Wealth does not usually end in scandal. It ends in silence, and the silence is enormous.
Wealth does not usually end in scandal. It ends in silence, and the silence is enormous.
Here is an exercise almost nobody in the wealth industry runs, because the answer is bad for business as usual. Take the roughly 1,000 wealthiest families in the world at the start of the 1900s. Assume nothing heroic: ordinary market returns on diversified capital, ordinary rates of marriage and children, ordinary taxes, ordinary spending. Compound for twelve decades. How many billionaire families should exist today from that cohort alone?
Legacy adviser Johann Kurtz did the arithmetic, in his book Leaving a Legacy: Inheritance for Thousand-Year Families and in the interviews he has given around it. His answer: something on the order of 16,000 billionaire families should be walking around right now, descended from that founding cohort, without any of them ever doing anything remarkable again. Money at normal returns doubles faster than families do. The math wants dynasties.
Now count the dynasties. Roughly 700 exist. Not 700 missing, 700 present. Somewhere between the arithmetic and the reality, about 15,300 families vanished from the top of the distribution. Their money did not underperform. It did not get out-competed by index funds. It dissolved.
Treat Kurtz's numbers as a model, not a census, because that is what they are. You can quarrel with his return assumptions, his fertility assumptions, his definition of a "family." Move every input against him and you still end up with thousands of missing families, because the gap is not close. The point survives any reasonable recalculation: the default outcome of a great fortune is disappearance, and disappearance on a scale the industry almost never says out loud.
The interesting question is not whether the families vanished. It is how. And for that, history left us a perfect controlled experiment: two American fortunes, born a generation apart, one of which followed the default and one of which refused it.
Cornelius Vanderbilt died in January 1877 as the richest man the United States had ever produced. The Commodore's estate came to roughly $100 million, a sum contemporaries liked to point out exceeded what the US Treasury held at the time. He had built it twice, first in steamships and then in railroads, and he left almost all of it to one son, William Henry, precisely because he understood that division kills fortunes.
For one generation, the plan worked. William Henry roughly doubled the pile to around $200 million by his death in 1885. At that moment the Vanderbilts were not merely the richest family in America. They were arguably the richest family in the world, sitting on the fastest-compounding asset base of the industrial age.
Then the default took over. The third and fourth generations built the Breakers in Newport and Biltmore in North Carolina, ran competing palaces up Fifth Avenue, bought yachts and horses and English titles for their daughters, and treated the principal as a renewable resource. No family office. No common governance. No mechanism for the branches to invest together, decide together, or restrain each other. Each heir received capital and total freedom, which is to say each heir received capital and no protection from himself. Estate divided among heirs, heirs multiplied, spending scaled with status rather than with income, and the railroad money met the twentieth century with no one steering.
Arthur T. Vanderbilt II, a descendant, wrote the family's obituary himself in his 1989 book Fortune's Children, and one scene in it has become the most quoted funeral in the history of American money. In 1973, 120 members of the family gathered for a reunion at Vanderbilt University. As Arthur T. Vanderbilt II records, among those 120 descendants of the richest man in America, "not one of them was even a millionaire." Ninety-six years, start to finish. Within three decades of William Henry's death, no Vanderbilt ranked among the richest people in the country. The Fifth Avenue mansions were rubble before the Second World War ended.
The coda is Anderson Cooper, great-great-great-grandson of the Commodore and son of Gloria Vanderbilt, telling interviewers flatly: "there's no trust fund." He said it without bitterness, and that is the detail worth noticing. By the sixth generation the fortune was not even a grievance. It was trivia. That is what disappearance actually looks like: not a crash, no villain, no single catastrophic decision. Just arithmetic, unopposed, for four generations.
Be precise about what killed it, because the popular version blames champagne. Spending was a symptom. The disease was structural: a fortune held as disconnected individual inheritances, with no shared vehicle, no shared decisions, and no shared identity to make restraint rational. Every branch optimized locally. The system did what unmanaged systems do.
Run the numbers on the palaces and the point sharpens. The Breakers cost around $7 million in the 1890s, Biltmore consumed a comparable fortune, and the Fifth Avenue houses cost millions each to build and hundreds of thousands a year to staff and heat. Set against a $200 million base, no single project was fatal. That is precisely the trap. Each branch could truthfully say its own spending was affordable, because each branch was measuring against its own slice, in its own ledger, with no one accountable for the aggregate. A family office would have seen the aggregate. Nobody saw the aggregate. Meanwhile the underlying asset, railroads, entered a half-century of decline that a coordinated family might have diversified away from and forty separate heirs could not. Fragmented ownership does not just spend badly. It also cannot reposition, because repositioning requires someone with the standing to move everyone at once.
