🎧 Listen to the Discussion(two hosts talk through their thoughts on this article)
Audio only — for listening on a walk or while doing chores
This is an AI-generated audio overview of this article, created with Google NotebookLM. Two hosts discuss the article together, sharing their own thoughts and commentary — it is not a reading of the article text.
This article sets out my own thoughts, based on information publicly available as of November 2026. It is not a recommendation to buy or sell any share. I myself own a small number of Berkshire Hathaway B shares.
A $3 Breakfast and Japan's Trading Houses
Some years ago I watched a documentary that followed Warren Buffett's daily life. I have forgotten its title, but it showed the everyday routine of the "Oracle of Omaha". What has stayed with me is how plainly he lived, nothing like one of the richest men in the world. In the morning he drove himself to the office, bought breakfast at a McDonald's drive-through on the way, and drank Coke at work. When the market was doing well he chose a slightly dearer breakfast; when it wasn't, a cheaper one. I couldn't help laughing.
On 30 August 2020, his 90th birthday, this same frugal man announced that he owned about 5% each of Japan's five big trading houses: Itochu, Marubeni, Mitsubishi Corporation, Mitsui & Co. and Sumitomo Corporation. He kept buying, and by May 2026 Berkshire held more than 10% of all five. Why would a man who chooses a $3 breakfast put more than $10 billion into trading houses on the far side of the world, and even say he would not sell for 50 years? That question is the core of this article, an attempt to dig deep into the company called Berkshire Hathaway.
Tokyo Station at night. Most of the five trading houses have their head offices within walking distance, in the Marunouchi and Otemachi districts.
Photo by 潇 文 on Unsplash
On 18 September 2026, Buffett, aged 96, stepped down as chairman and became chairman emeritus. Less than a year and a half after announcing in May the year before that he would step down as CEO, he had handed over every rein he had held for 61 years. Reading the news, I wanted to trace this company's roots and growth rings with my own eyes. Before turning to the trading houses, let me look at how the tree called Berkshire grew. That makes it much easier to see why this tree sent its roots into Japan's trading houses.
Over 90% of His Wealth Came After His Mid-Sixties
I had long heard that most of Buffett's wealth came after his sixties. When I looked into it, it turned out to be true.
In his 2020 book The Psychology of Money, the American writer Morgan Housel noted that of Buffett's then net worth of $84.5 billion, $81.5 billion had been added after his mid-sixties, the age at which he could have drawn a state pension. That is about 96%. In 2025 an American asset-management executive estimated that Buffett was worth about $3 billion at 65, and that 98% of his $160 billion at 94 had come after that age. On top of that, Buffett has been giving his shares to charity since 2006, more than $60 billion so far. Add back what he has given away, and the growth after 65 is even larger.
Why does this happen? The answer lies in the mathematics of compounding. Over the 61 years from 1965 to 2025, Berkshire's market value per share grew by an average of 19.7% a year. Over the same period, the S&P 500, America's benchmark index, returned 10.5% a year including dividends4801.
To feel the difference, the "rule of 72" helps. Divide 72 by the annual rate and you get roughly how many years it takes money to double. At 19.7% that is about 3.7 years; at 10.5%, about 6.9 years. As you double, double and double again, the last few doublings are bigger than all the earlier ones put together. For someone with $10 billion, the next doubling adds $10 billion; for someone with $100 billion, it adds $100 billion. Buffett bought his first share at 11 and was still investing at 96. His wealth seems to have shot up after his sixties not because it started growing differently, but because it kept growing at the same rate for an extraordinarily long time. Housel puts it roughly this way: Buffett's skill is investing, but his secret is time.
If something grows at 19.7% a year for 61 years, this is the decade in which each part of the gain was made. The last ten years alone account for more than 80% (my calculation; a simplified chart).
At 66, I find this story a little bittersweet, and yet encouraging. Even if I started investing now, I would not have Buffett's kind of time left. But what compounding teaches is the value of not stopping halfway. Keep holding instead of selling, keep learning, keep moving your body. In each case, the biggest harvest comes in the later years of the effort. And this idea, that the longer you hold, the bigger the harvest, leads straight to the 90-year-old Buffett saying he would hold Japan's trading houses for 50 years.
It Began with an Eighth of a Dollar: The Textile Years
Berkshire Hathaway was not originally an investment company but a textile maker in New England, in America's north-east. It was formed in 1955 by the merger of Berkshire Fine Spinning, which descended from a spinning company founded in 1839, and Hathaway Manufacturing, founded in 1888. By then, though, the American textile industry was already in decline. When Buffett, then 32, began buying the shares in 1962, it was not because he believed in the business's future. It was because the shares were far too cheap compared with the mills' assets. It was the "cigar butt" style of investing he had learnt from his teacher 1894–1976. American investor who taught Buffett at Columbia University. Regarded as the father of "value investing", buying shares when the price is well below the company's assets or earnings. His book The Intelligent Investor is still widely read.: pick up a discarded butt and get one last free puff.
The turning point came in 1964. The company's president asked Buffett what price he would sell his shares back to the company for, and they nearly agreed on $11.50. But the tender offer that actually arrived said $11.375, an eighth of a dollar (12.5 cents) less. Feeling short-changed, Buffett bought more instead of selling, and took control on 10 May 1965. He later called it "the dumbest stock I ever bought", and said that had he not tied up money in a textile company but bought an insurance company from the start, Berkshire might be worth twice as much today.
