Neither luxury nor fashion lies within my natural circle of competence. I grew up in a working-class family where frugality was close to a family motto, and long regarded luxury chiefly as expensive status consumption. Don Thompson’s The Curious Economics of Luxury Fashion—encountered through Guy Spier’s recommended reading list—did something more useful than provide a tour of the industry: it forced a prior belief to be revised.

The book made the economics of luxury more convincing, but the permanence of its moats less so. A luxury house can possess pricing power, controlled distribution and an identity that capital alone cannot reproduce. Yet its value depends on whether each new generation chooses to inherit the same cultural meanings.

Seeing the system

The simplest way to misunderstand a luxury business is to compare the selling price of a handbag with the apparent cost of leather and labour. The customer is not buying the physical inputs alone. The object may carry craftsmanship, history, scarcity, ritual, social recognition and a particular account of the person who owns it. The economic product is larger than the manufactured object.

That helps explain why more than one strategy can succeed. LVMH operates a portfolio of maisons across fashion, jewellery, watches, wines, spirits, beauty and retail. Its scale allows it to allocate capital and talent across brands and support a global distribution and communication apparatus. Hermès follows a more concentrated model. It controls distribution tightly, expands craft capacity gradually and accepts that production cannot be increased as quickly as demand might permit.

The contrast prevents “luxury” from becoming a single formula. LVMH’s portfolio breadth and Hermès’s concentrated identity can both produce strong economics. What they share is disciplined control over the context in which their products are made, presented and acquired.

The system is expensive to maintain. In 2025, LVMH generated €80.8 billion of revenue and €17.8 billion of recurring operating profit, a margin of 22 per cent—excellent, but far below the naïve margin one might infer from a product’s retail price. Stores, staff, communication, product development and creative renewal consume real resources. Some of this spending resembles maintenance capital expenditure for an intangible asset. A factory must be repaired if it is to keep producing; a brand must be repeatedly expressed if it is to retain meaning.

The analogy is imperfect. A machine’s deterioration can often be observed. The return on a fashion show, flagship store or artistic collaboration is much harder to isolate. Management can count visitors and impressions, but not precisely identify how much of next year’s willingness to pay came from one campaign rather than decades of association. The spending is discretionary in an accounting sense and may be necessary in an economic one.

A real, but renewable, moat

An unlimited-capital thought experiment makes the moat more concrete. With enough money, one could hire designers, buy property, build workshops and purchase advertising. That would still not be enough to recreate Hermès rapidly. Capital cannot instantly manufacture a brand’s history, trained artisan base, controlled scarcity, customer rituals or social permission to charge comparable prices.

That is a real moat precisely because its most valuable elements cannot be assembled on an ordinary corporate timetable.

But a moat of collective belief has an unusual vulnerability: management does not own the belief outright. It can control quality, supply, pricing, distribution and communication. It cannot order customers to keep caring, or guarantee that the symbols valued by one generation will be adopted by the next.

This creates a difficult balance. A successful house must grow without making itself ordinary. Excessive volume, indiscriminate licensing, discounting or confused positioning can turn a scarce cultural asset into a conventional consumer product. Refusing to change can be equally dangerous if the brand becomes culturally inert. The task is not simply to preserve a fixed identity, but to renew it without breaking continuity.

That distinction matters to investors. A company may report excellent current margins because it possesses a strong inherited position. The valuation question is whether today’s expenditure and decisions are strengthening the institution that will produce cash flows ten or twenty years from now. Financial statements show the cost of maintaining the system, but reveal only imperfectly whether the cultural asset is appreciating or being consumed. A second difficulty is determining whether its meaning will transfer to future customers—what might be described as an age–period–cohort problem.

The age–period–cohort problem

Kweichow Moutai provides a useful analogy, although it should not be pushed too far. Kweichow Moutai is a Chinese spirits company; Moutai is its flagship product; baijiu is the wider category. Hermès sells different products to a more global customer base. Yet both businesses depend partly on socially embedded patterns of consumption.

Kweichow Moutai has several qualities investors often admire. Its flagship product is scarce, highly profitable and deeply embedded in Chinese traditions of banqueting, gift-giving and relationship-building. Its more recent business results have weakened, making the durability of the franchise a more immediate question.

Kweichow Moutai’s 2025 results make the issue more concrete: revenue declined by 1.2 per cent and net profit attributable to shareholders by 4.5 per cent. Those figures do not establish permanent decline. They are consistent with cyclical, regulatory and cohort explanations. The danger lies in selecting whichever narrative best suits a pre-existing valuation.

One longer-term concern is whether younger Chinese consumers are less attracted to strong spirits and the rituals surrounding them. That observation is consistent with at least three explanations.

An age effect would mean younger consumers have rarely been the natural customers for premium baijiu. As they age, earn more and enter professional settings in which it is expected, many may adopt the category.

