The Sovereign Family

Why now

A fifth of the world's assets belong to people over seventy-five and will change hands within a generation. Roughly US$90 trillion accumulated by the baby boomers and Generation X passes to heirs over the next twenty to thirty years, the largest transfer of wealth on record.

The vehicles built to receive it are for the most part unprepared. Single family offices, some eight thousand of them, manage about US$3.1 trillion and are expected to manage US$5.4 trillion by 2030; in a 2026 survey of 333 of them, 86 percent had no succession plan for their own decision-makers and 68 percent no family council. In the rich world, private wealth stands at five to six times national income while net public wealth sits at or below zero, a position last seen in the eighteenth century. Governments tax the flow of income. The stock is in families, and it is about to pass to people who did not build it.

Many of those people were not raised in the enterprises they will inherit. Research among business families finds founders searching for a single successor in their own image, leaving the question to the last moment, and heirs never brought into the business seeking a liquidity event to leave it; in most of these families the wives and daughters take no part in strategic decisions at all. When a founder dies without a structure, the enterprise divides among heirs, each heir is an individual taxpayer under one jurisdiction, and custody of the capital passes to that jurisdiction through the ordinary operation of succession law, or to an asset manager, BlackRock, Vanguard, Blackstone or a peer, the waiting room in which wealth sits until an heir arrives to spend it or a government to tax it. The estate planners who manage the passage are notaries at a change of custody.

Jurisdiction planning has always mattered; the transfer makes it urgent. The capital now moving goes either into family structures or into individuals, from whom it is taxed, spent or parked, and which of the two happens depends on jurisdictions that most families never chose. They accumulated, account by account and passport by passport. Jurisdiction planning is the practice of choosing them.

Preface

This document sets out the purpose and method of jurisdiction planning and the standards its practitioners hold themselves to. It is published by the Journal for Jurisdiction Planning and maintained by the Association for Jurisdiction Planning, whose members subscribe to the articles below.

International financial and legal services have reached a stage at which jurisdiction planning has become necessary. Financial services produced the multi-family office when families needed an advisor who sold no product and could weigh the banks against one another; residence, citizenship, banking and structuring have arrived at the same point. States are run for the short horizon and families for the long one, and a family whose whole position sits under an institution run for the short horizon is exposed to it. Both fields have competent professionals who weigh jurisdictions within their own discipline: private bankers who choose where an account is booked, private-client lawyers who choose where a trust is settled, tax counsel who choose where a company is domiciled. What was missing was a professional who addresses the choice of jurisdiction directly, and who considers not only the services that depend on it but the jurisdictions themselves: what each one is, and what choosing it entails for a family in liabilities and in opportunities. That professional is the jurisdiction planning advisor. He forms his advice by consulting the legal and financial professionals licensed in each jurisdiction in question, and gives neither legal nor financial advice himself.

The document is written for heads of families and for the professionals who serve them. Of those professionals it asks one thing: that they take no gain at the expense of the family they serve.

The Journal publishes it in the belief that families which plan across generations pay for what takes longer than a lifetime to pay back, the hospital, the school, the company that outlasts its founder, and that every jurisdiction is the better for holding them.

Part One · Articles of the Association for Jurisdiction Planning

Article 1 · Scope

Jurisdiction planning is on the side of the family.

It represents no jurisdiction and sells no programme, account or structure. It advises the family on which jurisdictions to hold and in what combination, and it is paid for that advice and for nothing else.

The market in international financial and legal services is a market of providers. A bank offers its accounts, a migration firm the programmes it is licensed to sell, a trust company the structures it administers, a law firm the law it is admitted to practise. Each is competent in its jurisdiction and has no view beyond it, and none is asked whether the family should be in that jurisdiction at all. Most families answer that question by accident, or by whichever provider spoke first.

The family is the party that chooses and the jurisdiction the party chosen, and the two plan on different horizons. A state organizes the population of a territory, most of whom plan within a lifetime, and its leadership plans to the end of its term. A family organizes a lineage; it plans beyond a lifetime or it ends, and every person a state governs or a company employs was raised by one, which makes the family the only institution that produces its own members. An enterprise plans as far ahead as its owner does: a family-owned enterprise is planned for as part of the family, and an enterprise owned by no one in particular, dispersed among shareholders none of whom can direct it, takes the horizon of the managers who hold it for a term.

Jurisdiction planning does not take individuals as clients. The individual is the unit into which families dilute, and the counterparty every institution prefers: a state taxes him, a company hires him, a bank holds his account, and none of them needs to negotiate. An individual acquiring a passport acquires a document of convenience, and no jurisdiction wants him. A family acquiring a citizenship enters a covenant with a jurisdiction on behalf of people not yet born, and jurisdictions want that family, because it is the only party whose horizon is longer than the government's. The first question put to a prospective client is whether he wants this for himself or for his family.

An advisor who also sold the programme, the account or the structure would have a stake in the answer, and the family would be back where it started. The value of the advice is its independence, and the practice keeps that independence by having no product to sell.

Article 2 · Curation

Jurisdiction planning curates and arbitrates only.

Its clients are families and family-owned enterprises. For them it chooses, among the services, programmes and products that jurisdictions and their advisors offer, the state memberships and registrations the family will hold, and it gives no legal, tax or financial advice in any jurisdiction, including one in which the advisor is himself licensed. Professionals licensed in each jurisdiction execute what the curation decides.

A jurisdiction offers a family a passport or a residence permit, a tax regime, a body of company and trust law, courts, schools, hospitals, and the protection of its police and its army. Each comes with obligations, and each is offered through advisors who represent that jurisdiction whether or not they say so, because they are licensed in it and practise in no other. Curation is choosing among these offers on the family's behalf. It assesses a jurisdiction on what it offers a family across generations before any route into it is considered, and it counts every lawful route equally: a programme is one door, and ancestry, marriage, incorporation, residence and property are others. Arbitration is deciding between jurisdictions when their offers conflict, which they do: a residence that suits the family's tax position may not suit its succession, and the jurisdiction that suits its company may not suit its children.

Membership and registration are curated together because they are governed together. Where the family's people are resident and what passports they hold decides which law reaches the family's companies, holdings and trusts, wherever those are domiciled. A registration chosen without regard to the family's memberships is a structure built on a foundation nobody examined.

