If the generational wealth curse were real science, the institutions selling you protection from it would at least quote the same numbers. They do not. Read three mainstream guides to building family wealth side by...
If the generational wealth curse were real science, the institutions selling you protection from it would at least quote the same numbers. They do not. Read three mainstream guides to building family wealth side by side, as almost nobody does, and you find three incompatible accounts of the disease, each stated with total confidence, none carrying a citation you can follow to an actual study. This brief lays the exhibits on the table.
| Source | The claim | What it implies about gen 3 | Primary source cited | |---|---|---|---| | Regions Bank | 70% of generational wealth is lost by the second generation, 90% by the third | 10% of family wealth survives to gen 3 | None | | White Coat Investor | Only about 30% of wealth makes it to the third generation, about 10% to the fourth | 30% survives to gen 3 | None | | US Bank | Citing Harvard Business Review: most wealthy families remain wealthy when they plan and communicate | Majority survival is achievable, the curse is conditional | Secondary (HBR) |
Sit with the spread. Regions says one family in ten gets money to the grandchildren. White Coat Investor says three in ten. US Bank says most families who plan keep their wealth, which reads as a survival rate above one in two. These are not rounding differences. Between the gloomiest and the sunniest account sits a gap wide enough to hold your entire financial future, and each version is presented as settled fact in the same calm advisory voice.
Two of the claims cannot both be true. Arguably none of the three can coexist. And the strangest part is that the first two are plausibly the same underlying number wearing different clothes: "70% lost by generation two" and "30% reaches generation three" both orbit a 30% figure, drifted one generation apart somewhere in the retelling. The statistic is not merely unsourced. It mutates in transit, and the institutions repeating it do not notice, because none of them ever handled the original.
In 2022 James Grubman, a psychologist who has spent his career inside the family wealth industry, did what the marketing departments never did. He followed the citations. His paper, There Is No 70% Rule, published in the International Family Offices Journal, traces the entire curse family back to a single ancestor: John Ward's 1987 study of roughly 200 manufacturing companies in Illinois.
Note every load-bearing word in that sentence. Manufacturing companies, not families. Illinois, not the world. 1987, examining firms that existed in 1924, meaning survivors of the Great Depression and the Second World War. Ward found that about 30% of these firms made it into the second generation of family ownership. That 30% survival figure was inverted into a 70% "failure" figure, quietly reworded from companies to wealth, and set loose. Williams and Preisser's 2003 book Preparing Heirs then dressed the number in the authority of a claimed 3,250-family research effort. Grubman examined that too: the actual data was roughly 177 complete questionnaires from self-selected seminar attendees, asked for opinions, with no measurement of anyone's wealth outcomes at all.
There are also definitional landmines inside the ancestor study. Ward counted a company as failed if it was sold, merged, or taken public. A family that built a firm for 40 years and exited at a strong price, converting a concentrated asset into diversified capital, scored as a casualty. By that coding, the most successful outcome available to most business families is recorded as proof of the curse.
So the pipeline runs like this: a regional study of Depression-era manufacturers, with sale counted as death, gets inverted, generalized from firms to wealth, laundered through a book whose sample was two orders of magnitude smaller than advertised, and then photocopied for 20 years by institutions that each introduce their own copying errors. Regions inherited one photocopy. White Coat Investor inherited a different one. US Bank skipped the curse and cited a rebuttal instead. Nobody checked, because the number was doing its real job either way.
Follow the incentive, because it explains the entire pattern. Every one of these guides is content marketing for a planning relationship. Regions publishes its curse statistics in an article about how the bank can help you build generational wealth. US Bank's article sits inside its wealth management funnel. The curse is the cold open; the wealth plan is the product. A 90% failure rate is a superb cold open. It is specific enough to sound researched, terrifying enough to create urgency, and flattering enough to the reader, who naturally assumes they will plan their way into the surviving 10%.
This does not require anyone to be lying. It requires something more ordinary: nobody in the chain has a reason to check a number that converts. The compliance department verifies the APR disclosures, not the folklore in the blog post. A statistic that sells meetings gets repeated; a footnote that complicates it gets cut for length. Twenty years of that selection pressure produces exactly what we observe, a family of confident, contradictory, unsourced numbers, all descended from one Illinois study that measured something else.
The tell, once you see it, is everywhere in financial content: a precise percentage, a dramatic claim, and no named study. Real research arrives with inconvenient furniture, sample sizes, definitions, regions, caveats. Marketing arrives clean.
It is worth naming how cheap the fix would have been. Grubman's audit required no laboratory and no grant, only a working library card and the patience to read what the citations actually said. Any of these institutions could have done it in a week. That none of them did, across two decades and thousands of publications, tells you the number was never load-bearing in the argument. It was decoration on a sales letter, and decoration does not get fact-checked.
Here is where this brief refuses to become a mirror image of the problem, because the wrong conclusion is available and tempting: the curse is fake, therefore relax, wealth transfers itself. That is also false, and Grubman is explicit about it. Estates do get fought over. Heirs do arrive unprepared. Businesses do die in the second generation, often enough that every culture coined a proverb about it before any bank wrote a listicle. What the audit kills is the precision and the terror, not the underlying risks.
And the practical guidance that rode in on the fake number is, awkwardly, mostly sound. The same survey work behind Preparing Heirs, weak as it was for measuring outcomes, asked practitioners and families why transfers go wrong, and the answers ranked communication breakdown and unprepared heirs far above tax and legal failure. That ranking matches what US Bank's HBR-derived framing says from the optimistic side: families that talk, prepare heirs, and govern themselves tend to keep what they built. The behaviors are validated by practice even though the statistics marketed alongside them are folklore. Strip the terror away and the to-do list barely changes; what changes is why you do it. Calmly, as maintenance of something likely to survive, rather than frantically, as a lottery ticket against a 90% death rate. Fear-based planning has its own costs. It makes families secretive, rushes them into structures they do not understand, and hands negotiating power to whoever holds the scary slide.
The reading protocol, stated once, plainly: when a percentage appears in wealth content, look for the footnote. If there is no footnote, you are reading advertising. If there is a footnote, read what it actually studied, because there is a fair chance it counted Illinois factories, called a profitable sale a failure, and never measured a family's wealth at all.
Keep the behaviors, discard the terror. Concretely: continue, or start, the three practices every version of this literature agrees on, one family money conversation on the calendar, one heir given real preparation rather than a sealed envelope, one written plan reviewed annually. And separately, adopt the footnote rule for every statistic in every financial pitch you receive from today: no named source, no belief. Say it back to the advisor if you have to, and watch what happens to the slide deck.
This piece did its job if the next "90% of families lose it all" slide makes you ask one question out loud, "measured by whom, in what study," and you keep your planning appointment anyway, for the right reasons this time.