The Auspicious Tower

On the columbarium industry of the Chinese world, and what it does with the dead

A note on register. What follows attends to practice rather than to belief — to what is done with the dead, what is paid, and by whom, and to whom the money flows — on the principle that in this cultural world the doing is the substance and the believing is, at most, its accompaniment. The judgment such an account may seem to invite is left, deliberately, to the reader. The materials are arranged to be looked at; the looking is enough.

I. The brochure

Begin with what is being sold, in the seller's own words.

A columbarium is a building for the storage of cremated remains: a structure of niches, each sized to hold an urn or two, stacked in tiers along interior walls, sometimes to a great height. In the Chinese world it is now the ordinary destination of the dead. And in its commercial form it is marketed, with care and at considerable expense, in the vocabulary of two ancient registers at once — the cosmographic and the ancestral, fortune and filiality — applied to a shelf.

Consider the language. A private high-rise columbarium in Hong Kong, designed by a German architect in a wave-fronted fan of twenty-three thousand niches, is described by its operational director in terms that locate the dead squarely in the housing market: she has said that she imagined herself "as someone 'living' inside one of these niches," and asked "what kind of home I wanted to stay at when I'm gone." The building's floors are sorted by religion — an airy Buddhist level, a section under the eye of Guanyin, a secular floor whose doors are decorated with gold coins to signify a prosperous afterlife. In Taiwan the market leader operates a twenty-story tower in the hills outside Taipei, its interiors compared by visitors to a luxury hotel; niches there have run from the equivalent of a few thousand US dollars to many times that, priced, like apartments, by floor and aspect and view. On the Chinese mainland the dominant firm sells, alongside its standard plots, multi-floor mausoleums reached by keycard, with an on-site restaurant for the bereaved and an in-house atelier of sculptors producing memorial busts; some seven hundred celebrities are interred at its flagship Shanghai cemetery, and the management notes, without irony, that couples come there to take wedding photographs.

The vocabulary throughout is aspirational, and that is the point. The marketing performs a precise alchemy: it moves the dead out of the register of dread — the register of the ghost, the corpse, the pollution that Chinese funerary practice has always worked to manage and contain — and into the register of property and prestige, which is the least stigmatized and most desired register the culture has on offer. A niche on a high floor with a mountain view, fronted by water, certified auspicious, is not a morbid purchase. It is a good address. The brochure sells the dead a good address, and sells the living the relief of having secured one for them.

Now set beside the brochure a second document: the income statement.

II. The revenue curve

The dominant death-care firm on the Chinese mainland — listed in Hong Kong, incorporated in the Cayman Islands, and known half-admiringly as "the Moutai of funeral services" after the luxury liquor whose margins it was said to rival — posted, for the better part of a decade, the kind of numbers that draw private equity. Its revenue climbed without interruption: roughly 1.09 billion yuan in 2018, 1.22 billion in 2019, 1.45 billion in 2020, 1.62 billion in 2021, 1.79 billion in 2022, with net profit rising each year alongside. Its initial public offering, in an earlier year, had been oversubscribed many times over; a co-founder of the Carlyle Group had taken a personal stake after visiting the Shanghai cemetery; analysts compared its margins favorably to a technology company's. The investment thesis was simple and, on its face, irresistible: an aging population, a cultural mandate of filial care for the dead, strong pricing power because families want a resting place near home and will not haggle hard over a parent's grave. Death is certain; the market is captive; the margins are protected by piety.

Then, in the first half of 2025, the same firm reported that its revenue had fallen by roughly 44.5 percent against the same period the year before, and that it had recorded a loss — its first in over a decade — of some 261 million yuan. It declared a dividend nonetheless. Some months later it announced that the publication of its audited annual results would be delayed: its board and auditors had identified, in the firm's own published language, issues involving certain cash withdrawals, disbursements, and procurement-related payment arrangements requiring further investigation. The flagship of the industry, the market's model of recession-proof piety, had in the space of a year posted a historic loss and entered the fog of a financial probe.

