In one sentence: Phil and Danielle go through the 2017 Berkshire meeting (Phil watched the live stream): why book value understates Berkshire, why Buffett won't promise to buy the stock at a set price, how he now sees Apple and IBM, what he and Munger missed in Amazon and Google, and a CEO-polling trick for sizing up an industry.
Key ideas
- Book value can hide real value. Accounting rules write down losers but never write up winners, so a business bought for $25M that now earns $65M a year can still sit on the books at $25M. Phil gives a land example (acreage bought in 1910 for a few dollars an acre that is now worth thousands) and says Berkshire has dozens of such holdings, so its intrinsic value is well above book. [02:00–07:00]
- No floor under the stock. Buffett has said he would buy Berkshire at about 120% of book. Phil puts book near $119 at end-2016, so about $144. Asked whether he would hold that price, Buffett said no: the board will let the price fall through it and buy at the best price it can. He would rather buy at 80 than 144. Don't assume a famous buyer sets a floor. [07:00–10:00]
- Buybacks only help below value. Phil says buying back stock at $300 when it is worth $200 hurts the shareholders who stay. Executives with options often prefer a high price, while long-term owners gain when the company buys below value. [10:00–12:00]
- IBM: a lower price helps the holder who stays. IBM has bought back about half its shares since 1996. After Buffett sold about a fifth of his stake and the stock dropped, the remaining shares benefit because later buybacks retire more shares per dollar. Phil notes Buffett kept a big stake and said he might buy more. [11:00–14:00]
- Is "tech" still a useful category? Buffett said Apple is now more like a consumer-goods company and IBM like a business-services firm. Phil's test: a true tech company must invent its next generation or die (Intel needs its own chips). Apple, Microsoft and IBM can buy or copy. Danielle replies that integrating and choosing acquisitions is still innovation, and that the real divide is between early-stage and entrenched companies. Phil as a user is wrapped up in Apple's ecosystem, which is a switching moat. [12:00–19:00]
- Scuttlebutt, Buffett's version. To learn an industry, ask each CEO: if you had to put money in a competitor for ten years, which one? And which would you short? Phil and Danielle doubt real CEOs would answer honestly or anonymously, but it is a good way to think about who is respected. [19:00–22:00]
- Missing Amazon and Google. Buffett said he was too dumb to see Bezos's scale, and that Amazon's 1997 annual report laid it out. Munger said Amazon wasn't an obvious call and that Google was the real miss: Geico advertised on it, and every click paid Google. Phil says Google was clearly a major franchise after its IPO, and the open question was whether anyone would take its place. "We don't miss them all." [22:00–24:00]
- Teased for next time. Market-wide valuation, Buffett's legacy, EBITDA, USG and financial advisors. These appear in 113. [24:00–25:00]
How it maps to RuleOne
- Book value is a poor proxy for value, which is why the screen's value work (Sticker Price and the MOS price) is built from earnings and growth, not from the balance sheet alone. Use book value as a cross-check only.
- When a stock page shows buybacks, compare the average buyback price with your own estimate of value, as in the IBM discussion above.
- The tech debate is a circle-of-competence question (001): the label matters less than whether you can judge the business and its moat.
Buffett, Munger and Graham links
- Buffett's mention of intrinsic value versus book value is a recurring theme in the Berkshire chairman's letters (the owner's manual at the front of every letter); don't cite a number from this episode without checking the 2016 letter.
- "Scuttlebutt" is Philip Fisher's term (Common Stocks and Uncommon Profits, chapter 1); Phil credits the CEO question to Buffett at this meeting, so treat it as Phil's account.
- Buffett has said several times that he missed Google and other tech names. In Phil's account, it was Munger who named Google at this meeting. The exact words are paraphrase.
Words to know
- Book value: assets minus liabilities as reported by the accountants, which is often not what a business is worth.
- Buyback: a company buying its own stock, which helps remaining shareholders only if done below intrinsic value.
- Scuttlebutt: Fisher's method of learning about a business by asking customers, competitors and suppliers.
Try this
Pick one company on /stocks/ that does regular buybacks. On its stock page, compare book value per share, the price paid for buybacks (look at shares repurchased and dollars spent in the cash-flow statement) and your own estimate of value. Write one line: did the buybacks happen below or above value?
Check yourself
- Why can Berkshire's book value understate its worth?
Answer
Accounting writes down failures but won't write up wholly owned winners, so businesses bought cheaply stay on the books at cost. - Why did Buffett refuse to promise a floor for Berkshire's stock?
Answer
Buying lower is better for the continuing shareholders, and a promise would let others buy early and pre-empt him. - When do buybacks help the shareholders who stay?
Answer
Only when the company buys below the intrinsic value of the business.
Short quotes
"He's a brilliant execution… it's all right there in the 1997 annual report." (Phil, paraphrasing Buffett on Amazon, ~22:30, auto-transcribed)