Part 2 of 4: Spotting the fragile family — power, knowledge, and the guardrails we never built.

One decision. By one person. That is all it took.

I sat in the first meeting with another creditor discussing the restructuring terms, wondering how I got there. My palms were sweaty, and my thoughts were going in circles. I was 25 years old and honestly should not have been sitting there. Yet, the decision by my father to support Teak Holz International caused our predicament. And it did not stop there; many more followed. One of these was the decision that put me in the room, the decision to let me lead through this crisis. I did not know back then that this was the beginning of a long road ahead of me.

After Part 1 ( Signs of a Fragile Family Balance Sheet) of this four-part series on fragile families, it is time for Part 2, where I will show you how to spot structural and governance fragility in any family.

Centralised decision-making

Often, one decision is all it takes to put your family on a road of adversity. These decisions happen particularly often when decision-making is centralised. We all have this image of a strong leader in our head. My father glorified strong leaders, and thus he tried to be one. Many of us think like my father. Yet, most of us simply aren’t strong leaders. In our case, the decision-making was centralised. One person calling all the shots. Even after I took over the crisis management, my decision-making power was only on an operational level. My father still made the important decisions.

There is a paradox worth pointing out, however. In adverse times, the families that have a competent leader who makes decisions tend to excel. On first look, this seems like a counter-argument, yet it is not. A strong leader and centralized decision-making are not the same. Decentralised decision-making is the magic word. This gives people within any organization the power and responsibility to react. So instead of having only one leader who is allowed to make decisions, there are many who can decide freely in their area of responsibility. This is a combination of a strong leader and a democratic decision-making process. This is a truly antifragile leadership structure.

Overly democratic structures are not much better than one strong leader. The difference there is that the demise of the family often becomes gradual. Take the Pritzker Family, where 11 decision-makers as a group slowly ground down the enterprise. In the end, they decided to separate, incurring huge expenses and tax bills. While a single leader will blow things up in a more spectacular fashion. Take Maurizio Gucci as an example. His decision to overhaul production along with his lavish spending and external debt led to a sensational failure of Gucci and his personal finances.

Family leaders need guardrails. A free pass to do whatever they want poses huge risks. Guardrails can be a family council, a shareholder council, an advisory or supervisory board. There are many ways to put them in place. The absence of them is a clear sign of fragility. But the over-presence of them is also one. In our case, it was the former; in other cases it is the latter.

A 2023 systems engineering study tested command-and-control architectures — ranging from fully centralised to fully decentralised — across a spectrum of operating conditions. The finding: a centralised structure achieves the highest performance when conditions are favourable, but that performance degrades quickly the moment conditions worsen, undermining its reliability when reliability matters most. A decentralised structure achieves a lower peak, but a far more stable one — it holds up under volatility instead of buckling. The same researchers also tested hybrids, and found that centralising only the functions that genuinely need system-wide coordination, while pushing everything else down to whoever has the most local information, outperformed both pure models. Centralisation isn’t wrong. It’s brittle. It looks like strength for exactly as long as nothing tests it. This is exactly the combination of a strong leader and a decentralised decision-making structure in a family.

The Pentagon found this out the expensive way. In 2002, the U.S. military ran a $250 million war game to test its own doctrine against a future adversary. The opposing force was commanded by a retired Marine general, Paul Van Riper, who refused to fight on the terms the exercise had been built around. Instead of routing orders through the same high-tech, centralised command network the American forces were relying on, he ran his side the old way — motorcycle messengers, light signals, no radio traffic to intercept. He also relied on a decentralised command structure. Battalion commanders had instructions to engage the enemy without needing to reaffirm with central command. It shouldn’t have worked against a technologically superior opponent. It did. Within the first days, his forces had sunk sixteen American warships, a defeat so severe the exercise had to be stopped. What happened next is interesting. The Pentagon didn’t rebuild its command structure around what it had just learned. It reset the simulation, restored the sunken fleet, restricted what Van Riper’s side was allowed to do, and ran it again until the outcome matched what everyone had assumed going in. This can also be seen in families when the family, instead of improving and learning from an incident, prefers to keep things the way they have always worked. More on this in Part 3 of the series.

Lack of redundancy

The next signal for a fragile family is a lack of redundancy. In our case, we had maxed out all avenues with the investment. There was no more room to manoeuvre. My father had gone all-in on one investment. An urge I see over and over again in founder-led organisations. We read these heroic stories of the founder going all-in and making it. As the story goes at FedEx, when they could not pay the fuel bill, the founder took the last five thousand dollars and went to the casino. He made enough playing Blackjack to save the company. Great story, inspiring story, but how many people tried this and failed?

