Concept
Corporate tax residence answers one question: which state is entitled to tax a company's worldwide profit, not just the profit earned on its territory. Every other corporate tax test — permanent establishment, withholding at source, CFC attribution — divides up parts of income; residence assigns the whole base and, with it, the company's filing obligations and its right to invoke tax treaties as "a resident of a Contracting State." One question comes before it. Residence presupposes an entity that is an opaque person in its own right: a partnership, or a US LLC left transparent, has no residence of its own to assign, because the income lands on its members wherever they happen to be resident — and under the check-the-box rules an eligible entity can move between those tax classifications by election (Treas. Reg. §301.7701-3). That prior question belongs to entity classification.
Tax residence has concrete consequences. A company that is tax resident of, say, Ireland files Irish corporation tax returns on its worldwide profit, obtains an Irish certificate of tax residence, and on the strength of that certificate claims reduced withholding rates under Ireland's treaties. The same mechanism cuts the other way: a state that considers the company resident by its own test will assert tax on everything the company earns anywhere, whether or not another state does the same.
A trading company is incorporated in Delaware; its founder moves to London and runs it from there — negotiating, deciding acquisitions, approving budgets. Under US law the company is and remains domestic: IRC §7701(a)(4) defines a domestic corporation as one "created or organized in the United States or under the law of the United States or of any State," and neither the statute nor Treasury Regulation §301.7701-5 asks where management sits. Under UK law the same company has become UK resident, because the case-law test — central management and control — abides where its real strategic decisions are taken. Two full worldwide claims now overlap, and the 2001 US–UK treaty does not break the tie automatically: Article 4(5) sends a dual-resident company to the two competent authorities, and until they agree, the company "shall not be entitled to claim any benefit provided by this Convention" apart from a narrow core.
Three features of this test are routinely misread. First, it is not registration: a registered office, a legal address and a company number fix which company law governs the entity — they do not, in a management-test state, fix where it pays tax, and a certificate of incorporation proves nothing about residence there. Second, it is not permanent establishment: a PE gives another state the right to tax the profit attributable to a local presence — a slice, with its own rules — while residence claims the whole; having a PE somewhere does not make the company a resident there, and residence somewhere does not prevent a PE elsewhere. Third, documents alone are insufficient: in every management-test dispute of the last century, from De Beers (1906) to Development Securities (2020), the tribunal considered evidence beyond the board minutes to establish who actually decided.
Key parameters of the regime:
| Parameter | Value |
|---|---|
| What residence confers | The state's right to tax worldwide profit, plus filing obligations and treaty access as a resident |
| Transparent entities | A partnership or transparent US LLC has no residence of its own; check-the-box under Treas. Reg. §301.7701-3 |
| Incorporation test | Where the entity was legally created (pure case: the United States) |
| Management test | Where strategic decisions are actually made (classic case: the United Kingdom) |
| Dual residence | Both tests passed in two states at once; treaties resolve it by negotiation between tax authorities |
| Registered address | Fixes which company law governs the entity; fixes tax only in incorporation-test regimes |
What Residence Decides — and What It Does Not
Residence, registration and permanent establishment are three different legal objects that constantly get merged into one in practice. The table separates them before anything else, because half of all corporate residence errors are really category errors.
| Question | Incorporation / legal address | Tax residence | Permanent establishment |
|---|---|---|---|
| What it determines | Which company law governs the entity: organs, capital, insolvency, ability to migrate | Which state taxes worldwide profit; which treaty network the company can invoke | Whether a state where activity happens may tax the profit attributable to that activity |
| Established by | A registry act — one-time, documentary | The residence test of each interested state — factual in management regimes, documentary in incorporation regimes | Facts of presence: fixed place, dependent agent |
| Scope of the tax claim | None by itself | Worldwide profit, plus filing, withholding-agent and disclosure obligations | Attributable profit only, plus local registration and filing |
| Can it multiply? | One incorporation at a time | Yes — dual (or plural) residence when tests overlap | Yes — one PE per state of activity, in parallel with residence elsewhere |
The practical consequence of the middle column is the largest: the residence state taxes profits the company earns anywhere, requires returns and accounts under its own rules, typically makes the company a withholding agent for its outbound payments, defines it as "resident of a Contracting State" for treaty purposes — and, in many systems, charges a deemed disposal of assets when residence leaves. Irish law, for instance, treats a company that ceases to be resident as having disposed of its assets at market value, with capital gains tax on the result, subject to narrow exceptions (Revenue guidance). Residence is also the reference point for other people's tests: whether a shareholder's home state treats the company as a controlled foreign company is measured against where the company is resident and how it is taxed there.
