In January 2006, in an office in Osaka, a man named Masakazu Kongō signed away a company that had been in his family since before Islam existed.
In January 2006, in an office in Osaka, a man named Masakazu Kongō signed away a company that had been in his family since before Islam existed.
Not before the modern stock market. Before Islam. Before Charlemagne, before the Vikings reached England, before anyone in Europe had written a word of the Domesday Book. Kongō Gumi, the construction firm Masakazu led as its 40th master, was founded in 578 AD. It had survived 1,428 years, forty generations of the same family, every civil war Japan ever fought, the arrival of firearms, the arrival of Americans, the firebombing of its home city, and two atomic bombs dropped on its country.
It did not survive the 1990s.
When Takamatsu Construction Group absorbed its assets that January, Kongō Gumi was carrying roughly $343 million in debt. The oldest continuously operating company in recorded history, the ultimate proof that a family enterprise can outlast every empire around it, died of something as ordinary as a bad loan book. The name survives today as a subsidiary. The family does not run it.
If you are building anything you hope your grandchildren will hold, this is the single most instructive death in business history. So let's start at the beginning, because the beginning explains everything that worked, and the ending explains the one thing that didn't.
In 578, Prince Shōtoku, the regent who would push Buddhism into the center of Japanese life, had a problem. He wanted to build Shitennō-ji, one of Japan's first great Buddhist temples, and Japan had no one who knew how. Buddhism was an import, and so was its architecture. The prince did what any good operator does when the capability doesn't exist in-house. He imported it. He invited three temple carpenters from Baekje, a kingdom on the Korean peninsula, and one of them, a craftsman named Kongō Shigemitsu, stayed, founded a workshop, and gave it his name.
Shitennō-ji was completed in 593. Then it burned down. Over the following fourteen centuries it burned down again, was struck by lightning, was flattened by a typhoon in 1934, and was destroyed in the air raids of 1945. It was rebuilt every time, and nearly every time, the family doing the rebuilding was the same family that built it first.
Sit with that arrangement for a moment, because it is stranger and smarter than it looks. Kongō Gumi's anchor client was a temple. A temple does not go out of business. A temple does not switch vendors to save four percent. A temple burns, and when it burns, it must be rebuilt exactly, which means the only qualified contractor is the one who holds the original knowledge. The Kongō family had effectively attached itself to an institution with a longer time horizon than any government, and made itself the sole custodian of that institution's physical body. Revenue for the workshop was not a market. It was a relationship, renewed by fire and faith, for 1,400 years.
Longevity that long is never luck. Strip the romance away and Kongō Gumi ran on four mechanisms, each one a deliberate choice that most family businesses, then and now, refuse to make.
First: the most capable child led, not the eldest. Primogeniture is the default setting of family succession almost everywhere, and it is a coin flip on competence run once per generation. The Kongō family didn't flip the coin. Leadership of the firm passed to whichever son showed the health, responsibility, and talent for the role, and the family was willing to look past birth order to find him. When the situation demanded it, they looked past sons entirely: after the 37th master, Haruichi Kongō, died during the desperation of the early 1930s, his widow Yoshie took over as the 38th head of the firm. A woman running a major construction house in pre-war Japan was close to unthinkable. The family thought it anyway, because the alternative was handing fourteen centuries of accumulated trust to someone unfit to hold it.
Second: when the bloodline ran thin, they adopted. Japan has an old and pragmatic institution called mukoyōshi, in which a family legally adopts an adult son-in-law, who takes the family name and becomes a true heir. If a generation produced no capable son, the Kongōs would recruit one, marry him to a daughter, and graft him into the tree. The family treated the bloodline as a vehicle for the enterprise, not the other way around. What was sacred was the name, the craft, and the client. Who carried them was a staffing decision.
Third: they stayed inside their niche with almost fanatical discipline. Temple carpentry is a deep art. Joints cut to interlock without nails, cypress selected decades ahead of use, proportions handed down master to apprentice. The 32nd master, Yoshisada Kongō, wrote the family's creed down around 1801, sixteen precepts covering how to live and how to work: drink in moderation, dress within your station, treat others with respect, and above all, concentrate on the core business. Not a mission statement. House law.
Fourth: when survival demanded flexibility, they flexed the product, never the balance sheet. In the Meiji era, the government promoted Shinto and Buddhism fell violently out of favor. Temples lost their patronage and Kongō Gumi lost its market. The firm adapted by building commercial buildings with its temple-grade skills. During the Second World War, with no temples being commissioned and the country consumed by death, the workshop survived by making wooden coffins. It is hard to imagine a starker act of humility: the artisans of Shitennō-ji, building boxes. They did it, kept the crews together, kept the knowledge alive, and went back to temples when the world allowed. Fourteen centuries of pivots, and notice what never appears in the record: a bet-the-company loan.
Until one does.
In the 1980s, Japan lost its collective mind about land. Property in Tokyo was, at the peak of the bubble, notionally worth more than all the real estate in the United States combined. Banks begged companies to borrow against land, buy more land, and borrow against that. Nearly everyone did. It felt less like speculation than like citizenship.
Kongō Gumi, by then a limited company run by its 39th and 40th masters, did what the whole country was doing. It borrowed heavily to invest in real estate during the bubble. For the first time in its history, the firm's fate was tied not to the temple, not to the craft, but to the price of an asset it did not build and could not control.