Since the inception of just-in-time manufacturing, there has been an infestation of this concept in all areas. Not just in manufacturing, but also in business. Any idle resource is the enemy. We try to squeeze out the last drop from a resource. Incidentally, family-owned organisations tend not to do this as much as corporates. Yet, it is still the case often enough. If you have stretched yourself thin, you have nothing to fall back on when crisis hits. Redundancy might be a waste in good times, but it is a lifeline in bad times. I wish we had had some redundancies. It would have given me something to work with. Instead, we had to rely on the goodwill of people. Of consultants to give us time to pay. Of the banks to give me the chance to restructure. Of employees to stomach the stressful times. And oh behold, I was not given the chance to restructure, turning the crisis into a disaster.

The same principle shows up in how organisations with the highest possible stakes — nuclear plants, aircraft carriers, air traffic control — manage to operate for years without catastrophic failure. High Reliability Organisation theory, developed by researchers including Karl Weick and Kathleen Sutcliffe, identifies redundancy as one of the defining features these organisations share: multiple, independent channels of communication, decision-making, and implementation, so that no single failure — human or mechanical — can bring the whole system down. Crucially, that redundancy is paired with decentralised authority, so the people closest to a problem can act on it without waiting for a decision to travel up a chain of command and back down again.

The combination of redundancy and decentralised decision-making is a powerful set-up for adverse times.

From 2005, the two families (Porsche and Piëch) — who between them owned all of Porsche — pursued a secret strategy to take over Volkswagen, a company roughly thirty times Porsche's size, by building a stake through stock and options. It worked, spectacularly, for years: Porsche made vast profits squeezing short sellers and selling options between 2005 and 2008, and by late 2008 it had acquired 42.6% of Volkswagen outright, with the option to acquire another 31.5%. But that position was built entirely on leverage, not reserves — the company had taken on $13 billion in debt to finance the acquisition. The plan was resting on an assumption: that once Porsche crossed the 75% ownership threshold, it would gain access to Volkswagen's own $12 billion in cash reserves, which would retroactively cover the debt. There was no independent buffer. The whole structure depended on the next step of the plan executing on schedule, in calm conditions, with credit still available. Then, just five months before the acquisition was due to complete, the global financial crisis worsened. Thus Porsche ran out of money and the banks that had extended the $13 billion suddenly wanted it back. Porsche's debt had risen to a level it could no longer service, leaving it unable to secure the financing needed to finish what it started. The family that had been five months from controlling Europe's largest automaker instead had to watch Volkswagen turn around and acquire Porsche

Concentrated knowledge

This may seem similar to centralised decision-making, but it is distinctly different. Knowledge and power are often not concentrated in the same person. A classic example is a key employee who hoards knowledge. The same can be true in a family. This may be the grey eminence who has already stepped away, but kept knowledge to themselves to stay needed. This can also be any other family member that is taking care of a specific area. For example, the person that is taking care of the financial matters for the family. Another classic example is when the family has a trusted family lawyer. Said lawyer has done everything; they know everything from wills to business structure etc. What happens when this lawyer passes away? What happens when the patriarch who was close to the lawyer passes away? Both cases that can have dire consequences.

During the crisis I managed, knowledge was concentrated with me. Due to the relationship dynamics that existed in our family, I started to hoard information and keep it to myself. This had the effect of fueling more mistrust in the family than we already had. I did not pass away, but I could have thrown in the towel. Any of the two cases would have been a disaster. Concentration of anything is not good.

The academic world of family business has studied this under the following term: founder centrality, studied most directly by Kelly, Athanassiou, and Crittenden. They found that a founder's dominant position shapes decisions, the firm's entire culture, strategy, and its relationships with the outside world. That centrality often persists as "legacy centrality" long after the founder has formally stepped back. The question is what, exactly, makes that position so hard to hand over. The answer lies in a distinction the philosopher Michael Polanyi drew decades ago, and which family business researchers still return to: the difference between explicit knowledge, which can be written down and transferred, and tacit knowledge, which cannot. A founder's most valuable knowledge is almost always tacit: which supplier can be trusted under pressure, how a particular banker actually thinks, which employee's objection is worth taking seriously. It's precisely because this knowledge is so hard to imitate that it becomes the family firm's greatest asset. And it's precisely because it's so hard to transfer that it becomes a risk. One study of succession put the mechanism bluntly: the founder is typically the main source of knowledge in the family enterprise, and unless he deliberately works to transfer it while he still can, the family loses it the moment he's gone.