Residence also has limits. It settles which state may tax the whole base; it does not settle how large that base is. What a resident company may report on intra-group sales, licences and loans is governed by transfer pricing rules that apply wherever the parties are resident, and the slice another state may tax where activity actually happens is measured by permanent establishment rules on their own facts. Nor is domestic residence by itself sufficient to invoke a treaty. Treaty residence has a separate definition: a "resident of a Contracting State" is a person who, under that state's laws, is "liable to tax therein by reason of his domicile, residence, citizenship, place of management, place of incorporation, or any other criterion of a similar nature," and the definition excludes a person liable to tax there only on income from sources in that state (US–UK treaty, Art 4(1)). A company incorporated where no corporate income tax exists can therefore be resident for domestic purposes and have no treaty residence to invoke at all.
Two Families of Domestic Tests
National definitions of corporate residence fall into two families, and most real regimes combine them.
Incorporation-based tests attach residence to the legal act of creation. The United States is the pure case: a corporation is domestic if "created or organized in the United States or under the law of the United States or of any State" (IRC §7701(a)(4)), foreign if it is not (§7701(a)(5)), and the regulation ties the classification solely to the jurisdiction of organization. Management can sit anywhere on earth without moving the needle — which is why a Delaware corporation cannot "move" its US tax residence by relocating its board, and why a foreign-incorporated company run from New York does not thereby become a US resident (its US exposure runs through effectively connected income and treaty PE rules instead — a different, narrower claim; see how this plays out for a non-resident's US LLC). The main carve-out runs the other way: US anti-inversion rules (IRC §7874) can deem a foreign-incorporated acquirer of a US business to be a domestic corporation — an incorporation regime defending itself against incorporation-shopping.
Management-and-control tests attach residence to where the company is actually run. The United Kingdom is the source of the doctrine: a company incorporated abroad is UK resident if its central management and control — the highest level of decision-making, not day-to-day operations — abides in the UK. The rule is case law, over a century old, and its content depends on how tribunals apply it.
Most regimes are hybrids, and the direction of the hybrid matters:
| Regime | Incorporated locally | Incorporated abroad | Treaty override | CFC exposure if resident elsewhere |
|---|---|---|---|---|
| United States | Always domestic — IRC §7701(a)(4) | Never resident, wherever managed | No override of domestic status; dual residence handled treaty-by-treaty (e.g. US–UK Art 4(5)) | Subpart F and GILTI reach a foreign company controlled by US shareholders |
| United Kingdom | UK resident by CTA 2009 s.14 | UK resident if central management and control is in the UK (case law) | CTA 2009 s.18: a company treaty-resident elsewhere is non-UK resident for the Corporation Tax Acts | UK CFC charge on a non-resident company controlled from the UK, subject to the gateway exemptions |
| Ireland | Resident if incorporated on or after 1 January 2015 (pre-2015 companies transitioned by end-2020), unless treaty-resident elsewhere | Resident if centrally managed and controlled in Ireland | Built into the incorporation rule itself — treaty residence elsewhere switches it off | ATAD CFC charge where Irish significant people functions generate the undistributed income |
| Germany | Resident if the registered office (Sitz) is in Germany — KStG §1(1) | Resident if the place of management (Geschäftsleitung) is in Germany — either connecting factor is sufficient on its own | No domestic switch-off of the UK s.18 type: a treaty tie-breaker limits what Germany may tax, but §1 unlimited liability is not itself removed by the statute | §7 AStG attributes the passive income of a controlled foreign company |
| Netherlands | Deemed resident by the incorporation fiction for corporate income tax | Resident where the place of effective management is in the Netherlands, judged on facts and circumstances | The fiction stands; the tie-breaker governs what the Netherlands may tax | ATAD Model A, confined to low-taxed and listed jurisdictions |
| Luxembourg | Resident if the statutory seat is in Luxembourg (art. 159 LIR) | Resident if the central administration is in Luxembourg — either factor suffices on its own | Tie-breaker allocates; the domestic charge is not switched off | ATAD CFC (art. 164ter LIR), on a significant-people-functions basis |