The Taiwanese market leader shows the same shape at a different scale. After an era of spectacular profitability that made its founder one of the wealthier men in the territory, its revenue has eroded; its share price sits at a fraction of its former level. The pattern, across the two largest markets, is a steep ascent followed by a turn. The brochure goes on promising the auspicious view. The income statement has begun to describe something else.

The piece that follows is about the distance between those two documents — between what the industry says it is selling and what its own numbers, and its own cosmology, reveal it to be doing. The argument is not that anyone is lying. It is that a system performing exactly as designed, by everyone's lights, produces an outcome that the system itself has no language to assess.

III. Correct practice

To see why, one has to begin with how this culture has organized the relation between the living and the dead — and the organizing principle is not belief but practice.

The point is not a casual one; it is one of the better-established findings in the anthropology of China. James Watson, surveying the standardization of ritual across late imperial China, argued that what unified the culture was not shared doctrine but shared performance: the state and the elite enforced orthopraxy, correct practice, rather than orthodoxy, correct belief, and by doing so could fold people of wildly varying convictions into a single cultural order. "To be Chinese," Watson wrote, "is to understand, and accept the view, that there is a correct way to perform rites" — above all the rites of the life-cycle, the wedding and the funeral. One participated in the culture by performing its rites correctly; what one privately believed while doing so was one's own affair, and largely beside the point. (The thesis has been qualified since — Donald Sutton and others have shown that officials often tried and failed to standardize practice, that regional variation ramified beneath the appearance of uniformity, that some reported orthopraxy was the wishful writing of local elites. The qualifications matter, but they do not overturn the core: the test that mattered was the test of correct performance, not of correct conviction.)

This has a consequence that is easy to state and hard to feel one's way fully into, for a reader formed in a tradition where morality is imagined as belief enacted. In a world ordered by correct practice, the unit of moral life is the act-in-its-context, performed rightly or wrongly, and not the principle held in the heart. A child is taught, from the first spoonful, that this is guāi — good, correct, what one does — and that is not; the training is continuous, concrete, and prior to any articulation of why. One does not first acquire a principle and then apply it; one acquires, act by act over a lifetime, a vast and finely-calibrated sense of what is correct in each situation. Conviction is not the goal and is not required. The performance is the substance.

From this a great deal follows that looks, from outside, like contradiction and is not. If correctness is contextual — if the test is always what is the right thing to do here, in this relation, on this occasion — then two acts that a principle-based morality would find inconsistent may both be entirely correct, each in its place, with no tension between them, because there is no overarching principle demanding they agree. The same person may perform scrupulous rectitude in one context and, in another, do a thing that an outside observer would call deception or favoritism or worse — and experience no inner division, take pride in both, and be, by the only test the culture applies, a thoroughly good person in each. There is no hidden "real self," beneath the contexts, whose principles are being violated. There are the contexts, and the conduct, and the conduct's correctness. This is not the absence of a moral order. It is a different moral order, and a robust one, and it is the one within which the columbarium industry operates.

IV. Feeding the dead

The same practical logic governs the dead themselves, and here it acquires a cosmological edge that the industry's troubles bring to a point.

The classic anthropological account, owed to Arthur Wolf and refined by others, sorts the supernatural beings of Chinese popular religion into three categories that turn out to be three positions rather than three kinds. There are gods, who are fed and honored by many, beyond their own descendants, and who are correspondingly powerful. There are ancestors, who are fed by their own living descendants and are, accordingly, benign — present, attended, and often helpful to the line that maintains them. And there are ghosts: the dead who are not fed by descendants, who are therefore hungry, and who are, therefore, dangerous. The decisive word in each case is fed. What determines which category a dead person occupies is not the quality of the soul but the maintenance of the offering — whether someone, at the proper times, in the proper place, brings food. And because the categories are defined by this practice rather than by any fixed essence, a being can move between them. An ancestor whose line stops feeding him slides toward the condition of a ghost. As Wolf and Steven Sangren both observed, the matter is wholly relational: one lineage's revered ancestor is, to everyone outside that lineage, precisely a potentially hungry and dangerous stranger-dead. One household's ancestor is another's ghost.