Execution is excluded because knowledge of one jurisdiction distorts the comparison of several. An advisor licensed in one jurisdiction knows its law better than the law of the jurisdictions he compares it with, and the knowledge pulls the comparison toward it. The article removes the pull by dividing the roles. The advisor compares, having consulted the professionals licensed in each jurisdiction in question; those professionals implement what the comparison decides and confirm that it is lawful there; and the confirmation is recorded in the plan, so that for every element of it the family knows who stands behind it and under which law.

Article 3 · Fiduciary Duty

Jurisdiction planning advisors should never gain at a client's expense.

The advisor acts in the family's best interest, discloses in full every payment he receives from a third party, and lets no payment decide a recommendation. Where his interest and the family's diverge, he says so before the family decides.

A plan is advised because it improves the family's position. The advisor charges for the plan and the family pays providers to execute it, so a plan is advised only when the family is better off with it than without it after every fee is counted. Where the balance is against the family, the advisor says so and advises against the move. A curation that never recommends staying is a sales pitch, and the advisor believed when he says stay is the one believed when he says go.

The family's other advisors are already held to a fiduciary standard in the jurisdictions that regulate them, and a practice that claims to sit above those advisors cannot be held to less. Providers remunerate advisors throughout this market, and the article does not forbid it; it requires that the family see every such payment in full and that none of them decide which jurisdiction, provider or route is recommended. The undertaking a member gives each family under these articles is set out in the Association's Charter.

A practice that asks families to be independent of their providers cannot itself be dependent on its own.

Article 4 · Diversification

Jurisdiction planning diversifies between jurisdictions.

Diversification is the economic principle that no single exposure should be allowed to decide the outcome. Applied to jurisdictions, it means that no single jurisdiction should hold a family's residence, citizenship, capital and structures at once.

In the sense the word has in finance, a jurisdiction is an exposure: one legislature that can change the tax, one court system that will hear the succession, one currency and banking regime, one set of alliances that decides whose sanctions apply, and one government whose decisions reach every asset the jurisdiction governs, wherever the asset is kept. A family whose residence, citizenship, capital and structures all sit under one jurisdiction has concentrated its entire position in a single exposure, and most families of consequence hold that position without knowing they hold it.

The principle governs the practice from the first step. A curation begins by measuring how concentrated the family's position is and ends by reducing the concentration: residence, citizenship, capital and structures placed across jurisdictions chosen so that no single one, and no alliance of them, can reach the whole. Tax is the symptom families notice; reach is what diversification treats. The Armenians of New Julfa had no favourable tax regime in Isfahan; they had agents in Madras.

The principle is also why jurisdiction planning comes before wealth, tax and legal services rather than after them. The jurisdictions a family holds decide what those services can offer it: which tax regime applies, which trust and company law is available, which banks will open an account, which succession rules govern the estate. A wealth plan, a tax plan or a legal structure built before the jurisdictions are chosen is built on whatever jurisdictions the family happened to have, and each of those services changes substantially once the jurisdictions change.

A family that has diversified fully has no superior. Sovereignty is having no superior. No state has it alone: a state is sovereign because other states recognize it and no single state can compel it, and a family attains it by the same means, recognized by several jurisdictions and compelled by none. The fully diversified family is the standard every plan is measured against: a family that guarantees its own security, prosperity, justice and posterity without depending on any single jurisdiction or institution; that is lawful in every jurisdiction it chooses and absorbed by none; and that stands as protector of the citizens and counterweight to the leadership of each.

Part Two · The Laws of the Sovereign Family

Law 1

A sovereign family holds its citizenships by choice.

A child is made a citizen at birth by the place of the birth. Tax residence, liability to military service and the law of the land come with the citizenship, and no one is asked whether they accept them, neither the parents at the time nor the child later. A family with a single citizenship has complied with it because compliance was the only course open. A sovereign family is one whose citizenships it could have declined and did not.

Consent to a citizenship requires that the family could have declined it. A party that cannot refuse has not agreed, whatever it signs, and a family's relationship to its jurisdiction changes the moment it holds a second one: every obligation it keeps after that is an obligation it chose to keep. Hirschman described the mechanism in 1970. A member of any institution has three options, exit, voice and loyalty. Voice, the ability to change the institution from inside, is credible only where exit is possible, and where exit is impossible what remains is loyalty, which is the name the institution gives to the member's captivity. A family with one jurisdiction believes it has a voice there because of what it employs, pays and builds. It has as much voice as its ability to leave.

Hume made the objection to birthright consent in 1748: a man born on a ship, who cannot swim and knows no other vessel, cannot be said to have consented to the captain. For most of history the objection had no remedy, because allegiance was perpetual. A subject could not lawfully change his sovereign, and the attempt was treason; the United States and Britain went to war in 1812 partly over Britain's refusal to accept that a British-born sailor could become American. The right to change allegiance was won in the nineteenth century, treaty by treaty, and the right to hold more than one allegiance was tolerated broadly only in the last decades of the twentieth. The freedom on which this law rests is younger than most of the people who hold it, and most of them have never used it.

A citizen by birth is a conscript until he holds a citizenship he could have declined. The military sense of the word is intended: a state that can tax a person can also conscript him, and a person should at least be able to choose the regime he would be sent to die for. A family with several passports to regimes it would not defend has multiplied its conscription. The choice is therefore made by constitution rather than convenience. A sovereign family acquires no citizenship in a jurisdiction whose constitution it does not accept at least in spirit, however flawed the government of the day, because governments pass and the covenant remains; and when a government fails the family, the family withdraws its capital and keeps the covenant. A family that renounces a citizenship over a bad decade has thrown away the position from which the decade could have been outlasted.

Jurisdictions now compete for families as they have not since the sixteenth century. What they want is the covenant a family enters on behalf of generations, since a family is the only party whose horizon is longer than the government's; an individual with a passport is a document of convenience to them. A family that reads its citizenships as accidents of birth is a set of taxpayers. One that reads them as covenants it chose, and can withdraw from, is sovereign, and jurisdictions can tell the difference.

Law 2

A sovereign family decides its jurisdictions before it decides anything else.

Every plan a family makes assumes a jurisdiction. A tax plan assumes the tax law of somewhere; a will assumes the succession law of somewhere; a company assumes the courts that will enforce its contracts; a child's schooling and a parent's treatment assume the schools and hospitals of somewhere. Each plan is made by a competent professional who takes the jurisdiction as given, and each fails at the border when the assumption turns out to be wrong.