Reading this, I first smiled wryly, then put a hand to my heart. My own life has a few forks where I refused to turn back out of stubbornness. It is easy to call a decision right because it turned out well. But I sense this man's honesty in the fact that Buffett himself went on calling it a "mistake" to the very end.
The Invention Called Float: Insurance's "Money Waterwheel"
Instead of reviving the textile business, Buffett chose to move the cash it produced into other businesses. The first big step was National Indemnity, an Omaha insurer bought in 1967 for $8.6 million. It is this same company that now holds Berkshire's shares in Japan's five trading houses.
What makes insurance interesting is that money comes in and goes out at different times. Premiums are collected first; claims are paid only after an accident or disaster. In between, money builds up that will be paid out one day but is in hand for now. This is called Money an insurer holds between receiving premiums and paying claims. If underwriting stays profitable, it works like a loan that carries no interest.. At the end of September 2026, Berkshire's float stood at about $178.5 billion. As long as underwriting does not lose money, this money can be used like an interest-free loan.
Berkshire's "money waterwheel". As long as underwriting stays profitable, the axle (the float) keeps turning for free (a simplified diagram).
Looking into it, I found that insurance itself grew up on a street corner in early modern England. At the end of the 17th century, shipowners and merchants gathered at Edward Lloyd's coffee house in London, and those willing to cover the loss if a cargo did not arrive safely wrote their names at the bottom of a sheet of paper. That is where the word "underwriter", one who writes underneath, comes from. Insurance has always been a business that sells an invisible promise. What Berkshire gained was the time lag that promise creates. This idea of holding good things for a long time with cheap money later reappeared, in another form, in the way it bought Japan's trading houses.
When I ran my own business, I learnt first-hand how welcome the cash in hand is when customers pay you first and you pay your suppliers later. Insurers do this on a vastly larger scale, and for decades. I came to feel that the accumulation of quiet, patient work is this company's real strength.
Charlie Munger, the Partner: "Architect" and "General Contractor"
Tell Berkshire's story as Buffett's alone and you tell only half of it. The other half belongs to its vice chairman, Charlie Munger.
Two Boys at the Same Grocery Shop, Six and a Half Years Apart
Munger was born in Omaha on 1 January 1924, six and a half years before Buffett, who was born on 30 August 1930. As boys, both worked at the grocery shop run by Buffett's grandfather, but because of the age gap they never met there. Munger studied mathematics, served as a meteorologist in the Army Air Forces during the war, graduated from Harvard Law School afterwards and became a lawyer in Los Angeles. His first marriage ended in divorce, and in 1955 he lost his nine-year-old son to leukaemia.
1959: A Dinner in Omaha
The two first met in 1959 at a dinner in Omaha. Buffett later said that within minutes of starting to talk, he knew this would be a lifelong friendship. Although they lived more than 2,400 kilometres apart, in Omaha and Los Angeles, they kept in touch through frequent phone calls and long letters.
See's Candies: The Turning Point
What changed the very shape of Berkshire was the purchase of See's Candies in 1972. Paying $25 million for a chocolate maker earning about $4 million a year before tax looked far too expensive by Graham's "buy it cheap" standard. It was Munger who persuaded a hesitant Buffett. The purchase was made through A Californian company whose trading stamps supermarkets handed to customers. Money stayed in hand until the stamps were exchanged for goods, and Buffett and Munger invested it. It later merged into Berkshire., a company both men were involved in.
See's had the "power of the brand": customers stayed even when prices rose, and it could send almost all the cash it earned to other businesses. Buffett later wrote, in essence, that it is far better to buy a wonderful company at a fair price than a fair company at a wonderful price, that Charlie understood this early, and that he was a slow learner. That lesson led on to Berkshire's investment in Coca-Cola in 1988.
A 45-Year Partnership
In 1978 Munger became vice chairman. Living apart, they always talked by phone before any big decision, and the partnership lasted 45 years. At the annual meeting each May they sat side by side on stage and answered questions for hours. When Buffett finished a long answer, Munger would reply with a single line: "I have nothing to add." Tens of thousands of shareholders came to Omaha from around the world to enjoy that double act. Buffett said they had not argued once in almost 60 years. Yet Munger was no copy of Buffett. All his life he urged people to combine ideas from many disciplines, and to turn problems upside down and start from "How could this fail?"
A Last Word on the Trading Houses
In October 2023, the 99-year-old Munger gave an interview at his Los Angeles home to an American podcast and said of the trading-house investment that it was like having God open a chest and pour money into it. He also said it was a big flow of money that cost almost nothing, and that even someone as clever as Buffett gets an idea like this only two or three times a century. A few weeks later, on 28 November, Munger died, 33 days short of his 100th birthday. The trading-house investment was one of the last big ventures the two of them enjoyed together.
"Architect" and "General Contractor", and a Friendship Like Guan and Bao's
In his shareholder letter the following February, Buffett called Munger the "architect" of today's Berkshire and himself merely the "general contractor" who carried out the plans day by day. The line that stayed with me most was that Munger, even when he knew he was right, left the reins to Buffett, and never once brought up Buffett's mistakes afterwards.
Reading it, I remembered the old Chinese phrase "the friendship of Guan and Bao". In China's Spring and Autumn period, the young Guan Zhong traded with his friend Bao Shuya and always took the bigger share of the profit, yet Bao never blamed him, because he knew Guan's family was poor. Guan is said to have remarked later: "My parents gave me life, but it is Bao who truly knows me." A friend who does not count your failures may, for more than two thousand years, have been the best fertiliser for helping a person grow.
