The architecture comes before the product

Posted on: 25 May 2026

Saturday 23 October 2010, mid-morning. Patrick Thomas, then chief executive of Hermès International, was cycling alone through the rolling countryside of Auvergne, far from Paris, far from anything resembling a corporate office. His phone rang. On the other end was Bernard Arnault, chairman of LVMH and the wealthiest man in France. The message was almost surgical in its brevity. LVMH had acquired 14.2% of Hermès through a sequence of equity swaps quietly executed over the preceding months, the announcement would hit the wires within two hours, and Thomas was being informed as a matter of courtesy rather than consultation. Thomas, the first non-family chief executive in the brand's century-and-a-half history, pocketed the phone and tried to absorb what he had just been told. By the time he returned to Paris the news was already public, and the most audacious corporate raid in the modern history of European luxury was underway.

What happened in the eight weeks that followed is the part of the story worth pausing on. Not the attack itself, which has been reconstructed dozens of times as financial thriller, but the response. A family of artisans, roughly seventy descendants scattered across six generations going back to 1837, managed to build in less than two months a legal architecture that fifteen years later still underpins the entire strategic freedom of the brand. This is the structural detail that explains things which neither product quality nor marketing sophistication could ever explain on their own.

The growth pattern of Hermès is genuinely anomalous within the luxury sector. Self-limited expansion at roughly 6 to 7 per cent annually, whilst the rest of the comparable industry expands at 15 to 20 per cent. A systematic refusal to use celebrities as brand ambassadors at a time when LVMH, Kering and the wider sector treat celebrity endorsement as a fundamental commercial tool. Sixty manufacturing sites concentrated in France whilst nearly every comparable luxury house has accelerated production offshoring. An internal training school, the École Hermès des Savoir-Faire, that takes four years to produce a fully trained artisan. Multi-year waiting lists for the Birkin managed as a deliberate structural device rather than as the embarrassing side-effect of insufficient capacity. All of this is routinely narrated as evidence of brilliant brand strategy, of a family that understood luxury better than its competitors. The narrative is convenient and partially true, but it leaves out the structural condition that makes those choices defensible in the first place. The proprietary architecture of the brand is what shields these decisions from the standard erosive pressures of the capital markets, and that architecture was not the product of farsighted founder vision. It emerged in eight weeks as a defensive reaction to an attack.

Arnault had constructed his position through a sophisticated derivatives scheme. Beginning in 2008, LVMH had quietly accumulated approximately 5 per cent of Hermès through individual transactions kept just below the mandatory disclosure threshold, remaining legally invisible. In June 2010 LVMH activated a series of equity swaps with various banks, financial instruments that allowed it to obtain economic exposure to Hermès stock without the ownership formally appearing in official registers. When those contracts were exercised in October, LVMH suddenly held 14.2 per cent of the capital, well above every French mandatory disclosure threshold. In 2012 the Autorité des marchés financiers would establish that the entire accumulation method had been constructed to circumvent transparency obligations, fining LVMH ten million euros, but by then the operational effect had been secured. A position difficult to attack on legal grounds, built in the shadows, capable of converting itself into a gradual climb towards control.

The Hermès family, as often happens in multigenerational dynasties, was fragmented. Seventy-two adult descendants distributed across three main branches, Dumas, Guerrand and Puech, each holding variable individual stakes with no formal coordination mechanism between them. It was exactly the configuration a corporate raider hopes to find. Ownership nominally concentrated but operationally dispersed, where each family member could sell independently of the others, where an acquisition offer aimed at individuals could find cracks anywhere, where any meaningful defence would require a coordinating capacity the family had never built and which families historically struggle to improvise under pressure.

What the family did in the weeks following 23 October is a piece of institutional engineering of remarkable precision. Approximately fifty adult descendants were convened and signed a binding pact. All their shares would be transferred into a single holding vehicle, called H51, the name corresponding to the fifty-one founding members of the agreement. The operating rule was simple and unforgiving. Any member wishing to sell their shares had to first offer them to other H51 members at market value through obligatory pre-emption, and only if no other member wished to purchase could the shares be released to the outside market. In practical terms it became impossible for an external attacker to accumulate further stakes by buying from individual family members, because every potential transaction was intercepted from within before reaching the market.

