Five sigma — the book and the record it must explain
Chapter 1 of 16 · 11 min
When Robert Hagstrom published The Warren Buffett Way in 1994, no book existed that systematically accounted for how Buffett actually worked — only legends. Hagstrom, an analyst who later became a focus investor himself, wrote the book from public material: annual letters, purchases, speeches and accounts. This chapter presents the book's project, its three editions and the record the book must explain — according to Hagstrom, a five-sigma event.
The book's introduction opens with a question the course keeps for life: is Warren Buffett an exception that cannot be copied, or is his process a discipline anyone can learn — even if the outcome will never be the same? Hagstrom's answer is that the process is learnable but that it demands a temperament most people lack. To prove it, he divides his material into four parts: the education that shaped Buffett, twelve principles for buying businesses, the portfolio mathematics and the psychology.
The foreword to the third edition (2013) is written by Howard Marks, the second (2005) by Bill Miller, the first by Peter Lynch — three names that have themselves run billions and who all share the same conclusion: the record is extreme, but the method is public.
The record is the book's evidence and its problem. Buffett Partnership, started in 1956 with little more than a hundred thousand dollars of family and friends' money, returned on average over thirty percent per year for thirteen years against Dow Jones's barely ten — without a single losing year. Thereafter the book value per share of stock at Berkshire Hathaway compounded from 1965 at an annual pace of nearly twenty percent and the share price just above that, against an index of nine to ten percent including dividends — for almost half a century.
Such a curve, Hagstrom writes, is a five-sigma deviation: in a normal distribution it should occur roughly once in three and a half million. The course's controversy chapter (chapter 15) will take that number seriously — partly as tribute, partly as accusation.
The book exists in three versions and the course covers them as a whole. The 1994 edition established the twelve principles and the five original cases: Washington Post, GEICO, Capital Cities/ABC, Coca-Cola and General Dynamics. The 2005 edition added, among other things, Wells Fargo and American Express after the banking crisis of the 1990s and brought in the portfolio mathematics with the Kelly criterion.
The 2013 edition — the one the course mainly follows — added IBM in 2011 and H.J. Heinz in 2013 and updated the research on concentration. Hagstrom writes no biography: on the sugar, the father and Susie the course refers to the Snowball course; on the letters themselves to the Essays course. Here it is the process: what did he do, in what order and with which numbers.
For the AK1A ecosystem this book is especially useful for one reason: Hagstrom formulated a checklist a quarter century before checklists became investing's buzzword. The twelve principles — three on the business, three on management, four on the finances, two on the market — are in practice a structured fundamental model.
The course will therefore show, for each principle, where it lands in AKM1's twenty variables: profitability in V07-V09, the moat in V13-V15, capital allocation in V20 and capital burn in V19. That is the course's own thesis: Hagstrom showed that Buffett is a process; AKM1 makes the process measurable.