter Susie finds a text sitting in his flip phone — from Lehman, asking for the hotel's fax number. Buffett: "What's a text message?" Ben's clickbait summary: "How Warren Buffett could have saved Lehman Brothers if he was a little more tech-savvy." His note on being turned down has no heat: "they just didn't want me that bad."
• The B-share prospectus disclaimer (Ben's favorite): "Mr. Buffett and Mr. Munger believe that Berkshire's class-A common stock is not undervalued at the market price stated above. Neither Mr. Buffett nor Mr. Munger would currently buy Berkshire shares at that price, nor would they recommend that their friends or families do so." Ben: "It's like Eric Yuan going on Bloomberg and saying, it's too high."
• Li Lu: the successor who never took the job: in July 2010 the WSJ front page reports that Buffett's succession search was nearly done, with Li Lu as the choice; Charlie is quoted, "it is a foregone conclusion that Li would join Berkshire." Li grew up in China, took part in the Tiananmen protests, emigrated to the US, founded Himalaya Capital (mostly China-focused) with a remarkable record, and is a close friend of Charlie's — he introduced BYD to Charlie, which is how Berkshire's BYD investment happened. He never joined; the scuttlebutt is compensation — keeping 2-and-20 on his own fund made Berkshire uneconomic, exactly the logic of Buffett declining to take over Graham-Newman years earlier ("why run your firm where you keep a piece, when I can do my own thing?"). The rumored other candidate was Greenlight's David Einhorn (pure internet rumor, never confirmed). Both Ben and David say they'd never heard the saga and "about lost it" reading it.
• Ted Weschler's $5.2M lunch: in 2010 and 2011, the same anonymous bidder wins Buffett's charity lunch for a record $2.6M each year (an eBay auction, benefiting San Francisco's Glide Memorial Church) — $5.2M for two meals, and a job. Ted had run the $2B hedge fund Peninsula Capital Advisors for 12 years — 12x over 12 years, a concentrated portfolio, top holdings DaVita and DirecTV. Ben: "Wait, Ted gets the job because he paid for lunches with him twice? ... $5.2 million for a job at Berkshire." The lore: both the airlines and the Apple ideas started with Ted — one bad, one good, and Apple covers everything.
• The Sokol–Lubrizol scandal: in 2011 Berkshire fully acquires the chemical company Lubrizol for $9B; the first person interested was the energy chief and presumed CEO successor, David Sokol (who came with MidAmerican in 1999 — Ben planted it early: "he's just a little too good, as we'll see"). He built a personal position in Lubrizol, then recommended Buffett buy the whole company, and kept buying during negotiations — front-running, "for some literally unfathomable reason, it's not like he needed the money." Not prosecuted, but he left Berkshire; Buffett said publicly he couldn't believe it. The envelope swaps to Sokol's former number two, Greg Abel.
• IBM: the 30-year itch: at their 1991 first meeting Buffett's very first question was about IBM, and Gates's answer was "absolutely don't"; in November 2011 he pulls the trigger anyway — $10.7B. His public thesis: "hit between the eyes by how great IBM is and how strong and defensible its client relationships are." By then AWS had already established the cloud paradigm — old switching costs are a negative asset in a paradigm shift (the last friction against customer escape, not a growth engine). David: "If this is Buffett's first technology investment, maybe it's a good thing he didn't make any before." Ben: "Maybe it's half a century too late." David: "Warren, go back to the monkey throwing darts."
• The Visa/MasterCard pain: these two, previously locked inside bank federations, IPO'd (2008 and 2006) and opened to the public for the first time — Ben: "These stocks were criminally undervalued initial issuances and Buffett just watches them go right by, knowing the AmEx business, it's crazy."
• Ajit profile: born in India → IIT → IBM → Harvard Business School → McKinsey (the transcript garbles it "McKenzie") → joined 1985. An underwriting savant — Ben: "You say 'I'm going to strap this guy to a rocket and shoot it at a hurricane, and I want insurance that bypasses the force majeure clause,' and he says 'I know exactly how to price this.'" David calls him a "Jack Ringwalt reincarnate" (the National Indemnity progenitor of pricing weird risk; the transcript badly mangles the name). For decades he and Warren ran the whole insurance portfolio and every live deal over a nightly 10:00 PM call.
• 100 shares of Microsoft: on the first trading day after the 1991 holiday, Buffett buys 100 shares of Microsoft for his personal account in his signature "fateful, immediate, split-second, gut" style (and zero Intel). Ben calls it "confounding": a double loss — the upside doesn't reach Berkshire shareholders, and being a shareholder actually blocks him from talking sensitive information with Gates.
• The 60-year gum loop: David ribs Warren for getting his start as a kid buying gum in bulk and breaking up the packs to sell singles — 60 years later he spends $6.5B back on gum: in April 2008 Mars buys Wrigley for $23B (Juicy Fruit — the transcript garbles it "JC Fruits" — Doublemint, Big Red, Extra, Orbit, and post-merger also Skittles, Starburst, Altoids, Lifesavers); Wrigley had over $5B in 2007 revenue (the transcript typos it "20017"). Ben, doing his Buffett impression: "people chew a lot of gum in the past and chew a lot of gum in the future." David: "No, Warren."
• Todd's new species inside: in 2020 (Ben's guess: Todd) buys Amazon, Snowflake, and other non-Warren-style tech stocks — Snowflake notably as pre-IPO shares, with real gains — "a non-Warren-approved strategy already running inside." Todd's résumé at hire was strikingly plain: Florida state bank regulator → Progressive Insurance → Castle Point Capital (~$100M, 34% cumulative over five years, not annualized — Buffett did nearly that every year for 12 straight years in the partnership days). David's after-the-fact verdict: "Todd is great. This was a good hire."
• Buybacks and the Bloomstran frame: in recent years Berkshire has bought back its own stock at scale — there's nothing better in the market than the businesses it already owns. Christopher Bloomstran (the "epic" annual Semper Augustus letter): "As long as capital markets remain overvalued and private investors flush with cash persist in investing at low yields, share repurchases are a magnificent use of capital." Ben's counter: there's tension in the logic — most of Berkshire's holdings are mark-to-market public equities at historically high multiples, so a buyback is adding to those same companies at market price.
• The Wells Fargo Bitcoin analogy: bought starting in 1989 and added to all the way up, with years of gains ultimately evaporated by scandal. Ben: "You start buying Bitcoin around $15,000 and yet you keep buying all the way up through $60,000, then you're probably about break even." — long holding doesn't exempt you from governance risk.