H51 was incorporated in December 2010, eight weeks after the call from Arnault. On 7 January 2011 the AMF issued the decisive ruling. The pooling of family shares within the holding did not constitute a concerted action sufficient to trigger a mandatory tender offer on the remaining minority shareholders, an obligation under French securities law that would otherwise have made H51 economically impossible to implement. That ruling was the legal exemption that crystallised the defence. From that moment the family controlled operationally roughly 50.2 per cent of the capital through H51, with a thirty-year lock-up that prevents any disaggregation until 2031, and additional voting blocks on board nomination decisions. Arnault could still retain his eventual 23 per cent position, but he could no longer grow it, could not obtain board seats, could not influence strategic decisions. In 2014, under court-mediated agreement, LVMH released the entire stake by distributing the shares to its own shareholders.

Here is the part of the story that the standard reconstructions miss. H51 is not merely a defence against Arnault. It is a structure that permanently crystallises the strategic freedom of Hermès against any future pressure, whether it comes from public markets, from financial analysts, from activist funds, from minority shareholders requesting faster growth or higher payouts. The thirty-year lock-up means that for at least the next six years no one inside the family can break the pact, and given that H51 controls the majority, no one outside the family can force a strategic shift. The family has, quite literally, tied its own hands for four generations, voluntarily accepting that its shares are illiquid relative to the outside market in exchange for the certainty that the strategic direction of Hermès cannot be captured by standard capitalist logic.

This is the structural mechanism that explains the subsequent choices of the brand. Growing at 7 per cent rather than 20 per cent is not a moral choice, it is a choice made practicable by the absence of shareholders demanding higher growth. Refusing celebrity endorsement is not aesthetic purism, it is an operational freedom that exists only because no financial analyst can credibly force the management's hand. Maintaining sixty manufacturing sites in France, where labour costs four times what it would in Vietnam, is not patriotism, it is a decision that can be sustained because those making the decision do not answer to a board of directors where sovereign wealth funds demand margin optimisation. The École Hermès des Savoir-Faire trains artisans for four years before they become productive, an investment no chief executive of a standard listed company could defend before shareholders without paying a significant political price. Hermès can do all of this because H51 makes it possible. Fourteen years after the Arnault attack, the market capitalisation of the group has exceeded $250 billion, occasionally surpassing LVMH itself.

The structural pattern is clear. Most companies attempting today to follow Hermès as a luxury positioning model cannot actually replicate its choices, not because they lack the vision, but because they lack the proprietary architecture that would make those choices defensible over time. A standard listed company, with a significant free float and dispersed shareholding, is structurally incapable of sustaining self-limited growth for more than a few business cycles before analysts, activist funds, hostile acquisitions or simply the share price compel management to capitulate and accelerate. The post-hoc narrative that frames Hermès as a story of brilliant family vision is partially true, but it obscures the simpler underlying truth. Without H51, without the thirty-year lock-up, without the obligatory pre-emption, without the voting blocks, even the most refined strategic vision in the world would have been eroded within a few years by the standard pressure of capital markets.

There is a further detail worth attention, because it distinguishes Hermès from other luxury family configurations. One family member, Nicolas Puech, refused to join H51 and kept his 5.7 per cent stake outside the pact. The family has tolerated the dissent without managing to absorb it, and the Puech stake remains today an open element, the object of complex succession litigation that has included Puech's attempts to bequeath his shares to a Moroccan employee through declarations subsequently contested. Even the best architectures have cracks, and this is the H51 crack. But the general point holds. Fifty members out of seventy-two chose individual illiquidity in exchange for collective long-term solidity, and that choice produced the strategic freedom that today everyone studies and few can genuinely replicate.

Beneath the story of Hermès, then, is a different story from the one usually told. It is not the story of a family that understood luxury better than its competitors, but the story of a family that understood, under attack, the importance of a defensive proprietary architecture, and had the discipline to construct it in eight weeks, and maintained it for fifteen years by voluntarily forgoing the individual liquidity the market would have offered. The strategic choices that today distinguish the brand are practicable only because H51 exists. Without the call from Arnault on 23 October 2010, H51 probably would not exist, and without H51 Hermès would probably be something else today. Perhaps still a good luxury brand, but an ordinary one within the sector, subject to the same homogenising pressures that have transformed its competitors into variations of each other.

Anyone looking at Hermès and trying to understand its success must therefore look at H51 before looking at the bags, the ateliers, the school, the advertising. The architecture comes first, everything else is consequence.