The investor library

Warren Buffett

1930- · Berkshire Hathaway · The longest record there is.

Overview

Has run Berkshire Hathaway since 1965, compounding shareholder value at roughly twice the rate of the American market over sixty years — the longest such record there is. He is the most studied investor alive and the least usefully copied.

“It's far better to buy a wonderful company at a fair price than a fair company at a wonderful price.”

Warren Buffett

Background

Omaha, 1930, the son of a stockbroker who became a congressman. He filed a tax return at thirteen, claiming his bicycle as a deduction. Rejected by Harvard Business School — which he later called the best thing that happened to him — he went to Columbia instead, to study under Benjamin Graham, then worked for him at Graham-Newman. He ran his own partnerships from 1956 to 1969, compounding at roughly 29% a year before fees, and then closed them: he said he could no longer find anything worth buying. Along the way he had taken control of a failing New England textile mill called Berkshire Hathaway, which he has described as the worst investment he ever made.

Style, and how it evolved

He began as an orthodox Graham cigar-butt investor: buy statistically cheap, sell on reversion, feel nothing. Two people moved him. Munger argued that a great business at a fair price beat a fair business at a great one; Philip Fisher's writing supplied the vocabulary for judging one.

See's Candies, 1972, is the hinge. Berkshire paid $25m for a business earning around $4m before tax — several times book value, a price the Graham method forbids outright. Munger had to talk him into it. It taught him what pricing power is worth, and he has said plainly that without See's he would not have bought Coca-Cola.

The later evolution is about size rather than philosophy. As capital grew he moved from buying pieces of companies to buying whole ones, and from an edge in obscure securities to an edge in being the buyer who can act instantly at scale. He is candid that this costs him returns: the partnership-era numbers, he has repeatedly said, are unrepeatable at Berkshire's size, and anyone quoting the sixty-year figure should know most of it was earned in the early decades.

Performance

paid offa decisiona losswalked awaythe lifeshaded columns are the crashes — hover any mark

1965-2023: roughly 19.8% a year in Berkshire's per-share market value, against about 10.2% for the S&P 500 with dividends reinvested. Compounded across fifty-eight years, that gap is the difference between about 4,400,000% and about 31,000%.

✓ documented · ○ rests on secondary accounts rather than a filing

Case studies

1964 — American Express. A subsidiary had accepted warehouse receipts for salad oil that did not exist, and the fraud threatened to bankrupt the parent. The shares halved. Buffett went to Omaha restaurants and steakhouses and watched whether people were still paying with the card and buying travellers cheques. They were. He concluded the loss was a one-off liability and the franchise was untouched, and put about 40% of the partnership into it. The mechanism: separating a solvency event from a franchise event, and checking it by observation rather than inference. ✓

1972 — See's Candies. Covered above; the point here is the price. He has said the lesson took him twenty years to fully absorb: a business that can raise prices every year without losing customers is worth a multiple of one that cannot, and the multiple you pay matters less than that difference. ✓

2008 — Goldman Sachs. Five billion dollars of preferred stock paying 10%, with warrants attached, agreed in a matter of hours during the worst week of the crisis. Note what this trade actually required: not analysis, but being the only counterparty with cash and a reputation that made the money worth more than the money. It is the clearest evidence for the "machine" side of the open question. ✓

The other side of the record

A record this good is where scepticism is most worth spending, so:

A large part of the alpha has been explained. The academic work — Buffett's Alpha, by Frazzini, Kabiller and Pedersen — finds that his returns are substantially accounted for by leverage of roughly 1.6 to 1.7 times, sourced cheaply from insurance float, applied to a systematic tilt toward high-quality, low-volatility, cheap stocks. That does not make the record less real. It does mean the transferable part is "cheap leverage plus a quality tilt, held for decades", which is a much more mundane sentence than the folklore.

The best years were the small years. Compounding at 29% on a few million dollars in an under-covered 1950s market is a different activity from allocating hundreds of billions. Berkshire's returns over the last two decades are much closer to the index.

The advantages are structural and unavailable. Permanent capital that cannot be redeemed in a panic, float at negative cost, deal terms offered to him because he is him, and the latitude to do nothing for years without losing clients. Almost nobody investing has any of these.

The misses are substantial. Dexter Shoe, bought with Berkshire stock, which he has called the worst deal he ever made; IBM; the airlines round trip; the Precision Castparts writedown; and four decades of avoiding technology followed by a single enormous bet on Apple that has flattered the recent record.

The plainness is itself an asset. The folksy Omaha register makes insurance leverage sound like thrift. It is one of the most effective pieces of positioning in finance, and it is not an accident.

Key lessons

• The durable competitive advantage is the whole question. The structural reason a business can keep earning well without competitors arriving to compete the returns away. Growth without one is temporary by definition, however impressive it looks while it lasts. It is the difference between a good few years and a good business.
• Temperament beats intelligence. His stated view, repeatedly: the required IQ is ordinary, and what is scarce is the ability to do nothing while others act.
• Inactivity is a strategy. Most of the sixty-year record comes from a very small number of decisions held for a very long time.
• Structure determines what you can do. The permanent capital came first; the patience it permits is downstream of it. This is the least-copied part of him and possibly the most important.
• Stay inside what you understand, and say so out loud. The circle of competence is useful only if its edge is admitted.

Reading and links

• The Berkshire Hathaway shareholder letters, 1977 onwards — free, and the primary source for everything else on this page.
• The Essays of Warren Buffett — Lawrence Cunningham. The letters, rearranged by subject, which is how they are actually usable.
• The Snowball — Alice Schroeder. The authorised biography, and unflattering enough to be worth reading.
• Buffett's Alpha — Frazzini, Kabiller and Pedersen. The adversarial read, and the best one.

Marked ✓ where the figure is documented and ○ where it rests on secondary accounts rather than a filing. This is a profile of an investor, not a view on any security.

Primary sources

Go to the thing itself. These are the subject’s own publications or an institution’s own site. No bookseller links, and no referral arrangements — a reading list that earns per click is not a reading list.

• The Berkshire shareholder letters, 1977 onwardsFree, and the primary source for most of this page.
• Berkshire annual reports

Profiles of investors, not views on any security, and not personal advice. Figures are marked ✓ where they are documented and ○ where they rest on secondary or private accounts — investing biography is heavily mythologised and the well-known numbers drift with each retelling. Capital is at risk.

Corrections and right of reply. Where a criticism is made of a named person it is stated as a specific measure over a stated period, attributed to its source, and separated from opinion. Assessments are opinion, honestly held, on facts believed accurate at the date shown. If anyone profiled here — or anyone acting for them — believes a fact is wrong, it will be corrected promptly and visibly, and a reply will be published alongside it on request.

← John BogleCharlie Munger →