The jurisdictions come first because of custody. Custody of a family's assets follows the jurisdiction that governs their owner, not the place where the assets are kept. A family with accounts abroad and legal status at home has its accounts in the custody of home, because home governs the owner: it taxes his worldwide holdings, receives the existence of the accounts through automatic exchange, applies its succession law to them at his death, and can order him to repatriate them under penalties the foreign bank will never see. The vault protects the asset from theft, and from nothing else. The apparatus built over the last fifteen years, automatic exchange of information, beneficial-ownership registers, the reporting of controlled foreign companies, exists to attach every asset on earth to one natural person and that person to one residence.

Residence is therefore the variable, and it is acquired, counted and lost by rules that differ from one jurisdiction to the next: days present, a home available, the centre of family and economic life, and in some jurisdictions the holding of status alone. Many jurisdictions grant residence status without requiring residence, and a family registered as resident in one of them is governed by its law for the purposes of every bank that holds the registration; the registration must be substantiated to hold against a home jurisdiction that disputes it, and the substance is what a plan provides. Citizenship governs alongside residence, since some states tax their citizens wherever they live and most apply succession law by nationality or domicile, so a family can change its residence and remain governed. The governing jurisdiction may also not be the one the family sees. A home jurisdiction that depends on a foreign government for its security, its credit or its markets takes direction from that government in matters it does not disclose, and the direction reaches the family as domestic decisions: a regulation, a tax measure, a change in how contracts are enforced. Such a family is held at one remove by decision-makers it has never assessed and would not choose. In every one of these cases custody is changed by changing who the owner is and where he is registered, and never by moving the assets.

The family networks that governed trade before the nation-state, the Portuguese Nation in Amsterdam, the Genoese houses, the Armenians of New Julfa, all had a body above their bankers that decided where the family stood before the bankers decided what it held. That layer disappeared when states absorbed the families, and for a century families have been served from the provider's side. Jurisdiction planning restores it.

Most families arrive at this point with a succession plan. A succession plan is a document about who gets what and how the tax is minimized; it moves money from one taxable individual to several with the least friction, so that the money stays where the jurisdiction can reach it, and it ends with more citizens and less family. Founders choose it, or a foundation, out of a settled belief that a structure which keeps the family together is not possible, that money ruins children and dynasties always fail; the best of them then endow for strangers the structure they could have built for their own.

In practice the owner of a sovereign family's assets is, wherever possible, an Independent Family Structure rather than a person under one jurisdiction: an architecture that holds the family's capital, decisions and formation independent of any single jurisdiction, advisor, trustee or vehicle, so that none of them can capture the whole and each can be replaced without interrupting the family's custody of itself. Its trustees answer to the family's constitution, and to neither the founder nor the heirs; a rich family has a lawyer, and a sovereign family has a constitution and someone whose job is to enforce it against its own members. Its core capital is held for a purpose rather than for owners, as the endowments of the Portuguese Nation were, so that an heir cannot sell what he does not own and can only be included in or excluded from its benefit; no other arrangement prevents a liquidity event, because every other arrangement is a conversation. The great structures of the nineteenth century held the fortune in the house rather than in any member and spread it across five capitals, and remained dependent on one bank, one law firm and one trustee, through which they were eventually captured. The independent structure settles its own disputes in a tribunal it has constituted, seated where the tribunal's awards are enforceable, which under the New York Convention means in the courts of over one hundred and seventy states. A family whose final arbiter is its own tribunal, enforced by courts it does not depend on, has custody of itself.

The same rule governs the family's enterprises. Capital is held through vehicles, and the custodian of the vehicle is the custodian of the capital: whoever holds the company holds the money and gives it a planning horizon. A state that taxes companies rather than individuals, and invites the world's companies to domicile under its law, has understood this and is taxing other nations' citizens through their vehicles; a sovereign family uses the rule in the same direction and holds its vehicles where the vehicle is treated as a guest.

Every family already has a jurisdiction map, drawn by no one: an account opened where the business was, a company registered where a lawyer suggested, a passport inherited, a residence bought after a good holiday. Jurisdiction planning turns the accidental map into a designed one, as estate planning once turned inheritance from an event into a plan, and it does so first, because the four disciplines of family building all stand on the jurisdictions the first one chooses. Jurisdiction Planning keeps the family's security. Treasury Planning keeps its prosperity, and takes wealth and tax as one office, since a treasury does tax, investment, currency, credit and liquidity as one function, and an office, unlike a sum of money, can only be kept well or badly. Family Governance keeps its justice: the constitution, the council and the protocol that settle disputes inside the family before any court sees them. Generational Planning keeps its posterity, and is the subject of Law 4.

Law 3

A sovereign family holds residence or citizenship in at least two jurisdictions.

A single jurisdiction cannot have been chosen, since there was nothing to choose it against. A family with a single jurisdiction has no alternative to it and therefore no consent to it, no voice within it and no protection from it. Two is the minimum at which choice, voice and safety become possible.

Four facts about institutions make the second jurisdiction necessary. An institution plans as far ahead as its custodians do: a caretaker to the end of his term, an owner to the end of his life, a family to the next generations, and the family is the only institution with a built-in reason to protect the generations after it, needing no ideology to do so. The record bears this out: family-controlled companies carry less debt, invest more patiently, outlast their peers and keep their workers through downturns when public companies cut, and most chief financial officers of public companies say they would sacrifice long-term value to meet a quarterly target. Short-term planning consumes what long-term planning saved, and when the savings run out the short-term custodian takes what he can reach from the long-term ones; this is why states in debt tend to confiscate, and heirs in desperation tend to liquidate without restraint, the structure being the restraint. Everyone is born a short-term planner, and long-term planning is learned, so the majority is always short-term; the state is the institution that serves it, selling it the present in the form of the pension, the benefit and the subsidy, and any regime chosen by majorities moves toward short-term planning whatever its founding intent. And numbers decide the use of force: a short-term majority in debt has both the incentive and the means to take from a long-term minority, and given time it will.

The short horizon is the human condition under scarcity, and the long horizon is a duty because it is rare; short-horizon members arise in every family too, by temperament, illness or misfortune, and the family looks after them as it looks after the citizens of its jurisdictions, provided for and capped, without a hand on the principal. A family that plans long is therefore a minority in every jurisdiction, with something to lose and no way to win a contest with the majority at home. Resistance at home fails; the family's protection is positions the home majority cannot reach, and every resistance that succeeded had one. The barons of Magna Carta held castles the king could not take, and William of Orange ran the Dutch revolt from Dillenburg, outside Philip's reach. The second jurisdiction is also the condition of the family's own lawfulness. A family under one jurisdiction, when that jurisdiction fails, is forced to choose between its members and the law, and no family that protects its members chooses the law under extreme pressure; a family with other jurisdictions is never forced to choose. Far from being a way around the law, jurisdiction planning is what allows a family to keep it.