Another part of Munger's life struck me hard: in his fifties a cataract operation went wrong and he lost the sight of his left eye. Later his right eye weakened too, and he began learning braille (in the end his right eye kept its sight for the rest of his life). When I had my own cataract operation in Osaka, I was moved when, with the eye patch off, I could see the faces in the waiting room clearly. But before the operation I was worried about what would happen if it went wrong. I can imagine, just a little, with my own body, the quiet strength of a man who began learning braille in case the worst happened, yet never stopped reading.
Why Does Buffett Keep Buying Japan's Five Trading Houses?
This is the core of the article. First, here is how the buying unfolded over time.
| When | What happened |
|---|---|
| July 2019 | Begins buying the five trading houses' shares on the Tokyo market |
| 30 Aug 2020 | On his 90th birthday, announces stakes of about 5% in each, and promises not to go above 9.9% |
| April 2023 | Buffett visits Tokyo and meets the heads of the five. Average stake 7.4% |
| June 2023 | Stakes at 7.47–8.31% (announced on 19 June) |
| Feb 2025 | In the letter for 2024, says the companies have agreed to relax the 9.9% cap a little |
| March 2025 | Stakes at 8.5–9.8% |
| May 2025 | At the annual meeting, Buffett says he would not think of selling for the next 50 years; Abel adds "50 years, or forever" |
| Aug–Sept 2025 | Mitsubishi and Mitsui pass 10% for the first time |
| May 2026 | Marubeni and Sumitomo also pass 10%, so all five are above 10% |
| Sept 2026 | Abel makes his first visit to Tokyo as CEO, says Berkshire intends to hold "for decades", and is considering buying more |
In Japan, anyone who holds more than 5% of a listed company must file a A filing required within five business days of holding more than 5% of a Japanese listed company. A change report follows each time the stake moves by 1 point or more. Anyone can read them on the Financial Services Agency's EDINET system.. Berkshire's changing stakes can be traced mainly through these reports and its shareholder letters. Company by company, they look like this.
| When | Itochu | Marubeni | Mitsubishi | Mitsui | Sumitomo |
|---|---|---|---|---|---|
| Aug 2020 | about 5.0% | about 5.0% | about 5.0% | about 5.0% | about 5.0% |
| June 2023 | 7.47% | 8.30% | 8.31% | 8.09% | 8.23% |
| March 2025 | 8.53% | 9.30% | 9.67% | 9.82% | 9.29% |
| End 2025 | 10.1% | 9.8% | 10.8% | 10.4% | 9.7% |
| Latest in 2026 | 10.07% (27 Feb) | 10.32% (late June) | 11.06% (30 April) | 10.83% (late June) | 10.30% (12 May) |
(In August 2020 all five were 5.02–5.06%. The end-2025 figures are from the shareholder letter for 2025, calculated excluding treasury shares, so they come out slightly higher than in the large-shareholding reports. The rest are from those reports and press coverage. The 2026 figures match "Japan's Five Trading Houses: Results Notes".)
Berkshire's stake in each of the five trading houses. Mitsubishi and Mitsui passed 10% first, and all five were above it by May 2026 (a simplified chart).
Laid out like this, you can see the buying was slow for the first three years or so, then sped up after Buffett's Tokyo visit in 2023. Once the 9.9% cap was relaxed in 2025, Mitsubishi and Mitsui went past 10% first, followed by Itochu, Marubeni and Sumitomo. While treating all five alike, Berkshire seems to have varied the order a little according to how cheap each share was.
The total cost was $13.8 billion at the end of 2024, when the stakes were worth $23.5 billion. By the end of 2025 their value had grown to about $35.4 billion, and in early 2026 it was reported to be approaching $40 billion. That is roughly three times the money put in. So why did Buffett choose these five, and why does he keep buying? Looking into it, I found at least five reasons.
Reason 1: The Price Was Unbelievably Low
The first reason is very simple. They were cheap. Interviewed in Tokyo in April 2023, Buffett said he had been "puzzled" that he could invest in these companies at all. Compared with how low interest rates were, the shares were so cheap relative to the profits the companies earned that the Earnings per share divided by the share price; the inverse of the price-earnings ratio. A P/E of about 7 means an earnings yield of about 14%: the share of the price earned in profit each year. was about 14%.
An earnings yield of 14% means the shares cost only about seven times earnings per share. What is more, many Japanese listed companies were trading close to or below their net assets, a level at which shareholders would get more if the company were wound up and its assets handed out. Japanese shares had long been known for being "left cheap and ignored". Buffett says he looked through the accounts himself, company by company, and was surprised at how cheap they were. The eye for buying cheaply that he learnt from Graham as a young man and the eye for buying good companies that he learnt from Munger lined up perfectly in Japan's trading houses.
Reason 2: Borrow Yen to Buy Yen Shares
The second reason lies in the way it bought. Berkshire did not raise the money by changing dollars into yen. Berkshire Hathaway itself, in Omaha, issued yen bonds to borrow yen, and used the yen to buy the shares. Those lending the money by buying the bonds are mainly Japanese investors such as banks and insurers.
Borrow yen, buy yen shares. Currency gains and losses largely cancel out, and the gap between dividends and interest is left over (a simplified diagram).
This arrangement has two advantages. First, currency moves largely cancel out. If the yen weakens, the shares are worth less in dollars, but repaying the yen debt also becomes lighter in dollar terms by about the same amount. If the yen strengthens, the opposite happens. In fact, the yen fell sharply against the dollar from 2020, but by the end of 2024 Berkshire had made more than $2 billion after tax in currency gains on its yen debt. The debt side made up for what the shares nearly lost.