A cohort effect would mean the present younger generation has formed persistently different attitudes towards alcohol, hierarchy, business entertainment, gifting or status, and may not adopt Moutai to the same degree as it ages.

A period effect would mean current economic conditions, regulation, health concerns or changing norms are temporarily suppressing demand across several generations.

These explanations imply different futures, but observational data cannot cleanly separate them. Age, period and cohort are mechanically related: knowing any two determines the third. Statistical methods can add assumptions or evidence, but the data do not identify three independent linear effects by themselves.

The framework does not answer the investment question; it improves it. Rather than asking only whether young people consume less Moutai or buy fewer traditional luxury products today, an investor should ask what evidence would distinguish delayed adoption from permanent change. Relevant evidence might include the age of first purchase across cohorts, progression from entry products to core products, repeat purchasing and changes within the same individuals over time. None is decisive alone.

What valuation can—and cannot—do

The duration of exceptional economics matters most when the price already assumes that they will persist.

A separate, dated research exercise examined Hermès and LVMH. At the prices used—€1,659 and €471.30 respectively—the scenarios placed them in different positions. Hermès required outcomes materially stronger than the constructive case; LVMH appeared closer to estimated value, but still offered little margin for error or an attractive prospective return.

These results are not precise target prices. Long-duration valuations are acutely sensitive to required returns, terminal assumptions and the treatment of reinvestment. The research appendix therefore shows scenarios and sensitivities rather than presenting one number as fact.

There is a more important exercise: what must be believed for the observed price to make sense? For Hermès, the answer required not merely an exceptional company, but a long continuation of high growth, extraordinary margins, successful absorption of new craft capacity and continued cultural relevance. Each proposition may be defensible. The price left little protection if one proved less durable than expected.

For LVMH, whose research pack incorporates its first-half 2026 results, the valuation was less demanding because both earnings and the multiple had already fallen substantially from their 2023 peaks. Yet a share price being far below a former high does not make it cheap. The high was itself a market forecast, not evidence of intrinsic value. By 2025, revenue, recurring operating profit and margins had all declined from 2023, while Fashion and Leather Goods no longer combined its previous growth rate with margins near 40 per cent. The lower price reflected lower earnings and a lower multiple.

A reverse discounted-cash-flow model can reveal how much durability the market has priced in. It cannot, however, establish whether the cultural mechanism supporting that durability will persist. That remains a question of evidence, judgment and humility about what can be known.

Postscript: 29 July 2026

Two days after the €1,659 Hermès closing price used in the research, the company reported second-quarter sales growth of 6.7 per cent at constant exchange rates and a first-half recurring operating margin of 41 per cent. Its shares nevertheless fell 11 per cent in one session, closing at approximately €1,509, as investors focused partly on subdued growth in Asia excluding Japan and the absence of a convincing recovery in China.

The decline did not prove that the previous price was wrong, nor that the new price was attractive. It illustrated something narrower: an exceptional business can report strong absolute results and still disappoint when its valuation already capitalises exceptional expectations. The original model has not been rewritten after the market moved; the results are treated as subsequent evidence rather than retrospective validation.

The price of permanence

The inquiry began with the view that luxury was mostly an expensive expression of status. It ended by making businesses built around scarcity, craftsmanship, ritual and social meaning more intelligible. Their strongest assets may be absent from the balance sheet, difficult to copy and capable of supporting remarkable pricing power.

Those assets also demand continual expenditure, disciplined distribution, excellent management and the renewal of cultural relevance across generations. Luxury looks easy only if one sees the selling price and ignores the system required to sustain it.

The moat can be real without being permanent. The company can be exceptional without the share being attractive. And a valuation can be internally consistent while depending on a cultural forecast that the model itself cannot justify.

That is the price of permanence: before paying for decades of exceptional economics, an investor must decide not only whether a brand matters today, but why it should continue to matter to people who have not yet become its customers.

Disclosure: I hold no position in Hermès, LVMH or Kweichow Moutai as of 2 August 2026. This article is for research and discussion and does not constitute investment advice.

Sources and further detail

  1. Don Thompson, The Curious Economics of Luxury Fashion: Millennials, Influencers and a Pandemic.
  2. LVMH, 2025 results and first-half 2026 results.
  3. Hermès, 2025 regulated information and first-half 2026 results.
  4. Kweichow Moutai, Annual Report 2025.
  5. Martin Browning, Ian Crawford and Marike Knoef, “The age-period cohort problem: set identification and point identification”.
  6. Reuters, “Hermès shares dive as sales and Chinese weakness disappoint”.
  7. Linxuan Zhang, “Research Appendix: What Hermès and LVMH’s Prices Implied”, analysis cutoffs 28 and 29 July 2026.