The last time trade outran the reach of states, between roughly 1450 and 1700, families held the space between them, because protecting capital had come to cost less than seizing it. State power was fragmented, trade routes ran thousands of miles beyond any state's writ, treasuries were empty and borrowed from private houses, expulsions had scattered peoples into networks that could enforce a contract where no court could, and cities wrote charters to attract them, beginning with the Livornina of 1593. Six peoples of several faiths arrived at one design. The Portuguese Nation, from Amsterdam to Recife, was governed in each city by a council that held communal funds as a trustee, and its dowry endowment of 1615 paid for marriages across the whole diaspora on condition they were made inside it. The Genoese financed the Spanish crown and were governed at home by the Casa di San Giorgio, a private institution that held the republic's debt and at times ran its colonies. The Fuggers bought Charles V his crown in 1519 and held mines from Tyrol to Peru. The Armenians of New Julfa, a single suburb of Isfahan, ran silk and gems from London to Manila on kinship and a court of their own. The Hanseatic houses kept Kontors in London, Bruges, Bergen and Novgorod with their own law and the power to expel, so that a merchant finished in Bergen was finished in Lübeck. The Ismailis alone were never absorbed, and still govern themselves under a constitution of their own, with conciliation and arbitration boards for commercial, family and inheritance disputes. Each of the six was multi-jurisdictional by design, governed by a trustee institution rather than by any state, bound to purposes no heir could liquidate, able to expel, based in cities, and, in the case that survived, strictly neutral toward every host. Each held three kinds of position: a jurisdiction that recognized the family and protected its property, open cities where its capital was run and its sons were tested, and refuges under a different protector. The Nation held Amsterdam, Livorno and Curaçao; the Rothschilds later held London, the continental capitals and their own countryside. The cities that held these families were the cities of Rembrandt and Spinoza, Brunelleschi and Galileo.

The networks were absorbed rather than defeated. From the late seventeenth century states learned to borrow from their own publics, through the Bank of England and a permanent, tradeable national debt, and no longer needed the houses. The houses that had lent to crowns were ennobled, the Fuggers as counts, the Genoese as Spanish grandees, the Medici as grand dukes, the Rothschilds as barons in five countries, and the houses that stayed mercantile declined. The joint-stock corporation then supplied scale without lineage and the nation-state supplied legitimacy without family, and the twentieth century became the only period on record in which the dominant unit of wealth was not a family. The absorbed lost in two ways. The Fuggers became indispensable to one sovereign and could not leave him. The Nation concentrated in Amsterdam and could not move whole when Amsterdam's charter narrowed.

Every condition of 1500 has returned. Jurisdictions compete openly for families again; value chains run through digital platforms whose terms bind more merchants than any national commercial code; public debt exceeds the size of the economy across the rich world, and governments court private capital for what their treasuries cannot fund; a new diaspora of multi-passport families spans a dozen jurisdictions; and law is plural once more, with common-law islands inside civil-law states, freeports, private arbitration and competing trust regimes. Three things have no exact precedent. Bearer digital assets can cross a border in a person's memory, and though the state has drawn much of them back into regulated custody, the exit they offer disciplines every jurisdiction's conduct. Artificial intelligence makes capital productive without a national workforce, which breaks the bargain the twentieth-century state was financed on, the taxation of labour in one place. And the state has lost its monopoly on the story its citizens hear. States are becoming more important and less able to govern capital at the same time, which was the Habsburg position in 1550, and, as was predicted in 1997, they have turned predatory before they have weakened. The networks of 1500 were born of expulsion and new routes, not of inheritance; the modern return has a trigger the old one lacked, the largest generation of builders on record dying without having decided what it leaves.

In 1600 the winners were cities, not countries: Amsterdam, Livorno and London, not the Netherlands, Tuscany and England. Venice was founded by families fleeing Attila into a lagoon no state wanted, and was ruled by its merchant houses for eleven hundred years. Antwerp received the merchants of Europe, was sacked in 1576 and closed to trade in 1585, and sent its families north. London was a chartered free zone inside its kingdom before there was a Parliament, with its own courts and militia, the Lombards on the street named for them, the Hansa in the Steelyard, the Jews readmitted in 1656 for openly commercial reasons and the Huguenots after 1685; in 1688 the crown's finances were put under Parliament, and the Bank of England followed within six years. New York was founded by the Dutch and the Nation together, the first Jewish families arriving from Recife in 1654. The founding acts of the open jurisdictions were refusals of confiscation: Rütli against Habsburg bailiffs in 1291, the Dutch revolt against the Tenth Penny, a sales tax, and the Boston revolt against a tax on tea. The modern cases follow the same script. Dubai's rise began in 1902, when its ruler abolished customs duties to attract the Persian merchant families that a new tariff had driven out of Lingah. Singapore was founded as a free port in 1819 and refounded in 1965 as an island without resources whose one export was the enforcement of contracts. Miami lay under a Florida constitution that banned an income tax from 1924 and protected the family home from creditors from 1868, received its merchant class from Cuba in 1959 and from every Latin American seizure since, and within five years of a 2017 federal tax change that made the costs of the older centres bite in full, the largest funds and banks of those centres had opened there. Charter, diaspora, trigger, in that order, made Amsterdam as well. The jurisdictions that seized families decayed, Spain after 1492 among them, and the founding documents tell the two kinds apart: the Livornina, the Act of Abjuration and the Declaration of Independence speak of property, courts and consent, and the states that seize speak of purity, unity and the people. The founding memory of the open jurisdictions fades in about two centuries; the most far-reaching extraterritorial tax law ever written is administered by the heirs of a tax revolt.