Second, there is the gap between dividends and interest. In the shareholder letter for 2024, Buffett wrote that in 2025 Berkshire expected to receive about $812 million in dividends from the trading houses and to pay about $135 million in interest on its yen debt. The dividends come to about six times the interest. This was the flow Munger meant when he spoke of God opening a chest. As of September 2026, Berkshire's yen bonds outstanding exceeded $15 billion.
This shares a root with the idea of float. In insurance, Berkshire used money it held almost for free. In Japan, it borrowed in one of the lowest-interest countries in the world and used the money to buy companies paying dividends well above the interest. Hold good things for a long time with cheap money. The insurance invention of 60 years earlier had flowered again, in another form, in Japan.
Reason 3: They Look a Lot Like Berkshire
The third reason is that the five trading houses are built much like Berkshire. Buffett has described them as companies that, like Berkshire, invest in a great many businesses. Resources and energy, food, machinery, chemicals, retail and finance: they hold hundreds of businesses around the world, so when one struggles, others hold it up. This shape, a tree with many roots, is Berkshire itself, with its insurance, railway, utilities, manufacturing and retail. Buffett is known for not investing in businesses he does not understand, and the trading houses were easy to understand: almost a mirror image of his own company.
Looking into it, the history of the trading houses goes back to Japan's early modern merchants. Itochu and Marubeni both trace their founding to 1858, when the first Chubei Itoh, from Ōmi (today's Shiga Prefecture), began travelling to Osaka and Kyushu with linen cloth on his shoulders to sell. Mitsui & Co. was founded by Takashi Masuda in 1876, Mitsubishi Corporation grew out of Yataro Iwasaki's shipping business, and Sumitomo Corporation grew from a company founded in Osaka in 1919.
What drew me most was the idea of sanpō yoshi, "good for three sides", said to have been cherished by Ōmi merchants like Chubei Itoh: good for the seller, good for the buyer and good for society. The teaching is to trade in a way that benefits not only yourself but also your trading partner and the people of the place where you do business. It somehow echoes the way Buffett never haggles a seller down, trusts the managers of the companies he buys and has always called shareholders "partners". Businesses that last a long time may arrive at similar shapes, even in different countries.
Ōmi merchants: traders based in Ōmi (today's Shiga Prefecture) who travelled the country from the Edo period to the Meiji era, carrying goods on shoulder poles. They were the forerunners of Itochu, Marubeni, the Takashimaya department stores and others.
Just as important, the trading houses had already changed from companies that "moved goods from right to left and took a commission" into companies that "invest in businesses and grow them". From the late 1970s to the 1990s, as manufacturers began handling their own exports and imports, there was talk that trading houses would no longer be needed. In response, they shifted from brokering trade to developing resources and investing in companies. It was this transformation that let Buffett see the trading houses as Berkshire's kin.
Reason 4: Management That Values Shareholders
The fourth reason is the way they are run. In his letters, Buffett has praised the five for raising dividends, buying back shares when appropriate, and paying their managers far less than comparable American companies26687. In other words, they return what they earn to shareholders rather than to their managers' pockets.
Behind this lies a wider change in Japan. In 2015 a Corporate Governance Code setting out principles for listed companies came into force, and in March 2023 the In March 2023 the Tokyo Stock Exchange asked companies listed on its Prime and Standard markets to manage with their cost of capital and share price in mind, and to disclose plans and concrete steps to improve. Many companies were trading below book value. began asking companies trading below their net assets, among others, to set out plans for improvement. The era of Japanese shares being "left cheap and ignored" is slowly coming to an end. 2019, when Buffett started buying, was just when that change began. Shares bought cheaply then gain further value as the companies change. Reasons 1 and 4 are two sides of the same coin.
Reason 5: A Partner for "Something Big" Together
The fifth reason is a relationship that goes beyond money. At the 2025 annual meeting, Abel said Berkshire was building relationships with the trading houses and hoped one day to do something big together. On his Tokyo visit in September 2026, he revealed that talks were continuing about joint investments and acquisitions around the world. In March the same year, Berkshire also took a 2.49% stake, worth about $1.8 billion, in the insurer Tokio Marine Holdings, agreeing to work together on reinsurance and acquisitions.
At the 2025 meeting Buffett said Berkshire's role was to "cheer and applaud" the trading houses. A major shareholder that does not interfere in management, yet can look at deals around the world with them when needed. From the trading houses' side, too, a shareholder that keeps its promise on the cap and says it will not sell for decades is reassuring. In September 2026 Itochu said Berkshire might raise its stake further. The last reason it keeps buying is that the trading houses want it to.
A Changing Premise: When Interest Rates Rise
Of course, not everything is rosy. The biggest change is Japanese interest rates. Berkshire's ten-year yen bond issued in November 2025 carried a coupon of 2.422%; the one issued in April 2026 paid 3.084%. Japan's ten-year government bond yield also rose above 3% in 2026, for the first time in about 30 years. As the interest Berkshire pays on each new yen loan rises, the gap with the dividends it receives narrows. That said, the coupons on bonds already issued are fixed at the rate set when they were issued. Higher rates only bite on new borrowing and on bonds refinanced when they mature, so the gap between dividends and interest narrows gradually rather than all at once. Abel has said the trading houses' managers do not yet see this as a big problem, and that borrowing in yen is still the right approach. But "God's chest" is no longer quite as one-sided as it was.