This is not abstract doctrine; it is encoded in the smallest reflexes of daily practice. One does not stand one's chopsticks upright in a bowl of rice, because that is the posture of the incense offered to the dead, and to make the gesture casually is to issue an open, untargeted invitation to the dead at large — and the dead at large, the unattached and unfed, are exactly the dangerous ones. The offering is supposed to be specific, directed, lineage-bound: my dead, fed by me, at the altar or the grave that is ours. The whole architecture of safety lies in the specificity of the feeding. Generic hospitality to the dead is not generous; it is reckless, because it draws the hungry ones in.

Hold this beside the columbarium, and the difficulty comes into focus. The traditional ancestor was a fed presence in a fixed place maintained by a resident line — a grave on a family hillside, tended at Qingming by descendants who lived within reach of it, the feeding and the place and the lineage all bound together in one local circuit. The columbarium, by design and at scale, takes the bones out of that circuit. It moves them off the ancestral ground into a commercial tier of niches in a city, often far from where any descendant now lives; it replaces the tended grave with a maintained product, the descendants' offerings with an annual fee paid to a management company; and it does this just as the descendants themselves are dispersing — to the cities, abroad, out of reach of the hillside their grandfather was the soul of. The industry does not intend to sever the feeding-circuit that distinguishes an ancestor from a ghost. But severing it, at scale, is the structural effect of what the industry does. By the cosmology's own terms — not by any imported judgment, but by the tradition's own definition of what makes a dead man dangerous — the consolidation of the dead into distant commercial inventory, decoupled from the resident, feeding line, is a machine for producing the unfed.

The machine is not figurative, and it is running now. In the spring of 2026 three displacements proceeded at once across the region, each on a published deadline. In Tainan, the city resumed clearing the four-century-old Nanshan cemetery — sixty thousand graves, the oldest from 1642 — giving families weeks to remove the first batch, while preservationists gathered in the streets; on the ground, witnesses described construction crews breaking tombs open, bones left to the weather, teeth scattered like gravel. In Singapore, the National Environment Agency opened the ninth phase of its exhumation of the Choa Chu Kang Chinese cemetery, graves dating from 1965 to 1997, with claims due by the end of April and the digging to begin after June. In Hong Kong, the trade settled into the new routine of monthly lotteries for public niches, the long shortage easing at last as fresh towers came online. Three forms of the same motion — bones lifted off old ground and filed into new inventory — on three government timetables, in one season.

That the tradition has a standing institution for exactly this surplus — the seventh-month Ghost Festival, the Zhongyuan, at which the community collectively feeds the wandering dead that no lineage is feeding — is the measure of how old the problem is. The festival is the net the culture long ago strung beneath the lineage system, to catch the dead who fell out of it. The festival assumes the surplus is small. The industry is enlarging it.

V. Where the cash flows

Between the brochure and the cosmology runs the money, and following it reveals the industry's true shape — which is not a single business but a circuit of them, feeding one another.

The visible product is the niche. But the niche sits at the center of a wide ring of auxiliary trades, each with its own fee, each older than the tower. There is the geomancer, the fengshui master, retained to certify the siting — and the cosmographic vocabulary translates upward into the high-rise without difficulty: the auspicious grave wants sunlight, a mountain at its back, water before it, and no ill-omened flow cutting through, and so the auspicious niche wants a high floor with light, a mountain-facing aspect, a water view, and a position away from the disturbance of the elevator. The classical art that once sited graves on hillsides now grades shelves by floor, and does so with such seriousness that the suitability of mausoleum sites has been modeled, in a peer-reviewed geographic-information-systems study of cemeteries near Nanjing, against eight quantified topographic variables mapped to the old geomantic categories. The cosmography has acquired a methodology. It has also crossed the Pacific: brokerages in southern California advertise "auspicious Feng Shui plots" to Chinese-American families, the vocabulary preserved intact in diaspora as a marker of the identity it carries.