In a designed plan the minimum of two memberships becomes three positions, and a family holds a portfolio of jurisdictions as it holds a portfolio of assets. The Vault is the jurisdiction where the family's rights and property are guaranteed: trust and foundation law, testamentary freedom, enforceable family constitutions, property beyond the reach of any leadership's discretion. It is the jurisdiction whose power stands behind the family, it need not be lived in, and full compliance is owed to it without exception; the word is used in its plain sense, since a family is entitled to a vault, and in the vocabulary of the market it is a family-first regime. The Forum is where the family's capital is run and its next generation is tested against the world's other capable families; a family may hold several, and a Forum may be strong on prosperity and weak on posterity without disqualifying itself; in the market's vocabulary, a business expansion hub. The Refuge is where the family's life holds if the Vault fails or the world turns: non-aligned, self-sufficient in food and energy, distant from every theatre, and under a protector other than the Vault's; in the market's vocabulary, a chosen home, a retirement, a plan B.

The Vault and the Refuge must not share a protector; a Vault and a Refuge under one security guarantee count as a single position. The Rothschilds kept houses on both sides of the wars of 1815 and 1870, London against Napoleon's Paris in the first and Paris against Prussian Frankfurt in the second, and lost neither; the Nation held Protestant Amsterdam and Catholic Livorno. Today the Gulf's common-law islands are the strongest candidates for a Vault, with three risks a plan must price: a charter granted by a ruling family, security resting on an external guarantee, and no settled path from resident to member. The American trust states, South Dakota, Delaware, Wyoming and Nevada, offer the most protective trust law in the world inside the strongest state in the world, whose extraterritorial tax law is written by the same legislature; a Forum and a Vault, never a Refuge. Singapore is the Asian Forum, and the free port it succeeded in that role showed that a charter holds only as long as the leadership consents to be bound by it. Switzerland has not been invaded since 1798; its bank secrecy ended with automatic exchange, and what remains is a vault in the literal sense and the natural seat for a family's charter and tribunal rather than for its operating capital. The small European havens, Liechtenstein with the best foundation law in Europe and no means of defending anything, Luxembourg and Jersey with charters written elsewhere, are instruments inside a structure, not positions. The true Refuge is non-aligned, and the families now seeking one are finding it in large, self-sufficient, neutral countries under no one's umbrella. Because every jurisdiction drifts toward the short horizon, curation is continuous: each position is reviewed on a fixed cycle, and every review is dated.

A family distributed across jurisdictions without a governance of its own is more exposed than one that never left home. A family inside one jurisdiction is at least protected by that jurisdiction's courts; a family spread across five with no constitution is protected by nothing, and any member with control of the accounts can take the assets hostage and dare the others to find a court with jurisdiction over the whole. What makes the distribution safe is an institution between the jurisdictions, and it is built by several houses rather than one, because a single lineage can be well advised but cannot be sovereign, while several houses under shared trustees can pool risk, run credit on reputation, and accept arbitration because exclusion from the only market that matters is the alternative. The institution has a constitution no single family owns and every family subscribes to; a registry of houses cross-certified between nodes, as the Nation's councils cross-checked their books between Amsterdam and Livorno; shared trustees; a tribunal; an academy in which the heirs of every member house are formed as owners together, so that cadets who trained in the same cohort become, twenty years later, trustees who trust each other; a mutual fund, contributed to in proportion to each house's weight, for members facing evacuation, persecution or an asset freeze; admission by the sponsorship of two houses and a vote of the council; screening at entry and at intervals; and expulsion of any member who breaks the law of a host, since the network's standing is the shared asset of every house and is only as clean as its least careful member. An institution that cannot expel cannot govern, and a member finished in one node is finished in all of them. No advisory practice holds a vote in it, and its members remain free to use any advisor; an institution owned by its advisor is a sales channel. Nothing in the market provides any of this: banks belong to one jurisdiction, asset managers hold money and not families, family offices are treasuries without constitutions, trust companies serve one settlor under one law. The design has existed, in the Nation's councils and in the Ismaili constitution, and the modern models for its parts are the shipowners' mutuals, twelve clubs that compete and still share each other's losses and cover about nine tenths of the world's ocean-going tonnage; the Order of Malta's standing among states; and the Ismailis' architecture. The network provides, by other means and with no land, the four things a jurisdiction owes: the mutual fund keeps its members' security, the credit among them their prosperity, the tribunal their justice, the academy their posterity. In the old sense it is a good king with no land, and for that reason it can rent a state's protection without becoming a state.

More than fifty countries run investor programmes, and the ones that granted approval in weeks two years ago now take months or have closed. Families buying programmes off the shelf are buying doors to houses they have not inspected. The second jurisdiction is a position, chosen for what it holds, and the door is found afterwards.

Law 4

A sovereign family deals with every jurisdiction as an equal.

The relationship between a sovereign family and a host jurisdiction is reciprocal, as between sovereigns: the family owes the host its duties to the extent that the host keeps its obligations to the family. Sovereigns measure one another on these terms, and none may lord over another; between sovereigns the attempt is an act of war. The family answers as a sovereign answers, with the letter of the law and departure.

Every jurisdiction holds its families on a mandate, whether or not it remembers granting one. The families that plan across generations hold the capital that outlasts a term and raise the citizens who will inherit it; a jurisdiction that loses them keeps its majority and loses its future. Hobbes, the theorist of absolute sovereignty, wrote that the subject's obligation lasts as long as the sovereign's power to protect him, and no longer. A family that owes duties to a jurisdiction which keeps none toward it is in custody, whatever the arrangement is called.

This is the orthodox position of the Western legal tradition. Aristotle distinguished the king, who rules for the good of the whole, from the tyrant, who rules for himself. Aquinas, following Augustine, held that an unjust law is no law; John of Salisbury argued in 1159 that a ruler who breaks the law he is sworn to has ceased to be a ruler; the Vindiciae contra tyrannos of 1579 made the argument a political programme, and the Dutch nobles applied it in 1581, deposing Philip II for breach of covenant in the Act of Abjuration. A sixteenth-century forgery gave the Aragonese nobility an oath to their king, that they, who were as good as he, would accept him provided he kept their liberties, and if not, not; it spread across Europe because the nobles who repeated it already held the doctrine. Magna Carta was barons dictating terms; the Hungarian Golden Bull of 1222 wrote a right of resistance into a constitution. Absolutism suppressed all of it, and Versailles was built to turn the families that had granted the crown into courtiers who had forgotten they had. Montesquieu wrote that without intermediate bodies a monarchy is a despotism; Benjamin Constant turned the intermediate bodies into a moderating power, which Brazil's constitution of 1824 then wrote into law. That power was never the emperor's. It belonged to the bodies between the ruler and the people, and when they were dissolved into individuals it passed to no one.