| Cyprus | From 1 January 2026, Cyprus incorporation creates residence unless a double-taxation treaty provides otherwise | Resident if management and control is exercised in Cyprus — the historic sole test, still in place | The exception requires a contrary result under the applicable tax treaty; residence under another state's domestic law alone is insufficient | ATAD CFC applies to companies; no attribution to individuals |
| Switzerland | Resident if the seat is in Switzerland (art. 50 DBG) | Resident if the effective administration is in Switzerland | Tie-breaker allocates; federal and cantonal liability follows | No CFC regime |
| Singapore | Incorporation alone decides nothing: residence follows control and management | Resident if control and management of the business is exercised in Singapore — in practice, where the board meets and decides | IRAS issues the certificate of residence on the same control-and-management footing | No CFC regime |
| Hong Kong | Profits tax is territorial, so residence matters for treaties rather than for the charge; a Hong Kong-incorporated company is resident for treaty purposes | Resident for treaty purposes if normally managed or controlled in Hong Kong | The Certificate of Resident Status is issued case by case, with substance questions asked | No CFC regime; the FSIE rules do the equivalent work |
| UAE | A juridical person incorporated in the UAE is a Resident Person under art. 11 of the Corporate Tax Law | A foreign juridical person effectively managed and controlled in the UAE is also a Resident Person | Tax residency certificates are issued by the Federal Tax Authority | No CFC regime |
| Russia | Russian organisations are residents by the fact of Russian incorporation (art. 246.2 NK) | A foreign organisation is a Russian resident where its place of management is in Russia, on statutory criteria about where the executive body operates | A foreign organisation may also be a resident by treaty; a large part of the Russian treaty network is currently suspended | CFC rules (art. 25.13 NK) attribute undistributed profit to the controlling person |
| Italy | Resident if the registered office is in Italy (art. 73(3) TUIR) | Resident if the place of effective management, or ordinary day-to-day management, is in Italy for most of the period — the test rewritten with effect from 2024 | Tie-breaker allocates; the Italian charge is not switched off by statute | Art. 167 TUIR, on controlled low-taxed entities |
The Irish regime illustrates how approaches to corporate residence evolve. Until 2015 an Irish-incorporated company managed from a zero-tax jurisdiction could be resident nowhere useful — the architecture behind the "double Irish." Ireland closed it by adding an incorporation rule (TCA 1997 s.23A, as described by Revenue): incorporation now creates residence by default, and the management-and-control test continues to apply to foreign companies run from Ireland. The global trend is exactly this — regimes add the test they lacked, so the opportunities for a company to be resident nowhere keep decreasing, while the opportunities for it to be resident twice keep increasing.
Germany applies a management test outside the common-law system. Both connecting factors are statutory: a corporation is subject to unlimited German corporate income tax if it has either its registered office or its place of management in Germany, and that unlimited liability "extends to all income" (KStG §1(1)–(2)). The Fiscal Code then defines the terms — the registered office is the place fixed by law, articles or statutes (AO §11), and the place of management is "the centre of commercial executive management" (AO §10). That centre differs from the criterion used by the UK test. The British test looks for the highest level of strategic control, wherever the routine business is run; the German definition points at where the executive running of the business is concentrated. Two management-test states can therefore apply honest tests to identical facts and land on different countries — which is one of the ordinary ways a company becomes dual resident without anyone planning it.
Read down the first two columns and the pure incorporation regime turns out to be rare. The United States is the only one of the thirteen that ignores management entirely. Everyone else either began with a management test and later bolted on incorporation — Ireland in 2015, Cyprus from 31 December 2022, expanded from 1 January 2026 — or carried both connecting factors from the start, as Germany, Luxembourg, Switzerland, Italy, the Netherlands and the UAE do. Only two still run on management alone: Singapore, where residence follows control and management and the certificate of residence is issued on that footing, and Hong Kong, where profits tax is territorial and residence is a treaty concept rather than a charging one — set out in Hong Kong tax residence. The direction of travel is one-way: regimes add the limb they lack, and the space for a company resident nowhere keeps closing.