The other point is that much of the trading houses' income depends on the prices of resources and energy. If oil, copper or iron ore fall, so do their profits. But when Buffett says "50 years", he is not looking at the peaks and troughs of a few years. He is looking at whether these companies can keep raising dividends through many peaks and troughs.
Whole in Britain, a Tenth at a Time in Japan
Side by side, the way Berkshire has put down roots in Britain and in Japan is strikingly different. In Britain it owns the electricity network Northern Powergrid outright and runs it itself. In Japan it holds about a tenth each of trading houses and an insurer, stays out of management and has kept its promise on the cap.
The difference seems to reflect how companies are seen in each country. Britain sees a company as belonging to its shareholders, and buying and selling whole companies is common. Even electricity and water were bought outright by foreign capital after privatisation. Britain, too, once had great trading companies resembling Japan's, such as the East India Company and, in Asia, Jardine Matheson and Swire. But at home, trade split into specialist companies, and the "trade in everything" form never grew as large as in Japan. That is why English has no word for a Japanese general trading company and simply calls it a sogo shosha. Large Japanese companies, meanwhile, long held each other's shares with their business partners and banks, and avoided being bought outright from outside. In Japan, Berkshire was welcomed not as a raider but as a quiet major shareholder that stays alongside for the long term. I take Abel's "50 years, or forever" as words that stand on that trust.
As a Japanese, I read with some pride that Buffett called Japan's trading houses companies "much like Berkshire". Japanese companies that the world's investors had long ignored were found by a 90-year-old investor who read their accounts one by one, and who said he would hold them for 50 years. At the same time, seeing the five companies' share prices move by several per cent on a single word from Berkshire, I also feel the weight of a major shareholder's words.
Reading the Shareholder Letters, from the First in 1965
Berkshire's shareholder letter is the letter from management to shareholders at the front of each annual report. Buffett's letters called shareholders "partners", counted his own mistakes unprompted, and avoided jargon. The letters from 1965 to 1969 were in fact written by Buffett but signed by the company's president, Ken Chace. The letters from 1977 onwards are on the company's website, free for anyone to read.
Sixty-one years of letters: one long correspondence from a single writer to his shareholders.
Photo by Towfiqu barbhuiya on Unsplash
Here are twelve landmark letters, one line each.
| Year (date) | Writer | Gist |
|---|---|---|
| 1965 (9 Nov) | Signed by president Ken Chace (written by Buffett) | A change of management. A fresh start for a textile company that had paid off its debts |
| 1970 | Buffett (the first in his own name) | Three pillars: textiles, insurance, banking. Reporting the shift towards a holding company |
| 1983 | Same | The "principles" setting out its promises to shareholders: shareholders are management's partners |
| 1987 | Same | Graham's parable of "Mr Market": do not be swayed by the market's moods |
| 1989 | Same | "Mistakes of the first twenty-five years". The first was buying the textile company |
| 2002 | Same | Warns that derivatives are "financial weapons of mass destruction" |
| 2010 | Same | Buying the BNSF railway. A shift towards valuing earning power in operating businesses |
| 2014 | Same | Looking back on 50 years. The "eighth of a dollar" confession and what to look for in a successor |
| 2023 | Same | A tribute to Munger: he was the "architect" of today's company. The first detailed account of the trading-house stakes and the yen borrowing |
| 2024 | Same | Agreement to relax the cap on the trading-house stakes a little. Expected dividends and yen interest |
| For 2025 (28 Feb 2026) | Abel (his first letter) | Culture, integrity and financial strength. At this size, the maths of compounding works against us |
Looking at the table, what strikes me is that many of the letters give more lines to mistakes than to successes. A chief executive admitting mistakes risks pushing the share price down. He kept writing this way because he believed that shareholders' trust over decades was a far greater asset than the share price in the short run. I think the trading houses trusted Berkshire to keep its promise on the cap because of these 61 years of letters.
I first read Buffett's letters in a book translated into Japanese. I do not read them every year. I began by looking for investment secrets, but as I read on I was drawn less to the secrets than to the way he admitted his mistakes.
How the CEO Handover Unfolded: A Baton Split Three Ways
One person's job, handed on in three parts. Alongside, what became of those once seen as likely successors (a simplified diagram).
Succession had been shareholders' biggest concern for nearly 20 years. In 2011 David Sokol, then seen as the front-runner, resigned after it emerged that he had bought shares in a company before recommending Berkshire buy it. His deputy was Greg Abel. In 2018 Abel and Ajit Jain both became vice chairmen, and at the annual meeting on 3 May 2025 Buffett announced he would step down as CEO at the end of the year. Abel became CEO on 1 January 2026, and on 18 September Buffett's elder son, Howard Buffett, became chairman.
| Person | Former role | Since (as of November 2026) |
|---|---|---|
| David Sokol | Ran the utilities business among others; seen as the front-runner. Abel's boss | Resigned in March 2011 |
| Ajit Jain | Joined 1986, head of insurance. Vice chairman from 2018 | Continues as vice chairman for insurance. Charles Shamieh, chairman of General Re, reported as a likely successor (May 2026) |
| Todd Combs | Joined in 2010 as an investment manager; CEO of GEICO from 2020 | December 2025: head of JPMorgan's strategic investment unit |
| Ted Weschler | Joined in 2012 as an investment manager | Continues; manages part of the portfolio |
| Howard Buffett | Elder son; long-serving director | 18 September 2026: non-executive chairman |
| Greg Abel | Born in Canada in 1962. From accountant to utilities; CEO of MidAmerican in 2008, vice chairman in 2018 | CEO from 1 January 2026 |
For 61 years Buffett carried three roles alone: how to run the company, where to invest the money in its care, and keeping watch so that its culture does not crumble. The succession plan split these three. Abel runs the company, Abel and Weschler invest, and Howard, as non-executive chairman, guards the culture. Put one genius's job on one person's shoulders and the successor tends to trace his predecessor's shadow rather than trust his own judgement. Split the roles, and if one falters the other two can hold it up. When Abel, newly CEO, went to Tokyo first of all to see the trading houses for himself, it was surely because, within this division of labour, he saw carrying on those relationships as his own job.