Sam Steinberg built his mother's small Montreal grocery store into Quebec's largest supermarket chain (Steinberg Inc.) — by his death in 1978, it was grossing over a billion dollars a year. He ran the company for decades with a distinctive, personal style. For example, rather than a formal labour policy, Steinberg had no formal plan to deal with unions at all — his approach was simply that if workers wanted to unionize, so be it, and if not, that was fine too. That worked because it was Sam himself running it. His personal relationships, his judgment, his read of the room with union leaders, none of it written down anywhere, all of it living in his head. When he suddenly died of a heart attack, that laissez-faire approach hurt the company, and Steinberg employees unionised quickly in the vacuum he left behind. Layered on top of that, Sam had also left no clear succession plan, and disagreement among his surviving daughters. Each held equal controlling shares, stipulated to be voted together as one. This led to a prolonged power struggle that dragged on for years before the company was sold in 1989. It went bankrupt in 1992, seventy-five years after it began.

Absence of skin in the game

Who has something to lose if things go bad? Who actually has the most to lose? And who is in control? Who sits at the table when decisions are made? It may sound straightforward, but it is not. Often the person who decides has the least to lose. Sometimes the people who have the most to lose are not even informed properly. Skin in the game is a crucial criterion for an antifragile family. The absence thereof is a clear sign of fragility. In our case, there were actors inside the family who had more to lose and some who had less to lose. In the wider family, some had nothing to lose. It was interesting to see how some important decisions could not be made due to people sitting at the table with no skin in the game.

This is not just an important factor inside the family but also with employees, consultants or business partners. Who has skin in the game and who has not? Skin in the game is a great way to align incentives. But it is a double-edged sword. If the incentives are not set properly, skin in the game can become a problem. A classic example is when you have partners who have more to gain when the family business is sold. Whenever things get difficult, they will lobby and push for the sale. We saw that with one consultant in our case; he pushed for the sale. Especially when the first offer was on the table. He would have been paid handsomely; we would have had little left over. I stood my ground and refused and kept looking for a better offer. A much better offer came through.

Nassim Nicholas Taleb built his own framework around this criterion. The phrase itself — skin in the game — has become the shorthand for it: when the person making a decision doesn't bear the downside if it goes wrong, they take on more risk than they would if they were exposed to the consequences themselves. Taleb's own illustration is a parent who hands a teenager the keys to a Ferrari. The parent bears none of the risk the teenager is about to take with it. The underlying mechanism is old and well studied in economics under a more technical name: the principal–agent problem, formalised by Jensen and Meckling in the 1970s. Whenever one person (the agent) makes decisions on behalf of another (the principal) but doesn't fully share in the consequences, their incentives drift apart from the principal's. The agent optimises for what benefits them personally, not necessarily for what's actually best for the person or family they're deciding on behalf of. In a family fortune, this shows up constantly and rarely announces itself. The advisor is paid a fee regardless of outcome. The heir spending from a trust they didn't build. The family member managing a division whose failures land on the balance sheet rather than on them personally. None of them are doing anything as crude as fraud. They simply aren't the one who pays if it goes wrong.

The Sackler family controlled Purdue Pharma and sat on its board, directing the aggressive marketing of OxyContin even as evidence of its addictiveness mounted. The company had already pleaded guilty to federal charges over its marketing practices as early as 2007. The family's exposure to that decision-making, however, wasn't structured the way the exposure of an owner-operator usually is. As the company's legal liabilities grew larger and more certain through the 2010s, the family increased, rather than reduced, the pace at which it pulled money out of the business. In the years immediately before Purdue's 2019 bankruptcy filing, the Sacklers took roughly $11 billion in distributions out of the company. They moved much of it into private trusts, structurally separating their personal wealth from the entity that was about to absorb the legal consequences of the decisions they had made. By the time the lawsuits arrived in force, the money had already left the building the liability was aimed at.