The treaty-override column separates two attitudes to the same problem. The United Kingdom surrenders outright: a company treaty-resident elsewhere simply stops being UK resident for the Corporation Tax Acts. Ireland builds the surrender into the charging rule. The continental regimes do neither — Germany, Luxembourg, Switzerland, Italy and the Netherlands keep the domestic charge on foot and let the treaty allocate the base, which is why a company treaty-resident in another state still files where it was incorporated and still answers for its worldwide accounts there. Cyprus and the UAE sit closer to the British model on the incorporation limb. For Cyprus from 1 January 2026, the exception requires a contrary tax-treaty result, not merely residence under another state's domestic law.
The last column answers the question the table provokes: if management moves out, does the old state lose interest? Rarely. Germany attributes the passive income of a controlled foreign company under §7 AStG; the United Kingdom runs its own gateway analysis; Ireland, Luxembourg, Cyprus, the Netherlands and Italy each implement ATAD CFC on their own terms; the United States reaches a controlled foreign corporation through Subpart F and GILTI whatever its residence. Only the zero-CFC group — Switzerland, Singapore, Hong Kong and the UAE — genuinely stops asking, and Hong Kong replaces the question with its FSIE rules. Moving the board therefore changes which state taxes the company; it does not by itself change which state taxes the group.
Where to look first, by structure
| Structure | What the table predicts | What to establish |
|---|---|---|
| Holding incorporated in the Netherlands, Luxembourg or Cyprus, board sitting abroad | The local limb bites on incorporation; the other state's management test may bite as well | Whether the tie-breaker has been run, and whether the substance file supports the seat |
| Company incorporated in Singapore or Hong Kong, decisions taken in Europe | Incorporation creates no local residence; the European management test can capture it | Where the board actually decides, and whether a certificate of residence will issue at all |
| Delaware corporation run from London | Two full worldwide claims at once | The competent-authority procedure under the treaty, and what benefits are suspended meanwhile |
| UAE company with a European resident director | Art. 11 residence at home, plus a possible second residence where the director decides | Whether board decisions are taken and minuted in the UAE, and by whom |
| Russian-incorporated company managed abroad | Russian residence follows incorporation regardless of where management sits | Whether the foreign state also claims residence, and whether the relevant treaty article is in force |
Central Management and Control: Where Decisions Actually Abide
The UK line of authority is worth learning even outside the UK, because it is the most fully litigated version of the question every management-test state asks. The founding formulation is Lord Loreburn's in De Beers Consolidated Mines Ltd v Howe (1906): a company resides "where its real business is carried on … and the real business is carried on where the central management and control actually abides." De Beers mined diamonds in South Africa and was incorporated there; it was held UK resident because the board that actually directed it sat in London. From the first case, the test identified the actual decision-makers regardless of the place of registration.
HMRC's published approach (Statement of Practice 1/90) turns the doctrine into three questions asked in order: do the directors in fact exercise central management and control; if so, where do they exercise it; and if not, where and by whom is it exercised. The location of board meetings is "important in the normal case, but not necessarily conclusive" — directors who actually run a wholly-UK business from the UK do not transfer residence by flying out for formal meetings. And Unit Construction v Bullock (1960) illustrates the reverse situation: African subsidiaries whose constitutions required boards to meet outside the UK were still UK resident, because the London parent in fact ran them — the constitution differed from the facts, which determined the court's conclusion.
Three modern cases mark out where the line runs between a real board and a bypassed one:
- Wood v Holden [2006] EWCA Civ 26 — the taxpayer side of the line. A Netherlands company in a share-sale structure had a professional corporate director that did little more than consider and sign the key documents. HMRC argued the real decisions were made by the UK owners and their advisers. The Court of Appeal disagreed: the director had been "advised and influenced, but not bypassed nor stood aside." A board that makes few decisions, on advice, with strong shareholder influence, still makes them — influence is not usurpation.