There is the clergy. The Daoist priest and the Buddhist monk are, as the saying in the trade might go, busy — and they are busy because the dead require ritual labor, and ritual labor is billable. The elaborate Daoist Rite of Water and Land, running seven days or more, delivers the souls of the dead and feeds the hungry ghosts of the three worlds; the Buddhist Rite for Feeding the Burning Mouths performs the parallel office; in southern Taiwan, troupes performing the Soul-Guiding Array — a folk form found, tellingly, in the same southern districts where the grave-clearances have been most intense — sing and dance the souls along their way for a fee. There is the maker and burner of spirit-money, an industry that in Taiwan alone consumes on the order of a hundred thousand tons or more of joss paper a year, with a measurable cost in air pollution that has drawn government schemes to centralize and filter the burning — schemes that one temple leader called humiliating to believers, since they relocate the offering out of the worshipper's own hands. There is the specialist in secondary burial, the bone-cleaner, whose traditional trade of exhuming, cleaning, and re-housing the bones after their first interment is precisely the skill the relocation industry now deploys at municipal scale.

And there is the layer above all these, where the niche stops being a resting place and becomes an asset. In Hong Kong, where the public columbarium waiting list runs to many thousands and the wait exceeds the wait for public housing, niches bought years ago resell for many times their original price; a broker can speak of a niche at a celebrity-graced columbarium changing hands for sums that rival an apartment's down payment, and the most expensive niches reported by the territory's Consumer Council reach into the millions. The dead, in a market of engineered scarcity, become a speculative instrument. And where there is a speculative instrument there is fraud: Taiwan's prosecutors have called scams rife in the columbarium and funeral trade, indicting dozens; one Ponzi scheme promised investors returns above ten percent on niches and cemetery plots and drew nearly ten thousand investors before its principal fled the territory — a man whose own father had been a senior financial official of the old ruling party, the lineage of state money running straight into the business of the dead. Fraud rings trade lists of those already defrauded — "lists of fat sheep" — to defraud them again. On the mainland, a nationwide investigation into funeral-industry corruption has run across eight provinces, and a separate scandal exposed a company illegally harvesting and reselling thousands of corpses. In Malaysia, in the spring of 2026, the police confirmed that urns had been stolen from the columbaria of three memorial parks by what was believed to be a cross-border ring holding the ashes of the dead for ransom — the logic of the market followed to its conclusion, the ancestor become collateral. And the probe now hanging over the listed mainland market leader is, in this light, not an aberration but a late and large instance of a pattern that runs the length of the trade.

Trace the whole circuit and a closed loop appears. The bereaved family pays the funeral parlor, the priest, the geomancer, the spirit-money seller, the bone-cleaner, and the columbarium developer; the developer, having sold the niche once, has little recurring revenue from it but the maintenance fee; the firm therefore needs continual new construction and new sales to grow, which is the structural reason the revenue curve cannot hold; and when sales falter and niches go unmaintained and ashes are displaced — when the system's failures accumulate — those failures take the cosmological form of unsettled, unfed dead, which generates fresh demand for exactly the ritual services, the placation of hungry ghosts, with which the circuit began. The industry, in other words, manufactures both the storage of the dead and, through the structural failures of that storage, the restless dead who require the rites that are the industry's other half. The hungry ghost is not a byproduct the trade would disown. It is, seen coldly, a second line of business.

VI. "We consulted the priest"

There remains the question of how a family lives with this — how the felt wrongness of taking a grandfather off his hillside and filing him in a distant drawer is metabolized, in a culture that feels that wrongness keenly. The answer is the same practical logic once more, and it is the quiet center of the whole matter.