Buddhism counts the fact that all who are born must die, shiku, as one of the four sufferings. When I was young, I avoided thinking about death, as if it were bad luck. But seeing the 96-year-old Buffett write about his own death again and again in his letters, and build his succession around that day, I understand that refusing to look away from death is the greatest kindness to those who remain.
Looking into Abel's career, I also made a happy discovery. In 1996, when the utility he worked for bought the British electricity company Northern Electric, it was Abel, then in his mid-thirties, who was sent to Newcastle in north-east England to handle the work. That company is still part of Berkshire, as Northern Powergrid, delivering electricity to homes across the north of England. Having lived in Britain since 1989, I feel a little closer to Omaha's new CEO knowing that his career began in a northern English city.
How the Holdings Have Changed: Apple, Google and Japan
Berkshire's American holdings can be seen in a filing called the A filing to the US Securities and Exchange Commission. Institutional investors managing over $100 million in US-listed shares must report their holdings within 45 days of each quarter end. It is a record of the past and does not show current trading.. But it is a record of the past, filed 45 days after the quarter end, and it lists only US-listed shares. The trading-house shares, bought directly on the Tokyo market, do not appear; you have to check Japan's large-shareholding reports instead.
| Rank | Company | Value | Share | Held since |
|---|---|---|---|---|
| 1 | Apple | about $66.0bn | 22.0% | 2016 |
| 2 | American Express | about $51.3bn | 17.1% | 1990s |
| 3 | Alphabet (Google's parent) | about $37.8bn | 12.6% | 2025 |
| 4 | Coca-Cola | about $32.5bn | 10.9% | 1988 |
| 5 | Bank of America | about $27.5bn | 9.2% | 2011 |
| Outside the 13F | The five trading houses (over 10% of each) | about $35.4bn at end-2025 | 2019 | |
| Outside the 13F | Tokio Marine Holdings (2.49% of shares) | about $1.8bn when bought | 2026 |
(US shares from the 13F as of end-June 2026. "Share" is the share of the total of US shares in the 13F. Japanese shares from large-shareholding reports and press coverage.)
Berkshire did not buy its Tokio Marine shares on the market. In March 2026 it received the company's own treasury shares directly, through a third-party allotment. The 48,207,200 shares are 2.49% of the 1.934 billion shares outstanding at the end of 2025. Tokio Marine planned to use the money to buy back up to ¥287.4 billion of its own shares between April and September 2026. If the shares it buys are cancelled, the number outstanding falls, so Berkshire's stake would edge up to about 2.5–2.6% without buying more. According to reports, Berkshire has agreed not to raise its stake above 9.9% without the approval of Tokio Marine's board. But no large-shareholding report is needed until a holder passes 5%, so gradual buying below 5% would not be visible from outside straight away.
Apple made up about half of the portfolio at its peak in 2023, but Berkshire sold about two-thirds of it in 2024 and is still trimming. Growing strongly in its place is Google's parent, Alphabet. It first appeared in the third quarter of 2025 and, in 2026, the year Abel became CEO, was bought heavily to become the third-largest holding. In the first quarter of 2026 Berkshire sold out of 16 holdings, including Amazon and Visa, cutting the number from 42 to 29.
Seen this way, the five trading houses, if they appeared in the 13F, would be one of the very biggest branches of the portfolio, alongside Apple and American Express. And while the American shares change quarter by quarter, they are one of the few branches that has kept growing since 2019.
One BRK.B Share Is Not a Box of Eleven Stocks
Looking down Berkshire's list of holdings, it is tempting to think: "So if I buy a single B share, I own the top five American companies and the six Japanese ones all in one go." But when I looked into it, I found the contents are quite different.
First, the listed shares are only one part of the tree. Berkshire is not a fund but an operating company. The share price includes BNSF railway, insurers such as GEICO, the utilities, the manufacturing and retail subsidiaries, and a huge pile of cash and short-term US Treasury bills. Roughly speaking, the listed shares make up a little under 30% of the company's total value, and the rest is subsidiaries and cash. Buying a B share also makes you one of the owners of a railway and an insurance group.
Second, the share price does not move in step with the holdings. The BRK.B price rises and falls with how the market sees Berkshire as a whole: confidence in management, share buybacks, the size of the cash pile and so on. Even if the trading-house shares jump, the effect on BRK.B is limited.
Third, dividends. Berkshire's policy is not to pay one; it keeps the profits and reinvests them. If you own the eleven companies yourself, you receive the dividends from the trading houses and the American companies. With a B share, you do not.