In 1992 the bubble burst, and the value of Kongō Gumi's real estate collateral collapsed with it. That alone might have been survivable. What made it fatal was that the firm's ancient revenue engine was quietly failing at the same time. Postwar Japan was secularizing. Communities that had once funded temple construction through donations were aging and shrinking, and by the 2000s contributions to temples had fallen sharply while the firm's revenues slid year after year. The anchor client that had held for 1,400 years was dragging. The debt was not.
By 2006 the borrowings had swollen to about ¥40 billion, roughly $343 million, against a business doing a fraction of that in annual revenue. There was no 41st-generation succession plan clever enough to fix that arithmetic. In January 2006 the company was liquidated and its operations absorbed by Takamatsu Construction Group, which, to its credit, kept the name and the temple craftsmen employed. Kongō Gumi still builds temples today. But the world's longest experiment in family enterprise ended the moment the family lost the equity.
Here is the part worth underlining. The company did not die because the craft became obsolete. It did not die from war, though it survived dozens. It did not die from an incompetent heir, the thing family businesses fear most. Every mechanism the family had spent forty generations perfecting, merit-based succession, adoption, niche mastery, the anchor client, was still functioning in 2006. All of it was overridden by a single variable the mechanisms had never been asked to govern: how much the firm owed.
Forty minutes from the Sea of Japan coast, in the hot-spring town of Awazu, there is a counterfactual you can sleep in. Hōshi Ryokan was founded in 718 AD, an inn built around a spring a Buddhist monk was said to have discovered, and it has been run by the same family for 46 generations. It is still open. You can book a room this week.
Hōshi has faced its own crises, including the modern one that stalks every old family firm, a generation reluctant to take the yoke. But through thirteen centuries it has held to the same quiet financial posture that Kongō Gumi held for its first thirty-nine generations: own the asset, keep the operation small enough to control, grow at the speed of retained earnings, and treat debt as a poison you handle rarely and in small doses. The inn never tried to become a chain. It never mortgaged the spring to buy land in Tokyo. It is smaller than what it could have been, and it exists, which is the entire point.
Put the two houses side by side and the lesson stops being sentimental and becomes almost mechanical. Two families, both anchored to religious institutions, both practicing capable-heir succession, both masters of a narrow craft. One added a decade of speculative borrowing to the formula. That is the only structural difference between the company that died and the inn that didn't.
The standard corporate finance view says debt is neutral, a lever that amplifies outcomes in both directions, and for a corporation with diversified shareholders that view is roughly right. A shareholder can lose one bet among fifty. A family cannot. A family enterprise is a fifty-generation position with no diversification across time, and for that kind of position, debt changes its nature entirely. Equity can wait out a bad decade. Debt cannot wait out anything. It converts every downturn from a period you endure into a countdown you might not.
Kongō Gumi endured the Ōnin War, the collapse of the shogunate, state suppression of its only client's religion, and the deliberate destruction of its city from the air. Each of those crises took revenue to zero or near it, sometimes for years. The firm survived because zero revenue against zero debt is hibernation. Zero revenue against ¥40 billion of debt is death, and death on a schedule set by the lender, not the family.
This is why the old Japanese house codes, written by merchant families who had watched neighbors vanish, so often read like the notes of a paranoid accountant. Hold cash. Avoid borrowing. Stay in the business you understand. Modern readers see conservatism, a timid refusal to grow. That is a misreading. These families were not avoiding growth. They were avoiding clocks. Across a long enough horizon, some decade will be catastrophic, and no one gets to choose which one. The only preparation that works for an unchoosable catastrophe is a balance sheet that can sit still indefinitely. Debt aversion is not a temperament. It is the survival trait, selected for by fourteen centuries of dead competitors.
And the bubble is always convincing. That deserves saying plainly, because it is the mechanism by which disciplined families finally break. Kongō Gumi's leaders in the 1980s were not fools or gamblers. They were operating in an environment where every bank, every peer, and every price signal said that land only rises and that refusing cheap money was the reckless choice. The fortieth generation faced a test none of the previous thirty-nine had faced in that exact costume, and the costume worked. Prosperity is the predator that eats old families. Famine never got them.
So here is the choice this story leaves on your desk, and it is a real decision you can make this month, not a mood.
Write your family's debt covenant, and write it now, while nothing is on fire and no opportunity is whispering. One page. What may be borrowed against, and what may never be, the operating business, the family home, the core holdings. A hard ceiling on total debt as a fraction of unencumbered assets, low enough that a 50 percent collapse in asset values leaves the family solvent and bored rather than negotiating with a workout group. A rule that any exception requires the written consent of every adult member, recorded, so that no single confident generation can quietly bet the accumulated work of the previous ones. Then read it aloud once a year, the way the Kongōs read their precepts, so your children inherit the rule along with the reasons.
It will feel excessive. For nine decades out of ten, it will be excessive, and your covenant will do nothing but sit in a drawer costing you upside. Then the tenth decade will arrive wearing whatever costume your era dresses it in, and your family will still be at the table when it passes, the way Hōshi Ryokan is still by its spring, taking guests, 1,300 years on.
Forty generations of discipline bought Kongō Gumi fourteen centuries. One decade of debt spent it. Decide which line of that ledger your family is going to write next.