Era & Industry Trivia (tangents worth keeping)

• The full 1991 first-meeting scene: Kay Graham organized the Fourth of July gathering; Mary Gates forced her son to attend — history turned on her. Gates complained beforehand: "I don't know about meeting a guy who just invests in money and picks stocks... I don't have many good questions for him." He arrived late by helicopter, planning to fly back to Redmond right after dinner — the sun set, the helicopter left empty, and Gates stayed. At the Gates Sr. dinner, asked the single most important factor in getting where they'd gotten in life, Warren and Bill both blurted out focus — "two peas in a pod."
• Gates's prophecy and the DOJ: at the 1992 Vancouver Graham Group gathering, beyond "Kodak is toast," Gates gave a precise 30-year media forecast: the networks own the content and can repurpose it across media, while Kodak's core skill is "how to make film" — and if film disappears, that knowledge becomes utterly irrelevant. Same shock, two fates. David: "Gates just described the next 30 years of media and the Internet right there. This is 1992." But prophecy ≠ building — Ben counters with Buffett's own line: "Predicting rain doesn't count. Building the ark does." Gates saw media in 1992 and Jobs described the cloud in 1995, yet neither company became the winner of the domain it foresaw (David credits the DOJ antitrust case with cutting off Microsoft's "embrace and extend"). Ben on 90s Microsoft: "They were killers there... You should think about it like Uber in 2016."
• Don Keough and the most expensive "no" in history: Keough was Buffett's neighbor on Omaha's Farnam Street, a Butter-Nut coffee salesman and father of six. Buffett tried to recruit him into the Buffett Partnership; Keough asked his kids their impression of Mr. Buffett — "he's always home" — and concluded the man "clearly doesn't work very hard," and declined to invest. Butter-Nut was then bought by Duncan (1961), Duncan by Coca-Cola (1964), and Keough rose to president and COO of Coke — the Dan Burke to legendary CEO Roberto Goizueta's Tom Murphy. Years later they reconnected at a White House dinner, and Keough converted the daily-Pepsi-with-cherry-syrup Buffett with Cherry Coke. Shane Parrish's Farnam Street blog and The Knowledge Project take their name from the street the two once shared.
• New Coke and the Perelman to thank: in 1985 blind-taste data beats brand common sense and Coke pulls the old formula outright — Pepsi calls it "the greatest unforced error in history" and starts a price war, bottler conflicts erupt, the stock plunges, and Ron Perelman (the Revlon activist, former Marvel wrecker, a Salomon-era acquaintance) is rumored to be circling — which is exactly what creates the buy point for Buffett's 1987 white-knight entry. David: "I hope Warren and Charlie send a case of wine over to Ron Perelman someday — they got some deals because of that guy."
• The 1997 Sun Valley "ham sandwich" panel: Keough moderates; on stage sit Warren, Goizueta, and Gates. Gates goes off script, quoting Warren's own line back — your businesses could be run by a ham sandwich, whereas running Microsoft is a "high-wire act," and he expects he'll have to retire before 60 because tech needs young people to navigate constant change. Goizueta was 65, Keough 71, Warren 67 — "Gates is literally just slapping them all in the face here." The famously fiery Cuban-tempered Goizueta was hugely offended, never spoke to Gates again, and died suddenly of lung cancer later that same year. Warren just shrugged: "He's like a wild animal. You can't bring him in public too much." The verdict: "this is a major social faux pas on Gates' part, but he's totally freaking right."
• A footnote to the 1999 Fortune piece: the closing provocation — if he could go back to Kitty Hawk in 1903, he'd "do capitalists a favor and shoot the Wright brothers down." David: "Remember that line — he later buys every airline stock, twice." In March 2020 he dumps the airlines at the bottom, with only ~five days of buy-the-dip window before the Fed reflates — Ben: "I don't fault the sale (they could all have gone to zero without a bailout), I fault the buy — he'd long joked about a romantic fascination / dirty habit for airlines, and did it anyway. Thank you Jerome Powell." David: "It's like you can't have newspapers anymore, so he wants the airlines."
• The Giving Pledge mechanism and joke: launched 2010, four years after the 2006 mega-gift — the lag blamed on nobody wanting to make big gifts during the financial crisis. Ben: "Should have called it the Buffett Giving Pledge."
• Ho Nam and the Sequoia comparison: Altos founder Ho Nam's tweet, which Ben calls maybe the best one-line summary of Buffett anyone has: "He is the only investor to build a company worth over $500 billion. A few amazing founders have done it but no investor comes close." Ben on adaptation speed: "Sequoia always adjusts a year before the climate changes; Buffett is 30–40 years later — no, look at IBM, maybe 10–15 years later." At the close David realizes his carve-out — Phil Fisher's Common Stocks and Uncommon Profits (1958; the R&D paradox: not doing R&D costs more) — was also a Ho Nam recommendation. "My gosh. He's everywhere."
• How financial institutions die, and the Bitcoin Easter egg: Bear Stearns (like Salomon) died not of asset losses but of a counterparty trust collapse — $63 on Monday, bankrupt by Friday, with the Fed engineering a $2/share JP Morgan purchase backstopped by a government LLC absorbing the toxic assets, the template for crisis resolution. For a leveraged financial firm, trust is solvency. The crisis was also the starting line for Airbnb, Uber, and crypto — the Bitcoin genesis block embeds the headline "Chancellor on Brink of Second Bailout for Banks," widely read as a jab at fractional-reserve banking.
• The Marcus footnote: GE's consumer savings bank was sold to Goldman after the crisis, hastily rebranded GS Bank for a couple of years, and later became the foundation of Marcus — "they were both Warren Buffett bailouts, 2008 sweep-ups." (The transcript's "70 years ago" is likely a mishearing of "7 years ago.")
• Closing color: the episode opens by comparing itself to the Halo trilogy ("finish the fight"); the whole series ran 9–10 hours (David says 10, Ben corrects, "nine, David, don't be ridiculous"), hundreds of hours of research, six books; the finale recorded until 11 PM, trapped in the room for four hours. David riffs on T.S. Eliot to sign off: "Not with a bang but a whimper." Ben answers with the Jobs-ism, "the journey is the reward." Ben's carve-out Goodfellas came out in the early 90s — "shortly before the Salomon scandal"; David: "while Warren was still in his real heyday."