The wrongness is not reasoned away, because in a practice-ordered world reasoning toward a principle is not the available move. It is ritualed away. The family consults the geomancer; an auspicious date is chosen; the rites are performed; the master pronounces the new placement sound — and the dread is thereby converted into correctness. The decision, almost always made by the senior male of the line, is rendered correct not by an argument that it was right but by the proper performance of the rites that attend it. And because the conversion is ritual, it can be purchased, and it is. The priest casting the divination is selling, beneath the metaphysical service, the thing the family most needs: the authoritative verdict that what was done was guāi, that they are good descendants, that the move was an act of care. The ritual fee is, at bottom, the price of the discharge of the stigma — and the more elaborate the rite, the more complete the discharge, so that the wealthy can purchase a fuller absolution than the poor, and the relief, like the niche, is sorted by what one can pay.

This is why the explaining differs by audience, and why none of the explanations is a lie. To the women of the house who keep the actual ritual life and feel the wrongness most exactly, the move is narrated as an improvement: the old grave was damp, or overgrown, or hemmed by the new road; the new niche is dry, high, bright, and better sited; the master approved it; the ancestor will rest more comfortably. To the children, who are being trained, it is the cleanest version, a lesson in correctness: we moved him somewhere beautiful, we did everything properly, now bow. To the neighbors it is told as prestige: the family could afford the premium niche, the high floor, the good fengshui — the uprooting reframed as a filial success only a prosperous house could manage. To the younger brothers, who have standing to object but no authority to decide, it is presented as accomplished fact and collective benefit at once: it is handled, the master approved, here is your share of the cost, there is room in the niche for all of us and our sons — and to object now is to insult the verdict and refuse one's own future resting place and stand revealed as the unfilial one, so the objection is swallowed. To the state and the record it is pure compliance: conducted per the municipal program, all rites observed, auspicious date selected by a qualified master, remains placed in a licensed facility. Nobody has lied. The bones are in a drawer three hundred kilometers from the hill they were the soul of, and every form has been correctly filed.

The whole truth is carried by no one, because the system is built so that no one need carry it. It is distributed across the household as residue — in the silence of the aunt who knows exactly what happened and says nothing at the table, in the second question of the child who remembers that the grandfather had planted the trees on that hill and asks whether he did not love it there, in the swallowed rage of the younger brother. And residue that the practice has no channel to discharge does not vanish. In this culture it has, for many centuries, surfaced in a particular place: the strange tale, the ghost story, the account of the dead who return because something was not made right. That is the register in which the culture says what its practice cannot. It is not the register of this essay. But it is where the rest of this matter has always been told, and it is not done being told, and the towers will furnish it, in time, with everything it needs.

VII. What the looking shows

Set the documents side by side, then, and the picture is consistent at every level — and consistent in a particular, unsettling way.

At the level of the brochure: a resting place sold as a good address, the dead moved from the register of dread into the register of property. At the level of the income statement: a captive market of guaranteed demand that has, against every expectation of its investors, turned downward, with the flagship firm in a historic loss and under investigation. At the level of the cosmology: a machine that, in consolidating the dead into distant commercial inventory severed from the resident feeding line, structurally manufactures the unfed dead the tradition has always feared. At the level of the household: a wrongness genuinely felt and ritually, purchasably, discharged, the discharge sorted by wealth, the residue distributed as silence and swallowed objection. At the level of the trade: a closed circuit in which the failures of the storage business become the demand of the ritual business, the hungry ghost a second revenue line.

