The architectural failure behind the third-generation curse

Posted on: 12 May 2026

There is a statistic that has circulated in wealth management circles for decades and is repeated at conferences as though it were a law of nature: seven family fortunes out of ten dissolve by the third generation, ninety per cent by the fourth. The most familiar formulation is "shirtsleeves to shirtsleeves in three generations", an Anglo-Saxon proverb whose equivalent exists almost word for word in Italian, Chinese, and Japanese folk wisdom. The cross-cultural recurrence is significant, because it tells us the phenomenon is structural rather than locally contingent.

The mainstream explanation is almost always the same: the children lack hunger, the grandchildren lack culture, the family failed to transmit its values, and so the wealth dissolves because those who inherit it have not earned it. This is a moral reading, and for that reason it is reassuring. It preserves the comforting idea that better education of heirs is enough to break the curse, and by extension it preserves the agency of the founder, who can still imagine themselves teaching, transmitting, shaping the next generation through force of personal example.

I have watched this interpretive scheme applied countless times in conversations with entrepreneurs who built everything themselves and worried about the fate of what they had built. The logic was always identical: if only the children understood, if only the grandchildren had the same intensity, if only we could transmit the meaning of the work. The structural response to this anxiety was invariably educational: financial literacy courses for heirs, formative travel, apprenticeship periods in the family business, ethical letters attached to the will. All internal scaffolding, all anchored to the idea that the problem lies in the quality of the people.

The empirical evidence accumulated over the last forty years suggests something quite different. The fortunes that survive into the fourth generation do not survive because the heirs are better; they survive because someone built an external architecture that prevents dispersion even when the heirs are unremarkable. The fortunes that fail do not fail because the heirs are worse; they fail because no one built that architecture, and when the founding patriarch leaves the scene there is no structure to prevent the natural fragmentation of the system. The difference is not one of virtue but of planning, and this changes everything.

Elinor Ostrom, who won the Nobel Memorial Prize in Economics in 2009, dedicated her academic life to documenting an apparently paradoxical phenomenon. Common-pool resources, according to classical theory, ought to be destroyed by individually rational behaviour. And yet there exist communities that have managed grazing land, forests, irrigation systems, and fisheries for centuries without dispersion. Ostrom went out to see these communities in the field, from the Swiss Alps to the Philippines to the lobster fisheries of Maine, and she discovered that they functioned not because the participants were more virtuous than average, but because they had developed governance architectures with recurring and precise characteristics: clear boundaries defining who may access the resource, rules adapted to local context, mechanisms of reciprocal monitoring, graduated sanctions for those who violate the rules, and low-cost methods of conflict resolution. The communities that applied these principles survived for generations. Those that did not dissolved within measurable time horizons. It was not a question of collective morality. It was a question of architecture.

The multigenerational family fortune is structurally a common-pool resource. Once the founder leaves the stage, every heir has an individually rational incentive to extract value for themselves, even at the cost of eroding the collective base. Without architecture, the system behaves like any ungoverned commons: it is depleted, slowly or quickly depending on the pressure, until exhaustion. With Ostrom's principles seriously applied, it can endure for generations. The component that the wealth management literature rarely places at the centre is implacability. The architectures that work are not the gentle ones, not those that leave room for flexibility, not those that adapt to the preferences of the heirs of the moment. They are the ones that bind asymmetrically, that impose non-renegotiable separations, that produce calculated friction precisely where individual incentive would push towards dispersion.

Nassim Taleb has articulated the same principle on the individual scale through the language of skin in the game and protective asymmetries. His observation is that sustainable systems do not rest on the good faith of participants; they rest on the structural alignment between risk and reward. When this alignment is missing, the system is fragile regardless of the quality of the people who inhabit it. When it is present, the system holds even when the people are mediocre. The serious family office applies this antifragility logic to the multigenerational fortune. It is not the adviser who explains to the children how to invest; it becomes the architecture that prevents decisional fragmentation even when the children decide badly. The separation between custody and discretionary management, the imposition of formal governance processes, the consolidated monitoring that reveals patrimonial erosion before it becomes structural, the collective negotiation of costs that neutralises the asymmetric informational power of banks, the multigenerational succession planning, these are all elements of an external architecture that works because it binds the hands of the decision-maker precisely where the decision-maker, left free, would produce harm.

It must be said, however, that the term "family office" today covers two profoundly different professional postures, even when the business card is identical. On one side there is the family office as technical extension of the client, a structure that competently executes what the client decides, optimises flows, consolidates reporting, negotiates with banking counterparts, and maintains a long-term fiduciary relationship that never interferes with the internal relational dynamics of the family. On the other side there is the family office as architect of the system, a structure that imposes collective discipline even when the individual client would prefer not to have it, says no when saying yes would be easier, and defends governance against pressures from individual family branches that would bend it to their interests of the moment.