Now run the experiment's other arm. John D. Rockefeller's fortune was born a generation after the Commodore's, peaked far larger in relative terms, and faced everything the Vanderbilt money faced: heirs, taxes, depressions, wars, and the corrosive glamour of being famous for being rich. The outcome could not be more different. The family is now in its seventh generation, more than 170 living descendants, still wealthy, still meeting, still recognizably a family in the operational sense of the word.
The difference was not better children. It was earlier structure. Rockefeller established what became the family office in 1882, while he was still building Standard Oil, decades before he had grandchildren to worry about. That office, which the family came to call Room 5600 after its address on the 56th floor of 30 Rockefeller Plaza, became the fortune's central nervous system: one professional staff managing investments, philanthropy, tax, and record-keeping for every branch at once. The Rockefeller Archive Center's records of the family office show what this looked like in practice across a century: correspondence, accounts, trust administration, generation after generation flowing through one coordinated institution rather than through forty separate lawyers.
Around that spine the family added soft structure with the same seriousness. Twice-yearly family meetings that descendants actually attend. Cousins introduced to the enterprise and to each other young. Trusts designed at Room 5600 in the 1930s that still discipline distributions today. Philanthropy run as a shared family project, which gave every generation a reason to sit in the same room that had nothing to do with dividing anything.
Notice what the Rockefeller design concedes. It concedes that heirs will be ordinary. It assumes no future genius, no second John D., no generation of exceptional restraint. The structure carries the mediocre generations, and every family gets mediocre generations. The Vanderbilts bet on individuals and got individuals. The Rockefellers bet on institutions and got a dynasty.
One fortune, no structure, gone in three generations. One fortune, structured from year one, intact in its seventh. Same country, same century, same tax code, same temptations. The variable is not luck and it is not virtue. It is engineering.
Pull the camera back to Kurtz's 15,300 missing families and the pattern generalizes. The families that vanished did not fail at investing. Markets were doing the compounding for them; that was the whole premise of the arithmetic. They failed at the family layer. Division without coordination. Inheritance without preparation. Wealth without a shared answer to the question "what is this for?" The money performed. The system around the money did not exist.
Name the mechanisms and you can see why no amount of investment skill fixes them. Division: heirs multiply geometrically, so a fortune merely holding its value shrinks per person every generation. Preparation: heirs who first encounter the wealth at probate make their worst decisions in their first two years, exactly when the decisions are largest. Purpose: money with no agreed job defaults to the job each holder invents for it, and status spending is the invention that requires no meeting. Coordination: without a shared vehicle, the family cannot act on any opportunity or threat bigger than one branch. Every one of the 15,300 disappearances is some blend of those four, and every one of the four is a governance problem, not a market problem.
This inverts the assumption most wealth holders carry, which is that persistence is the passive outcome and loss is the event. The Kurtz gap says the opposite. Loss is the passive outcome. It requires no mistakes, only absence of structure, and it claims roughly 95 percent of the cohort. Persistence is the event. It happens only where somebody built machinery for it, and the machinery is unglamorous: an office, a meeting, a trust, a written agreement, a habit of deciding together.
The encouraging part hides inside the discouraging math. Nothing the Rockefellers built required genius or even originality. A family office in 1882 was just a room with accountants in it. Family meetings are chairs and an agenda. The barrier to entry for endurance is not intelligence or luck. It is the willingness to build boring institutions before they feel necessary, which is to say while the founder is alive and the money still feels like one thing.
Scale the machinery to the fortune; the principle does not care about the size. A family with $2 million and a functioning annual meeting, a shared investment policy, and heirs who have practiced deciding together is running the Rockefeller design. A family with $200 million and none of that is running the Vanderbilt design, and the 1973 reunion is already on its calendar, RSVP pending.
Name the one structure your family will build this year. Not someday, not "when the estate plan is updated." This year. Pick from the short list that separated the 700 from the 15,300: a standing family meeting with dates already booked, a written family agreement on what the wealth is for, a single coordinated point of financial administration however modest, or a trust architecture your heirs have actually been walked through while you can still answer questions. Choose one, put a completion date on it, and tell the family which it is.
This piece did its job if you can now say, in one sentence, which structure you are building in the next twelve months, and the sentence has a date in it.