- Laerstate BV v HMRC [2009] UKFTT 209 (TC) — the other side. A Netherlands company's sole active mind was its UK-resident shareholder; the nominal director signed what he was told to sign without discussion or deliberation. The tribunal located central management and control with the shareholder in the UK: a director who abdicates to a dominant outsider is not exercising control, he is recording it.
- Development Securities plc v HMRC [2020] EWCA Civ 1705 — the refinement. Jersey subsidiaries were set up with local professional directors specifically to be Jersey-resident during a short critical window of a tax scheme; the directors met in Jersey, took advice, and approved transactions that made no commercial sense for the companies themselves because the UK parent had decided they should happen. The Court of Appeal held the companies were UK resident in the window: the Jersey boards had examined whether the companies could enter the transactions — legality, authority, documentation — but the decision whether they should had been taken in London, and the boards simply acceded to it.
The doctrine includes two further propositions that preclude a simple formal answer. Control can be divided: in the Union Corporation cases (34 TC 207) the Court of Appeal treated a company as resident wherever acts of controlling power and authority are exercised "to some substantial degree," rather than requiring a single place of control, and Swedish Central Railway Co v Thompson (9 TC 352) had already observed that it is "rather easier to think of a company having two places of residence than it is of a natural person" (HMRC INTM120210). A genuinely split board is a route into dual residence, not a hedge against it. And dependence is not abdication: HMRC's own review guidance accepts that "it remains possible for central management and control to lie with a subsidiary's directors even where it seems improbable that they would act other than in accordance with the directors of the parent" (INTM120180). The question is whether the board considered the decision and could have refused it, not whether anyone expected it to.
Local substance concerns a separate question and does not by itself determine residence. Payroll, premises and expenditure recorded under an economic substance regime show that people and costs are where the company says they are; they do not show where a decision was taken. Substance filings and the substance dossier behind them are useful evidence in a residence dispute and have never been a defence to one.
Board Minutes Against the Substance of Decision
Every residence-planning memo says "hold board meetings in the right place and minute them." The case law shows why that is necessary but nowhere near sufficient. A minute proves that a meeting occurred, who attended, and what was resolved. It does not prove that the decision was made there — and that is the fact in issue. Tribunals reconstruct the decision from the whole evidence trail, and that trail carries greater evidential weight than the minute book:
| Decision type | Evidence that supports local control | Evidence that destroys it |
|---|---|---|
| Strategy, budgets, business plan | Papers circulated in advance; directors' comments and questions; alternatives considered; a rejected proposal on record | Strategy documents authored at the shareholder level and adopted verbatim; no board-level record of alternatives |
| Acquisitions and disposals | Board sees the deal before terms are final; negotiates or amends; instructs its own advisers; can and sometimes does say no | Term sheet signed or price fixed before the board ever met; board resolution dated after the economic decision; could-not-should review of a done deal |
| Financing and security | Board weighs terms, considers the company's own interest, records why the terms are acceptable | Uncommercial terms accepted without recorded reasoning — the Development Securities fact pattern |
| Distributions | Reserves and solvency actually reviewed; timing decided by the board | Dividend "requested" by the shareholder and paid the same day; resolution drafted by the recipient |
| Appointments and key contracts | Board selects, negotiates fees, supervises | Appointees imposed from outside; board learns of contracts after signature |
| Day-to-day operations | Largely irrelevant either way — the test looks at the highest level of control, not administration | Also irrelevant: heavy local administration cannot compensate for strategic decisions made elsewhere |
The assessment of this evidence follows a general principle: the paper a company generates before a decision — drafts, questions, dissent, time to consider — is worth more than the paper generated at it. Emails are part of the evidence trail: in the litigated cases, the sequence of correspondence (who proposed, who instructed, who was informed) carried greater evidential weight than any formal document.
Dual Residence and the Treaty Layer
When an incorporation state and a management state both claim the company — or two management states read the same facts differently — the company is dual resident, and each state applies its full worldwide claim. Domestic law rarely solves this alone; the treaty between the two states is the instrument that can, and its mechanics changed in 2017.