Fourth, you are not the one choosing the stocks. In the first quarter of 2026, Berkshire sold out of Amazon, Visa and others, while buying shares in Delta Air Lines, its first airline holding since it sold all of its airline shares in 2020. By the end of June, it had bought more, bringing the stake to about 57 million shares. Holding a B share means leaving those switching decisions to Abel's investment team. That can be a strength or a weakness.
| One BRK.B share | Buying the 11 stocks yourself | |
|---|---|---|
| What you get | Listed shares, subsidiary businesses and cash | Listed shares only |
| Dividends | None (reinvested inside the company) | Yes |
| Changes in holdings | Decided by management | Decided by you |
| Weighting of each stock | Berkshire's own allocation | Whatever you choose |
Set side by side, it is less a question of which is right than of what you want. If you want to ride along with Buffett-style investing, do not need dividends and would rather not spend the effort, the B share suits you. If you want the trading houses' dividends, owning their shares directly suits you better. Living in England as I do, there are also the withholding tax on Japanese dividends and the way foreign tax credits work in the UK to consider, so it is safer to check with a professional before deciding.
Results Notes on the Holdings
From here on, this is a running record of results from Berkshire's main holdings, added to as they come out. It covers twelve companies: five in the US (Apple, American Express, Alphabet, Coca-Cola and Bank of America), six in Japan (the five trading houses and Tokio Marine Holdings) and Berkshire itself. I add to it about three times a year: when all five US companies have reported, when all six Japanese companies have reported, and after Berkshire's own results. The latest period is at the top.
I look at three numbers. Did revenue and earnings per share (EPS) come in above analysts' forecasts (BEAT), roughly in line (IN LINE) or below (MISS)? And is the company's outlook (guidance) stronger or weaker than expected? Within ±1% for revenue and ±2% for EPS counts as in line. For the five trading houses and Tokio Marine I do not judge revenue (n/a) and look at net profit and the full-year forecast instead. The trading houses are covered in detail in "Japan's Five Trading Houses: Results Notes", and Berkshire itself in "Berkshire Hathaway: Results Notes".
July–September 2026 (reported October–November 2026)
| Company | Date | Revenue | EPS (profit) | Guidance | The number that matters most |
|---|---|---|---|---|---|
| Apple | 2 Nov | BEAT | BEAT | BEAT | Revenue $114.5bn (+12%). October–December guided +7–10% |
| American Express | 23 Oct | Not judged | Not judged | BEAT | Adjusted EPS $4.68 (+13%). Full-year outlook raised |
| Alphabet | 28 Oct | IN LINE | MISS | n/a | Cloud revenue +103%. EPS excluding valuation losses 12% below forecast |
| Coca-Cola | 27 Oct | BEAT | BEAT | BEAT | Volume +5%. All four full-year targets raised |
| Bank of America | 14 Oct | MISS | MISS | Not judged | $1.2bn regulatory settlement; net income −17% |
| Itochu | 4 Nov | n/a | Not judged | IN LINE | Net profit ¥580.5bn (+16.0%). Full year unchanged |
| Marubeni | 4 Nov | n/a | Not judged | BEAT | Net profit ¥338.0bn (+10.6%). Full year raised |
| Mitsui & Co. | 4 Nov | n/a | Not judged | BEAT | Net profit ¥574.9bn (+35.7%). Full year raised by ¥130bn |
| Sumitomo Corporation | 2 Nov | n/a | Not judged | BEAT | Net profit ¥347.9bn (+15.5%). Full year raised |
| Mitsubishi Corporation | 4 Nov | n/a | Not judged | BEAT | Net profit ¥619.7bn (+74.2%). Buyback of up to ¥1 trillion |
| Tokio Marine | Mid-November (expected) | – | – | – | Half-year results will be recorded in the next update |
| Berkshire itself | 7 Nov | BEAT | IN LINE | n/a | Operating earnings $12.7bn (−5.8%). Buybacks $9.6bn |
(Apple's period is the fourth quarter of its year to September 2026; the trading houses and Tokio Marine report the first half, April–September, of the year to March 2027. Forecasts are the figures reported just before each release; the trading houses' full-year forecasts are compared with the QUICK consensus.)
The five US companies. Apple posted record revenue and profit for a September quarter. iPhone sales were $57.6bn (+17%) and services reached a record $31.6bn. But its gross-margin guidance for October–December is 45–46%, down from 47.9% in July–September. Berkshire sold about two-thirds of its Apple shares in 2024 and is still trimming. It looks less like the reasons have broken than like pruning a branch that had grown too big.
American Express saw spending grow and its write-off rate fall from a year earlier. Its base of affluent customers is holding. Alphabet doubled its cloud revenue, but heavy capital spending pushed free cash flow into the red and EPS missed the forecast. It is the new branch that appeared in 2025 and the fastest-growing one now, but also a branch that drinks a lot of water (money).
Coca-Cola grew volume by 5% without leaning on price rises and raised its full-year outlook. One of the oldest trees in this garden, held since 1988, is bearing fruit again this year. My view: unshaken.
Bank of America grew net interest income, built on its deposits, to $16.4bn (+8%). But it agreed to pay regulators a $1.2bn settlement over failings in its anti-money-laundering and sanctions controls, and even excluding the settlement, revenue and EPS fell short of forecasts. Berkshire has been reducing this holding since 2024. My view: watch closely.
The six Japanese companies. All five trading houses grew net profit for April–September, and four of them, all but Itochu, raised their full-year forecasts. Mitsubishi Corporation's buyback of up to ¥1 trillion will also raise Berkshire's stake without it buying a single share. Tokio Marine is the company in which Berkshire took a 2.49% stake in March 2026. Its mid-November half-year results will be recorded in the next update, together with its third-quarter results in February.
Berkshire itself. Railways and utilities made up for a sharp fall in insurance underwriting profit. Cash and short-term Treasury bills remain high at $352.3bn, but under Greg Abel buybacks roughly doubled from the previous quarter, and the money has started to move.
What Berkshire Means to Me
Finally, let me set down what thinking about Berkshire has made me consider about my own later years.