What a host owes a family has to be measured, or the law becomes a licence for grievance. The Journal's measure is four conditions, Security, Prosperity, Justice and Posterity, each with a breaking point at which the host has failed: Violence Outbreak, Economic Collapse, Lawlessness, Demographic Exhaustion. The four are old. The Circle of Justice, carried from Persia through the Ottoman mirrors for princes, held that there is no kingdom without an army, no army without a treasury, no treasury without prosperity and no prosperity without justice; the conditions fail together, and the Jurisdiction Success Index, which scores every jurisdiction on the four, is built so that a catastrophic failure on one cannot hide behind good scores on the other three. A family may adopt a measure of its own, provided it has one and applies it before the grievance rather than after; when a host fails a condition, the family does not remain in it. The same measure applies inside the family. A head of family holds his office from the elders on the same terms, and loses it when he fails the same conditions for the members.

The counterweight a family provides to a jurisdiction's leadership is exercised by three means, exit, audit and service. The barons at Runnymede, the nobles of 1581 and the seven who invited William in 1688 are the history of the mandate; a family that declared itself the overseer of rulers today would be a rival, and rivals are destroyed or absorbed. The modern moderating power is a family that leaves when a condition fails, publishes what it measured, and protects the citizens in the condition the leadership abandoned, as families have always done: financing defence when security failed, as the Genoese did for Spain's armies and the Rothschilds for Wellington's; credit when prosperity failed, as the Rothschilds did for the Bank of England in 1825 and Morgan for the American banks in 1907; migration when posterity failed, as Edmond de Rothschild did for the agricultural colonies of the 1880s, Baron Hirsch for the settlers of Argentina and the Aga Khan for the Ugandan Ismailis in 1972. The only sanction it applies is the departure of its capital. It no longer finances what it does not consent to, and it keeps the covenant.

Where a host's own leadership breaks the law, the family keeps it. A jurisdiction is lawless when its own institutions fail the test its own law sets: when courts do not bind the state, rules change retroactively and property is taken outside the law. The reading comes from the Index. In such a jurisdiction the family owes the letter of the law, because the purpose of a law can no longer be identified with a leadership that has abandoned it, and it holds the state to its own constitution, which in most jurisdictions already forbids confiscation, retroactive taxation and the taking of property without process. It reports what the law requires and volunteers nothing; the duty is defined by the law and not by the administration's appetite, and an indebted administration treats what is volunteered as the starting point for what it asks next. The families who survived the last lawless age teach the method and its cost. They complied with the letter of the demand made on them and refused its intent, and the hidden layer they kept became the lever their persecutors pulled. The family that stayed in one place under one name lost its memory in a mountain town; the families that stood lawfully in two jurisdictions built the credit of a republic and the trading class of a new city. The modern improvement is distribution in place of secrecy. A family visibly lawful in several jurisdictions cannot be held hostage in one and has nothing for an inquisitor to find. Nor is any of this evasion. Every regime a family uses was written by a state that wanted it used; a flat tax, a trust statute and a charter island are laws, not loopholes, and a family that uses them does what the Livornina invited merchants to do in 1593. A family whose whole life is in a jurisdiction that has failed does not leave outright; it complies fully where it now stands, keeps a lawful presence in the country it came from, and waits until a leadership arises or helps one to. What it never does is let its survival depend on breaking a host's law, because that is the hostage position and the shortest road to absorption.

Hosts breach the covenant under a banner, and the banner is never invented; it is one the short-term majority already salutes. The oldest form is seizure of the vehicle that capital moves through. Philip IV arrested the Templars in 1307 on charges of heresy and absorbed the largest financial network in Christendom, to which he owed a fortune; Henry VIII dissolved the monasteries in the 1530s in the name of reform; Portugal, France and Spain expelled the Jesuits between 1759 and 1767 and took their estates; the democracies of Western Europe nationalized coal, steel, rail, banking and energy between 1945 and 1975 in the name of the public. Indulgences were a fiscal instrument dressed as salvation, and the Reformation began as a complaint about a moralized tax. The Inquisition made its tribunal self-financing and its neighbours its investigators. The modern forms are quieter. Every emergency enlarges the state's reach, and the reach recedes only in part when the emergency passes; Higgs documented the ratchet in 1987. Companies answer increasingly to regulators and to the proxy votes of the funds that hold their shares rather than to the families that founded them. Forfeiture without conviction funds the enforcer in the name of crime; emergency procurement moves public money to suppliers under a vocabulary no one can oppose; private raters whose criteria a state then writes into law, and the withdrawal of banking from whole categories of lawful clients that regulators themselves call de-risking and warn against, impose costs on owners and confer benefits on administrators. Two instruments will decide the next round: a bureaucracy run by machines applies rules at scale without discretion and never loosens, and the same ledger technology yields either a currency issued by a central bank that records every transaction or a bearer asset no state can read. The vocabulary changes and the balance sheet does not. No banner is immune, including the ones raised against the state, because a movement that promises good leadership must court the same majority to win.

The condition families fail most is the one the host fails last, posterity. A founder does not want his work preserved; he wants it continued, by owners able to act as he acted, and a fortune frozen in a holding under managers who hold it for a term is to him a kind of death. Wealth beyond what a lifetime can spend has one use, to pass to whoever will keep the long horizon, so the test for who receives it is horizon rather than blood, and an heir who will not keep it is provided for rather than made an owner. Property is inherited; the long horizon is taught. A family must make its blood long-horizon or make the long horizon its blood, by formation, alliance or adoption, and a family that refuses all three has chosen to end.

The evidence most often cited against family capital argues for this. Heirs who take over the management of the family company themselves tend to underperform, and the structure is the answer: the family owns, the best available executive runs, and ownership is separated from management by design. Separation is two thirds of the answer. An owner who was never taught to own reads the structure as an obstacle between himself and his money; the third part is formation, which teaches a member to read a balance sheet, hold a manager to account, sit on a council and plan in the horizon the founder planned in. Leadership is renewed each generation rather than inherited; every type the family produces has a place, the executive, the entrepreneur, the member who wants a dividend and a quiet life, the one who wants nothing to do with the business; and where an individual of ability exists, command is given to that individual, because individuals are fast and committees are not. The youngest heirs, surveyed, rank purpose first among their values and money last. A structure that holds the capital for something larger than any member, and gives each member a place without giving any member the principal, is what they describe when they say they do not want the chair.