The two figures use the same vocabulary. Both speak of wealth preservation, of generational transition, of family governance. But they produce structurally different outcomes over the long term. The first preserves the commercial relationship. The second preserves the wealth. The two are not always congruent, and when they diverge the family office must choose which side to stand on. The truth that the category rarely admits aloud is that the gravitational pressure points entirely towards the first type. A family office that says too many noes risks losing the client. A family office that imposes constraints the founder does not understand is perceived as an obstacle rather than a resource. A family office that defends collective discipline against a powerful heir wishing to extract liquidity finds itself exposed, because that heir can remove it from the relationship and find someone more accommodating. The structural forces of the market reward compliant execution and penalise implacable architecture, at least over the short term.

Only over the long term, when the patrimony dissolves or survives, does the difference become visible. By then it is too late to tell which was which.

This temporal asymmetry is the real problem of the category. The family office sells a service whose effects are measured over horizons of thirty or forty years, but is chosen, evaluated, and paid over horizons of three or four. The metric available to the client in the short term is the quality of the relationship, the punctuality of the reporting, the courtesy of the interlocutors. The metric that actually matters, whether the architecture will hold when the founder is no longer there, is not observable until it is too late to change course. The lucid founder, the one seeking the right family office for the next fifty years, cannot rely on immediate metrics. They must look for indirect signals of structural posture: how the family office responds when the client wishes to do something technically legitimate but strategically questionable, how willing it is to say no, how it survives the pressures of individual family branches in subsequent generations.

There is a subtlety worth making explicit. The legacy of the founder is not the financial wealth itself, it is the architecture. A fortune without architecture is merely an episodic event, destined to dissipate according to predictable dynamics. A well-built architecture, on the other hand, can survive the initial fortune and reconstitute it, because it contains the mechanisms of collective discipline that allow the family to accumulate again. The great European families that have crossed centuries are not families with extraordinary heirs in every generation; they are families with extraordinary architectures that produce acceptable outcomes even with ordinary heirs. This reverses the standard way of thinking about wealth management. The added value of the family office is not the quality of investment decisions in any given year; it is the robustness of the architecture across time. It follows that the choice of asset manager is less important than the choice of governance structure. Portfolio return is less important than the formal separation of powers between custody, management, control, supervision, and succession.

One operational consequence of this reading is that entrepreneurs approaching generational transition who devote exclusive attention to the financial education of their heirs are making a category mistake. Individual training is useful but marginal investment. Governance architecture is the structural investment that determines whether the wealth will survive or dissolve. And the proportion of attention and resources dedicated to the two levels, in the families I have observed in the field, is almost always inverted relative to what would produce the highest probability of intergenerational survival.

There is an explicable psychological resistance behind this inversion. The founder prefers to believe that the problem is educational because the educational solution preserves their agency. They think: I can still transmit something, I can still teach. The architectural solution is structurally more uncomfortable, because it requires accepting that one's own success is not replicable through personal virtue alone, and that the heirs will need external constraints precisely because they will not be able to recreate the irreproducible combination of circumstances that allowed the founder to prosper without constraints. To recognise this is to recognise the limit of one's own exceptionality, and few founders arrive there spontaneously.

The final observation is that the request for a family office is not a request for a financial service; it is a request for an architecture. The clients who understand this distinction are those who extract real value from the relationship. Those who seek in a family office a better adviser are buying the wrong product for the wrong reasons, and they will typically be the ones who join the seventy per cent statistic.

The three-generation curse is not a law of nature, then, but a predictable and documented architectural failure, produced by decisions taken or not taken at the moment when the founder is still fully capable of building the structure that will outlive them. That moment is also the only one in which it can be seriously built. Afterwards, the fragmentation of interests among heirs makes it structurally impossible to agree on rules that should have applied to everyone.

All of this leads to a question that the family office category, I think, should ask itself more often. How many of its clients have built an architecture that will survive, and how many are simply having their time well managed while awaiting dispersion? The answer to this question says much, not so much about the clients, but about the professional posture that the family office has chosen to practise. Because a fortune that dissolves after thirty years of continuous relationship with a technically impeccable family office is still a fortune that has dissolved. The continuity of the commercial relationship and the survival of the wealth are two distinct metrics of success, and the profession is slowly realising that conflating them is no longer enough.

The families that endure gravitate, over time, towards the second type of family office, the one that has chosen to be architect rather than executor. It is a slow movement, often silent, almost always invisible from the outside. But it is the most reliable filter that exists for distinguishing those who will survive the third generation from those who will end up in the statistic.