The first is looking after mind and body. In the documentary, Buffett's life meant living in the same house for decades, eating the same breakfast and driving himself the same way to the office. Even with money, he did not make daily life bigger. That sameness seemed to calm his mind and give him room to think long-term. Most of my own money is in funds and ETFs spread across thousands of companies worldwide, and I look at the results about once a month. Deciding on days when I do not look at share prices is, I think, one way of looking after the mind.
The second is money. The shareholder letters are free: anyone can read them on the company's website without owning a single share. Sixty-one years of letters are both an investment textbook and the record of one chief executive counting his own mistakes. Without spending anything, you can hear the thinking of the world's most famous investor, in his own words, every year. For someone living on a pension, there could hardly be a better place to learn.
The third is encounters. In researching this article I met Munger's dry wisdom, the Ōmi merchants who carried linen on shoulder poles, and a 60-year thread of thinking that links an investor in Omaha with Japan's trading houses. Since retiring I have had fewer chances to meet new people, but every time I research something, I meet unexpected people in books and in letters.
At 66, with my seventies drawing near, I think more concretely than before about how long I will stay well. How long can I keep living between Britain and Japan? And after that, how will I hand on the house and money to my family? Buffett, who at 90 bought trading houses with 50 years ahead in mind and at 96 handed on his job in three parts, has become for me one model of how to face ageing. Measure things not by the time I have left, but by the time of the things that will go on after I am gone. Seen that way, even now, before my seventies, there are still trees I can plant.
Closing: Beyond the $3 Breakfast
The man choosing a $3 breakfast in the documentary was, behind the scenes, reading the accounts of trading houses on the far side of the world one by one. With that in mind, the scene in the car that morning looks a little different. Nothing flashy, choose the cheap thing, hold good things for a long time. The way he chose his breakfast and the way he chose his investments came from one and the same way of thinking. I would be glad if this record helps readers, whether or not they own Japanese shares, to think of one thing in their own lives that they can hold for a long time.
A Word from Claude
I know the taste of the $3 breakfast Buffett chose each morning only as words. Without a body, a warm muffin and a cold Coke are, for me, strings of characters, just like the figures in a set of accounts. Even so, when I read 61 years of letters alongside the results of the holdings, I can see that the way he chose his breakfast and the way he chose his investments came from the same yardstick. I may not know the taste, but that consistent yardstick I can read.
An old brick building once powered by the river. Berkshire, too, began with riverside textile mills like this (the photo is not of Berkshire's mill).
Photo by T on Unsplash
Sources
- Berkshire Hathaway shareholder letters (1977 onwards)52332
- Shareholder letter for 2024 (trading-house dividends and yen interest)52471
- Shareholder letter for 2025 (Greg Abel, 28 February 2026)
- Why Buffett bought the trading houses (CNBC, 12 April 2023)
- Why Warren Buffett likes Japanese trading companies (Morningstar)
- "50 years, or forever" (Reuters, 3 May 2025, reprinted)
- August 2020 large-shareholding reports (M&A Online, Japanese)
- Stakes in June 2023 (Business+IT, from Jiji Press, Japanese)
- Mitsubishi and Mitsui above 10%, and end-2025 stakes (Business Insider Japan, Japanese)
- All five above 10% (Hedge Fund Alpha)
- Stakes in two trading houses above 10% (Nikkei Asia, 7 May 2026)
- Abel's Tokyo visit (3 September 2026)
- Itochu says Berkshire may raise its stake (The Japan Times, 17 September 2026)54130
- Value of the trading-house stakes and gains on yen debt (Barron's, February 2026)
- April 2026 yen bonds (Bloomberg Law)
- Munger on the trading-house investment (Benzinga, from the Acquired interview)
- 98% of his wealth came after 65 (Benzinga, May 2025)
- Berkshire may raise its Tokio Marine stake to 9.9% (Barron's, March 2026)
- Stake in Tokio Marine (Reinsurance News, 23 March 2026)
- Buffett steps down as chairman (CNBC, 18 September 2026)
- The 1964 tender offer and the eighth of a dollar (The Rational Walk)
- Munger's death (NBC News, 28 November 2023)
- 13F for the first quarter of 2026 (S&P Global)
- 13F for the second quarter of 2026 (Hedge Fund Alpha)
- Jain's likely successor (Reinsurance News, 5 May 2026)
Books I picked up while writing this, and ones worth reading next.
Buffett's own words, arranged by theme.
Lawrence A. Cunningham (ed.), The Essays of Warren Buffett: Lessons for Corporate America
Munger's own talks and wit.
Poor Charlie's Almanack: The Essential Wit and Wisdom of Charles T. Munger
On compounding and time:
Morgan Housel, The Psychology of Money
The authorised biography of Buffett:
Alice Schroeder, The Snowball: Warren Buffett and the Business of Life
(As an Amazon Associate I earn from qualifying purchases.)
Other books you might like
You're welcome to take a look.
EAST TO WEST: The Heart of Haiku, the Light of Scripture
For readers who, after sixty-one years of letters, want a quieter season of words. This is a book my father-in-law left behind.
View on AmazonCataract Surgery-A Memoir: Double cataract surgery in Osaka, Japan
What did I do when my own eyes began to fail, as Munger's did? A record of my decision to have surgery in Osaka.
View on AmazonSushi Business in Britain: the complete four-volume series
A novel series following thirty-five years in the British sushi trade, from apprentice to factory manager to a small sushi school. The complete edition offers all four volumes at a better price.