The old rule in treaties following the pre-2017 OECD Model broke the tie automatically: a dual-resident company was deemed resident only of the state of its place of effective management (POEM). One fact — where key management and commercial decisions are in substance made — decided the treaty outcome without anyone's discretion. Many in-force treaties still carry this rule.
The 2017 rule abandoned the automatic tie-breaker. Article 4(3) of the 2017 OECD Model — carried into existing treaties by Article 4 of the MLI where both sides' positions match — sends the case to the two competent authorities, who "shall endeavour to determine by mutual agreement" the company's treaty residence, "having regard to its place of effective management, the place where it is incorporated or otherwise constituted and any other relevant factors." The consequences follow from the default: "in the absence of such agreement, such person shall not be entitled to any relief or exemption from tax provided by the Covered Tax Agreement except to the extent and in such manner as may be agreed." No agreement — no treaty benefits. The Model alone does not establish whether a given treaty was actually modified: it depends on both states' MLI notifications and reservations, checked pair by pair — the same matching exercise described in the tie-breaker guide.
Some treaties wrote the negotiated approach in directly. The 2001 US–UK treaty, Article 4(5): where a person other than an individual is resident of both states, the competent authorities "shall endeavour to determine by mutual agreement the mode of application of this Convention," and failing agreement the company cannot claim any treaty benefit except relief from double taxation, non-discrimination and the mutual agreement procedure itself. HMRC's guidance on such cases is candid about the interim position: while the authorities have not agreed, the company simply remains dual resident — and resident in the UK.
The corporate tie-breaker differs from the sequence of criteria for individuals. There is no "permanent home," no "centre of vital interests," no habitual-abode step for companies — that sequence belongs to individuals under Article 4(2) and its own guide. Applying that sequence in a corporate analysis means using an inapplicable test.
For a company that finds itself dual resident, the next analysis is concrete and runs in order: identify the exact treaty in force between the two states and read its Article 4 text as modified (or not) by the MLI — the automatic POEM rule and the negotiated rule lead to different strategies; if POEM applies, assemble the evidence of where effective management sits, because the outcome follows the facts directly; if the negotiated rule applies, prepare the competent-authority file — chronology of decisions, board composition, incorporation history, economic ties — and initiate the mutual agreement procedure, while planning cash flows on the assumption that treaty relief is suspended until agreement; and in parallel, check what each domestic law does in the meantime — a UK example: under CTA 2009 s.18 a company that wins treaty residence elsewhere becomes non-UK resident for domestic purposes, which itself can trigger exit consequences.
Proving Residence to Someone Else
Residence is a question of fact in the abstract and a document in practice: a foreign payer withholding at its full domestic rate will not drop to a treaty rate on assurances. Each state runs its own certification channel, and the certificates say less than the people relying on them assume.
HMRC certifies that a company "is a resident of the UK within the meaning of the DTA" — not residence at large. The request has to identify the treaty and the Article and income source it relates to; HMRC checks the position and may refuse where it is clear the claimant would not qualify under that Article, and its guidance states plainly that a certificate "will not guarantee" a successful claim, because the overseas authority decides (INTM162010). The US channel is a form exercise with a price: an entity applies on Form 8802, whose use is mandatory and which carries a user fee, and receives Form 6166 — a letter on Treasury stationery certifying that the entity is a US resident for income tax purposes (IRS).
Two limits follow. A certificate records the issuing state's view under its own test; it does not bind the other state, and in a dual-residence case both states can certify the same company for the same year. And residence is necessary, not sufficient: under the principal purpose test written into modernised treaties, a benefit "shall not be granted" where it is reasonable to conclude that obtaining it "was one of the principal purposes of any arrangement or transaction," unless granting it would accord with the object and purpose of the relevant provisions (MLI Art 7(1)). The certificate answers who the company is; the anti-abuse rule asks why the structure exists, and a claim to reduced withholding needs both answers.
Leaving: The Exit Charge and What Stays Behind
Residence is expensive to acquire and expensive to give up. A state that can lose a company's worldwide base as a matter of fact — a management regime above all — levies an exit charge, taxing gains that have accrued but not been realised.