Cross-domain Notes

Honest verdict: no strong overlap — company history and investment methodology are essentially orthogonal to PH's geopolitical thesis, and no link is forced. Three weak resonances worth noting. First, this episode's mechanics of 2008 — financial firms dying of counterparty-trust collapse, the Fed backstop template, "injecting literally more money than God into the economy," the Bitcoin genesis block's jab at bank bailouts — is a business-side counterpart sample to PH's dollar-system and financial-power narratives. Second, the frame "the world's greatest status quo investor fails in a world of accelerating change" rhymes methodologically with PH's "an old paradigm can't price a new board" — the failure mode of a forecasting framework is itself transferable. Third, Buffett's meta-creed "believe in America" sits in contrast (contrast, not evidence) with PH's "American decline/restructuring" thesis — and David's rewrite of it into "never bet against the Internet" is exactly the move of migrating faith from an old vehicle to a new growth vehicle.

Pages Worth Creating

• Entities: Warren Buffett（沃伦·巴菲特）, Charlie Munger（查理·芒格） (trilogy person pages; the finale can crystallize career-level frames: status quo investor, the four-tier grading, the succession structure, "call Charlie early")
• Episodes: Berkshire 之一：Buffett 合伙基金年代（1930-1970）, Berkshire 之二：Munger 与伟大企业年代（1970-1992） (the first two extraction pages of the trilogy, cross-linked into a set with this one)
• Amazon.com：从网上书店到万物商店: the Buffett–Bezos thread runs through this episode — Bezos's "never rest on your laurels, change every day" intensity is the mirror image of the status-quo thesis ("This is not Coke"); the bear case's "internal growth engine" argument is measured directly against Amazon, and prior guest Brad Stone's Amazon Unbound is cited repeatedly
• Concepts: 7 Powers 护城河框架 (the series' standard framework; though this episode has no itemized segment, the Buffett halo as a countercyclical Cornered Resource — monetized, then zeroed, then expiring — is a distinctive case for the framework)

Source · acquired