The founders who give their fortunes away have diagnosed the problem correctly and chosen the wrong remedy. Carnegie wrote the doctrine in 1889, at the moment states were completing the absorption of the families, and its signatories today build perpetual, purpose-bound, trustee-governed endowments and exclude the family from them; one of them gave each of his children a foundation to run, forming them as custodians of purpose-bound capital and then pointing the capital away from them. The same structure turned inward, with the family and its allies in the governing seats, preserves both the capital and the horizon. The Ford Foundation drifted until a Ford resigned from its board in 1976; the Rockefellers have kept a foundation and a family office side by side since 1913. Structure without formation fails as surely. A recent European founder built almost the whole architecture set out under Law 2, and within four years of his death heirs never formed inside it were finding their way out. The syndrome Brazilian practitioners call síndrome de vira-lata, the heir who would rather take a salary at a multinational than own the family firm, because the ownerless corporation carries the prestige of a state, is a short horizon plus no formation, and formation is its cure. The remedies have been known for centuries. The mirror for princes exists because every crown faced an heir whose horizon might not match the office, and answered with the entail and the tutor; Rome's best emperors were adopted; the oldest companies in Japan survive by adopting capable heirs; the Nation's dowry fund reproduced the Nation by marriage. Generational Planning, marriage, succession, health and education designed as one plan, is where a family keeps this condition for itself, and it accommodates the member with no vocation for ownership without letting him near the principal: a capped allocation of his own, freedom in it, and no vote on the rest.

Most families of consequence live in jurisdictions already failing at least one condition, usually justice or posterity, and have not noticed because security and prosperity still hold. The Index exists to make the failing condition visible before it becomes the only one that matters, and the covenant so that a family knows, when it does, what it owes and what it is owed.

Law 5

A sovereign family contributes to every jurisdiction that benefits it, on terms its citizens welcome.

A jurisdiction's obligations to a family are matched by the family's obligations to the jurisdiction. A family that takes a jurisdiction's protection, its courts, its markets or its passport owes that jurisdiction something in return, and the measure of what it owes is what the jurisdiction's citizens would be glad of if they knew the terms. A family enters no jurisdiction on terms the citizens would refuse, corrupts no law of the host and diminishes it in nothing, and gives what the jurisdiction asks of it in exchange for what it receives.

The reason is, first, self-interest. A programme the host's citizens would refuse, if asked, is withdrawn once they are. Programmes resented by the people of the jurisdiction that granted them are repealed, usually within a decade, and the families that entered under them lose the ground they stood on; a family that has contributed on terms the citizens welcome has a standing no change of government removes. The families that survived the last age of sovereign families did so because their hosts wanted them to stay. The Livornina of 1593 passed because the merchants it invited brought what Livorno wanted, and the Portuguese Nation kept Amsterdam's protection for two centuries because it financed the republic's war and its trade; the Fuggers financed the Empire, the Genoese financed Spain, and the houses of every network in Law 3 paid their hosts in the currency the host valued. The Ismailis obtained a seat agreement from a European state in 2015, making their premises inviolable, after decades of running schools, hospitals and development programmes, over a thousand institutions in more than thirty countries, open to anyone regardless of faith. Services first, then the trust of the state, then status; an institution that asks for recognition before it has earned it is read as a rival.

Contribution is directed at the condition the host most needs, and a family knows it is a protector of a jurisdiction's citizens by what it already guards there. Charity aimed at a failing condition, a hospital where health fails, a school where posterity fails, credit where prosperity fails, is the proof of the family's mandate, and it is the form of counterweight a host can welcome.

The Association's contribution to jurisdictions is a standard. Jurisdictions that wish to receive families need programmes their own citizens will accept, and parliaments need something to point to when they pass one. A published standard for what a lawful programme looks like, whose terms the host's citizens would welcome, is the first rung of a ladder that institutions with no land and no citizens have climbed before. The Holy See was recognized as a sovereign subject for sixty years in which it held no territory. The Order of Malta holds diplomatic relations with over a hundred states and permanent observer status at the United Nations in exchange for neutrality and a humanitarian work states find useful. The International Committee of the Red Cross is a private association with treaty-based privileges in over a hundred countries. Every rung was earned in the same order, service, agreement, status, and the network of sovereign families climbs the same ladder and never seeks the top rung, because a network that became a state would be one more state, and its families would need another network.

The network therefore encourages states and never replaces them. The majority plans short and always will; the state is the institution the majority chooses, and any stateless federation loses to its arithmetic. The state is permanent, and the family that intends to survive wants it strong, well run, and good at the two things a network cannot do for itself, defending territory and governing the many. It rents from the state the protection every one of its jurisdictions provides and pays the rent in the contribution this law describes. Three rules keep the relationship from becoming what absorbed the families before: no privilege the network could not survive losing, since every charter and favour will one day be withdrawn; no leverage used to dominate, since a federation that states answer to is a rival state; and the network as arbitrator of states and arbitrated by none, settling among its members what the states cannot settle, using their courts freely, and letting no single one adjudicate the whole. There is a name for an institution of this shape, and it has been tested for longer than any state has existed: a nation without a state, carrying its own law across the open cities of the world, protected by hosts it never depended on. The network is that model with its exile removed and its membership opened.

The moderating power the old aristocracies claimed and never held has one lawful modern form: a body that jurisdictions are glad to be measured by. States already submit to being rated on their debt. A network that publishes a fair reading of every jurisdiction's standing toward the families it holds, and is believed, has begun the climb; a network that demanded privileges first would have ended it.

Readings

The Journal states what it expects, with dates, so that it can be checked. These are its readings for 2026, revisable as events answer them.

1. By 2030, heirs will liquidate family enterprises at the highest rate on record, and the proceeds will pass overwhelmingly to the largest asset managers rather than into family structures; concentration in asset management will become a political question in at least one major economy.
2. By 2030, at least three jurisdictions will compete for families with explicit charters of family residence, offering their own courts, tax certainty and a path to membership.
3. By 2032, high-tax states will answer with exit taxes and wealth taxes, and the answer will accelerate the sorting rather than reverse it.
4. By 2035, multi-family structures with shared trustees, constitutions and tribunals will be a recognized class of institution, as the multi-family office followed the single family office.
5. Through the 2030s, family capital will finance states again, through co-investment and infrastructure rather than taxation, and the families that do so will be the ones their hosts protect.