The UK rule is the model. Where a company ceases to be UK resident, it "shall be deemed for all purposes of this Act to have disposed of all its assets … immediately before the relevant time" and immediately to have reacquired them "at their market value at that time" (TCGA 1992 s.185). The charge attaches to the moment of departure and to no transaction at all — which is what gives the date on which management moved a price. The carve-out is the instructive part: assets that remain in the UK and are used in, or held for the purposes of, a trade carried on through a UK permanent establishment are excepted from the deemed disposal (s.185(4)). The departing state gives up the worldwide claim and keeps the slice it can still reach — the two objects the opening table separated, now operating in sequence.
The EU generalised the design. Under the Anti-Tax Avoidance Directive a Member State taxes market value less value for tax purposes where, among other cases, "a taxpayer transfers its tax residence to another Member State or to a third country, except for those assets which remain effectively connected with a permanent establishment in the first Member State"; the taxpayer has a right to pay that exit tax in instalments over five years where the destination is an EU or qualifying EEA state (Directive (EU) 2016/1164, Art 5). Ireland's deemed disposal on ceasing to be resident is the same idea in national form.
Three Scenarios, One Evidence Trail
The scenarios below change one variable — who really decides — and follow both the residence outcome and the paper that proves it. Facts and figures are illustrative teaching constructions, not case reports.
Remote board, external decision-maker
A holding company is incorporated in a no-tax jurisdiction with three local professional directors. Board meetings happen quarterly by video, minuted impeccably. Every transaction, however, arrives pre-negotiated by the beneficial owner, who lives in London: he agrees terms with counterparties, instructs lawyers, and the board's packet contains final documents for approval. This is the Laerstate/Development Securities fact pattern: the directors verify that the company can sign; whether it should was decided in London before they ever met. On the UK test, central management and control abides with the owner; the company is UK resident, its worldwide profit within UK corporation tax, and its home-jurisdiction status is irrelevant to that claim. The evidence trail that decides the case is the one nobody curated: e-mail chronology showing decisions preceding meetings, term sheets dated before board packets, minutes recording no questions and no alternatives.
Local directors with real authority
Same shell, different practice. The board receives proposals as proposals: drafts circulate two weeks ahead; directors ask for valuation support; one financing is renegotiated because the board considered the security package too onerous; one acquisition is declined and the refusal minuted with reasons. The owner is consulted and his views carry weight — but the record shows deliberation happening at the board, on the company's own interest. This is Wood v Holden territory: advised and influenced, not bypassed. Central management and control sits where the board exercises it, and a UK-side challenge fails on the same chronology that convicted the first company — because here the chronology shows decisions being formed at the meetings, not merely recorded there.
Moving management without re-incorporating
A Delaware-incorporated company's founder relocates to London and continues running it. Nothing was filed anywhere, yet the tax position changed completely: the company remains US domestic forever under IRC §7701(a)(4) — an incorporation regime cannot be exited by moving people — and it has become UK resident by central management and control from the day decisions began abiding in London. Dual residence arises not from a plan but from a house move. Under US–UK Article 4(5) there is no automatic tie-break: the competent authorities may agree a mode of application, and until then treaty benefits are largely unavailable.
The evidence question now runs in both directions — the company needs a defensible date on which management moved (lease of the London office, first UK board meeting, relocation of the decision-maker), because that date starts UK worldwide taxation, and, in the mirror case of management moving out of a management-test state, the same date typically triggers that state's exit charge on unrealised gains and, where a treaty override like CTA 2009 s.18 operates, the loss of domestic residence itself. Management migration is a taxable event; if its date is undocumented, that date will later be chosen by the party taking the least favourable position for the company.
Q/A
Which of these regimes still decides residence by management alone?
Two of the thirteen in the table. Singapore looks only at where control and management of the business is exercised, so incorporation there creates no residence by itself. Hong Kong taxes on a territorial source basis, so residence matters for treaty access rather than for the charge, and a company incorporated outside Hong Kong qualifies if it is normally managed or controlled there. Everywhere else in the table at least one incorporation-type limb now exists — Ireland added one in 2015, Cyprus from 31 December 2022, expanded from 1 January 2026 — which is why a company resident nowhere has become hard to build and a company resident twice has become easy.