Succession planned for death, not incapacity

Thinking about one’s own death or incapacitation is uncomfortable. Especially if one has been extremely successful, the idea of death goes against the grandeur that has been achieved. Due to this, succession is often not planned at all. Even in the families that do plan for death, there is often no plan for incapacitation. Many wealth holders plan to hold onto their wealth till “death do us part”. A family where this is the succession plan is clearly fragile. Not only do our decision-making capabilities and energy levels degrade with age, but the chance of sudden death or incapacitation becomes larger. My father planned to hand over relatively early as long as we still had our company. After the sale, he changed his mind. We found a partial solution, thank god. His money management in his later years was horrible. And this is something I see over and over again. The patriarch or matriarch holding onto the wealth too long.

Antifragile families hand over wealth, responsibility and authority gradually. It is important to slowly increase the responsibility and power of the heirs. They do not only bring fresh eyes and ideas to the table, but also clearer mental capacity and energy. Times change and so must the family. More practically, there needs to be an emergency plan. Who takes over when something happens to the leader? Are the needed documents and certificates in place? And even when there is some planning in place, incapacitation becomes troublesome. My father suffered from dementia in his final months. I had a power of attorney for this situation. But we had no concrete plan of how to assess this issue. Even worse, when it became clear he had lost the ability to decide, it took us weeks to get a doctor to sign off on it. What we did not think about was that many doctors refused, due to the risk that the wealth posed. We had to get our lawyer involved to get it through.

This marker doesn't have the same body of peer-reviewed research behind it as the others. However, there is a good body of work in estate law and succession-planning practice rather than academic journals. Even the professionals who deal with this for a living keep flagging the same blind spot, over and over, without anyone having formally studied why it persists. The observation they keep converging on is this: death is, paradoxically, the easier of the two events to plan for. It triggers a clean, unambiguous process — a date, a certificate, a will or a buy-sell agreement that activates, often with life insurance already in place to fund it. Incapacity offers none of that clarity. There is no single moment that marks it, no document that self-executes, and most governance agreements never even define what "unable to fulfil the role" actually means, let alone who gets to decide when that line has been crossed. And the ambiguity isn't a remote, unlikely concern reserved for the elderly: the U.S. Social Security Administration estimates that a quarter of today's twenty-year-olds will experience a prolonged work absence due to disability before they even retire.

There's a deeper psychological reason this gap persists, and it has its own body of research behind it. In 1973, the anthropologist Ernest Becker published The Denial of Death, arguing that most of what humans build — status, achievement, legacy, culture itself — functions as a buffer against the unbearable fact of our own mortality. A decade later, psychologists Jeff Greenberg, Sheldon Solomon, and Tom Pyszczynski turned Becker's idea into a testable framework they called Terror Management Theory. The experiments since have held up: when people are made to confront their own mortality, even briefly, they don't just feel anxious; they actively avoid situations that force further self-reflection, reaching instead for whatever gives them a sense of permanence and significance. For a patriarch or matriarch who has spent a lifetime building status, legacy, a name that outlasts them, succession planning isn't a neutral administrative task. It's the one conversation that asks them to look directly at the thing they have no control over. That's not a character flaw unique to any one family. It's closer to the mechanism Becker and his successors spent decades documenting: the more someone has built as a shield against death, the harder it becomes to plan for it.

If death is supposed to be the easier of the two to plan for, Prince's estate shows what happens when a family doesn't even manage that. He died unexpectedly in 2016 at 57, leaving behind a business empire built entirely around his own name, catalogue, and personal control. His estate was reportedly worth over $150 million without a will, trust, or succession plan of any kind. Because he died intestate, Minnesota's default inheritance laws took over, and the courts had to first determine who his legal heirs even were before anyone could make a single decision about the business. That task fell to six siblings and half-siblings, several of whom barely knew one another, none of whom had ever been prepared to run anything together. What followed was six years of legal proceedings, tens of millions of dollars in fees, court costs and taxes, and two of the original heirs dying before the estate was even settled. Prince’s death was exactly the kind of scenario the succession-planning literature says families are supposed to be equipped for. And still, nothing had been put in place. If a family can fail to prepare for the easy version of this problem, then not being prepared for incapacity — the more difficult one — becomes almost guaranteed.

Final Words

There is so much we learned from our governance fragility. How our chain of command led to the shattering of our estate (acquired in 1730) and the family. The signs were clear. The research on this is clear. Yet, we did not change anything. Just as many families. Consider this article — the whole four-part series once it is done — carefully and act on it. Act on it for the good of your family or your clients. Don’t wait till trouble comes your way. Prepare, and maybe then the trouble will turn into an opportunity.

ACCESS THE ARCHIVE ON SUBSTACK

Signs of a Fragile Family's Command Structure