No part of this requires anyone to be acting wrongly by the standards of the system. The developer sells a licensed product. The geomancer certifies by the inherited art. The priest performs the rite that is his to perform. The senior son arranges, correctly, for his father's relocation and explains it, correctly, to each who must be told. The family purchases, correctly, the discharge of its unease. Every act is guāi, correct in its context, performed without deception by the only test the culture applies. And the aggregate of all this correct action is an industry that takes the dead from the ground that made them ancestors, files them in depreciating inventory at the moment their descendants scatter, manufactures hungry ghosts as a structural byproduct, sells the placation of those ghosts as its other half, and has now, at its commercial peak, begun to falter in a way its own logic cannot read as failure — because the system has no test for outcomes-across-contexts, only for correctness-within-them.

What the late phase of such a curve looks like is not a matter of speculation; it has already happened, in the West, a century ahead. San Francisco banned new burials in 1902 and cremation in 1910 and sent its dead south to the town of Colma, where they remain, outnumbering the living many times over. One grand structure was left behind in the city: a domed columbarium of 1898, some eight thousand niches, its rooms named for the Olympian gods. Cut off from new burials, it had no recurring revenue, and for decades it passed from hand to hand — in one caretaker's phrase, like a hot potato — neglected, nearly derelict, its niches quietly traded on a secondary market that persists to this day, a glass-fronted shelf near a stained-glass window offered, as the listings put it, open to offers. That is the far end of the curve the towers are on: the inventory sold once, the descendants gone, the building outliving its business and becoming a thing changed hands rather than a place tended. The Chinese towers are newer and far larger, and their demographic window is closing faster. The end state is on the map already.

That is the finding, and it is offered as a description rather than a verdict, because the description is the more damning thing and needs no help. The tradition that built the most durable apparatus in the world for keeping the dead in their place is now operating, at industrial scale and for profit, a machine for taking them out of it — and doing so, at every step, correctly. The grandfather kept the time of his hillside: the seasons, the plantings, the festivals, the offerings, the slow agrarian clockwork that bound a line to a place across generations. The tower keeps a different kind of time — the quarterly kind, the kind an income statement keeps — and the two have, for a moment, in the brochure, been made to look like the same thing. They are not the same thing. The distance between them is the subject of this essay, and the measure of what is being lost is simply the distance itself, which the documents, looked at plainly and set side by side, are enough to show.

A note on sources

The financial figures are drawn from the public filings and reporting on the two listed market leaders — the mainland firm's 2018–2022 revenue and net-profit series, its first-half 2025 results and the subsequent delay of its audited annual report, and the Taiwan firm's market performance — as carried by Hong Kong Stock Exchange disclosures and financial press through early 2026. The orthopraxy argument follows James L. Watson, "The Structure of Chinese Funerary Rites" and "Rites or Beliefs?", with the qualifications entered by Donald S. Sutton and others in the 2007 reassessments; the gods/ghosts/ancestors framework follows Arthur P. Wolf, "Gods, Ghosts, and Ancestors" (1974), with Steven Sangren and C. Stevan Harrell; the feeding mechanism draws on Stuart Thompson's "Feeding the Dead" and the essays gathered in Watson and Rawski, eds., Death Ritual in Late Imperial and Modern China (1988). The Hong Kong niche market and the Private Columbaria Ordinance, the Taiwan grave-relocation programs and fraud prosecutions, the mainland corruption investigations, the joss-paper environmental measures, the fengshui-GIS study of Nanjing mausoleums, the Daoist and Buddhist deliverance rites, and the diaspora brokerage language are documented in the regional reporting and scholarship gathered for this study. The spring-2026 developments — the resumed clearance of Tainan's Nanshan cemetery, the ninth phase of Singapore's Choa Chu Kang exhumation, Hong Kong's shift to monthly niche lotteries, and the Malaysian urn-ransom case — are drawn from regional press of that season; the San Francisco Columbarium's history follows the standard accounts of the Odd Fellows cemetery, the 1902 and 1910 municipal bans, and the building's later passage between owners. Specific citations are held in the working file and available on request. The personal observation underlying the closing belongs to the author, and to one family, and to one hillside that is now a drawer.