The Journal's instruments are the Jurisdiction Success Index, its country readings, the curation standard, and its principles, of which Generational Planning is the next.

Not by ruling and not by standing above, but by being always free to leave, and therefore never worth the trouble of capturing. Every article and every law in this document is a condition of that freedom. The last century belonged to the corporation and the nation-state. The next belongs to the family that holds its capital across jurisdictions rather than under one, keeps it beyond the reach of majorities that plan short, protects the citizens of every jurisdiction it chooses, and is absorbed by none. The ages of prosperity, art and science that history remembers were built by long-horizon families operating structures that outlasted their own lives, in cities open enough to let them. The families that make the transition will be the merchant republics of this century. Those that do not will be its liquidity.

Index of terms

Sovereign family. A long-horizon family that guarantees its own security, prosperity, justice and posterity without depending on any single jurisdiction or institution; that is lawful in every jurisdiction it chooses and absorbed by none; and that stands as protector of the citizens and counterweight to the leadership of each jurisdiction it chooses.

Long-horizon family. A family organized to hold capital, decisions and formation across more than one lifetime. Not an individual, whose planning ends with his life; not a class, since wealth is what a long horizon attracts, never what defines it; not a bloodline, since the horizon can be taught, married or adopted in.

Independent Family Structure. The architecture that holds a family's capital, decisions and formation independent of any single jurisdiction, advisor, trustee or vehicle, so that none can capture the whole and each can be replaced.

Membership, registration, curation. Membership is citizenship or residence in a jurisdiction, by any lawful route. Registration is the domicile and governing law of the family's companies, holdings and fiduciary structures. Curation is the selection, for one family at one time, of the jurisdictions for each, reviewed as the world moves.

Core Jurisdictions. The Vault, where a family's rights and property are guaranteed (a family-first regime); the Forum, where its capital is run and its next generation is tested (a business expansion hub); the Refuge, where its life holds if the world turns (a chosen home, a retirement, a plan B). Vault and Refuge never share a protector.

Four Conditions of Sovereign Success. The Journal's measure of what a jurisdiction owes the families it holds, and what a family owes its members: Security, failing as Violence Outbreak; Prosperity, failing as Economic Collapse; Justice, failing as Lawlessness; Posterity, failing as Demographic Exhaustion. The Jurisdiction Success Index scores jurisdictions on the four.

Family Building Disciplines. Jurisdiction Planning keeps Security. Treasury Planning (wealth and tax) keeps Prosperity. Family Governance keeps Justice. Generational Planning (marriage, succession, health and education) keeps Posterity.

Jurisdiction; state. In this text a jurisdiction is a place a family chooses, holds, leaves or is held by, with the law and leadership that come with it. A state is the institution in general. A crown appears only as history.

Sources and figures

The sources are printed as they appear in the PDF edition. Links from each figure to its primary document are in preparation; noted 1 October 2026.

Wealth and family offices: UBS Global Wealth Report on the intergenerational transfer; Thomas Piketty and Gabriel Zucman, "Capital is Back: Wealth-Income Ratios in Rich Countries 1700–2010," Quarterly Journal of Economics (2014), on private and public wealth; Deloitte Private, "Defining the Family Office Landscape" (2024), on the number and assets of single family offices; J.P. Morgan Private Bank, Global Family Office Report 2026, for the survey of 333 family offices; Brazil Journal (17 September 2026) on research among Brazilian business families. Horizon and ownership: Adolf Berle and Gardiner Means, The Modern Corporation and Private Property (1932); Ronald Anderson and David Reeb, "Founding-Family Ownership and Firm Performance," Journal of Finance (2003); Ronald Anderson, Sattar Mansi and David Reeb, "Founding Family Ownership and the Agency Cost of Debt," Journal of Financial Economics (2003); David Sraer and David Thesmar, "Performance and Behavior of Family Firms," Journal of the European Economic Association (2007); Andrew Ellul, Marco Pagano and Fabiano Schivardi, "Employment and Wage Insurance within Firms," Review of Financial Studies (2018); John Graham, Campbell Harvey and Shiva Rajgopal, "The Economic Implications of Corporate Financial Reporting," Journal of Accounting and Economics (2005); Vikas Mehrotra, Randall Morck, Jungwook Shim and Yupana Wiwattanakantang, "Adoptive Expectations: Rising Sons in Japanese Family Firms," Journal of Financial Economics (2013); Credit Suisse Research Institute, "The CS Family 1000"; Andrew Carnegie, "Wealth" (1889); Robert Higgs, Crisis and Leviathan (1987). Political thought and founding texts: Aristotle, Politics; Augustine, On Free Choice of the Will; Thomas Aquinas, Summa Theologiae I–II, q. 96; John of Salisbury, Policraticus (1159); Ralph Giesey, If Not, Not: The Oath of the Aragonese and the Legendary Laws of Sobrarbe (1968); Magna Carta (1215); the Hungarian Golden Bull (1222); Vindiciae contra tyrannos (1579); the Act of Abjuration (1581); the Livornina (1593); Thomas Hobbes, Leviathan (1651), ch. 21; David Hume, "Of the Original Contract" (1748); Montesquieu, The Spirit of the Laws (1748); the Declaration of Independence (1776); Benjamin Constant, Principles of Politics (1815); the Constitution of the Empire of Brazil (1824), Title 5, ch. 1; Albert Hirschman, Exit, Voice, and Loyalty (1970); the New York Convention (1958). Historical material: Miriam Bodian, Hebrews of the Portuguese Nation (1997), and Jonathan Israel, Diasporas within a Diaspora (2002), on the Portuguese Nation; Fernand Braudel, The Mediterranean and the Mediterranean World in the Age of Philip II, on the Genoese and the Casa di San Giorgio; Sebouh Aslanian, From the Indian Ocean to the Mediterranean (2011), on New Julfa; Philippe Dollinger, The German Hansa; Mark Häberlein, The Fuggers of Augsburg (2012); Niall Ferguson, The House of Rothschild (1998); Malcolm Barber, The Trial of the Templars; Geoffrey Parker, The Dutch Revolt (1977), on Antwerp and the Tenth Penny; the Constitution of the Shia Imami Ismaili Muslims (1986) and the agreement between Portugal and the Ismaili Imamat (2015); the International Group of P&I Clubs on the shipowners' mutuals. Figures marked as approximate are the Journal's reading of the cited sources; the online edition traces each to its primary document.