If we move the board out of the country, does the old state stop taxing us?
It may stop taxing the company and still tax the group. Where residence is decided by management alone, moving the decision-making moves residence. Where incorporation is also a limb — Germany, Luxembourg, Switzerland, Italy, the Netherlands, Ireland, Cyprus, the UAE, Russia, the United States — the company stays resident where it was created, whatever the board does. And even a clean move does not exclude the application of CFC rules: §7 AStG in Germany, the UK gateway, the ATAD implementations, Subpart F and GILTI in the United States all reach a foreign company controlled from home. Only Switzerland, Singapore, Hong Kong and the UAE have no CFC regime to apply.
Our company's legal address is in a zero-tax jurisdiction — is it taxed only there?
The legal address answers the company-law question, not the tax question. It fixes residence only for incorporation-regime purposes in that jurisdiction itself. Any management-test state from which the company is actually run — where its strategic decisions are made — can claim the company as its own resident and tax worldwide profit, regardless of the registry. The registered office is one fact among many, and in the litigated cases it was the least important one.
We have a permanent establishment in Germany — has the company become a German tax resident?
No. A permanent establishment gives Germany the right to tax the profit attributable to the German presence — a defined slice, with registration and filing for that slice. Residence is a separate test on separate facts (for Germany, principally the place of management), and a PE does not create it. The two claims coexist: a company can be resident in one state and hold PEs in five others, each taxing only its attributable share.
The company turned out to be dual resident — what exactly do we do next?
Three steps in order. First, find the applicable treaty and establish which Article 4 text governs: the old automatic place-of-effective-management rule, or the post-2017 negotiated rule — checking both states' MLI positions, since the MLI rewrites Article 4 only where notifications match. Second, if POEM decides, assemble the management-location evidence and apply it; if the negotiated rule applies, prepare a competent-authority file (decision chronology, board composition, incorporation, economic ties) and initiate the mutual agreement procedure. Third, plan the interim: until agreement, treaty benefits may be suspended, and each state's domestic law — filing, withholding, exit rules — continues to apply in full.
If our board minutes are in perfect order, is residence safe?
No. Minutes prove a meeting happened and what was resolved — not that the decision was made there. Tribunals reconstruct decisions from the whole trail: papers circulated in advance, questions asked, alternatives weighed, timing of resolutions against the underlying transactions, e-mail sequences showing who proposed and who instructed. In Development Securities the Jersey minutes were in order; the companies were UK resident anyway, because the boards had reviewed legality while the commercial decision arrived from London.
Our directors follow the owner's recommendations — is that already fatal?
Not by itself. The case law distinguishes influence from usurpation: in Wood v Holden a board that decided little, on advice, under a dominant shareholder still controlled the company because it genuinely made the decisions it signed. The line is crossed when the board stops deliberating — signs without consideration, or examines only whether the company can do what has already been decided elsewhere. The decisive point is the board's failure to deliberate, rather than the owner's influence itself.
Can our US corporation become non-US by moving its management abroad?
No. US status follows incorporation alone: a corporation organized in a US state is domestic wherever it is managed, and stays so until it ceases to exist under US law (re-domestication is a corporate transaction, not a management move — and anti-inversion rules can even deem some foreign-incorporated companies domestic). Moving management abroad does not remove US taxation; it can add a second residence in a management-test state on top of it.
What happens in the old country when we move management to a new one?
Three things to check. Dual residence first: if the old state uses an incorporation test, its claim survives the move entirely. Exit charges second: management-test states commonly treat outbound migration as a deemed disposal of assets at market value — Ireland's rule is explicit — so the move has a tax cost even with no sale. Treaty mechanics third: where the new residence wins under a treaty, override rules like the UK's s.18 convert the company to domestic non-resident, which is itself the trigger for exit consequences. The date of the move should be documented deliberately, because every one of these consequences attaches to it.
Is the corporate tie-breaker the same as the one for individuals?
No, and borrowing it is a standard error. The individual ladder — permanent home, centre of vital interests, habitual abode, nationality — exists only for individuals under Article 4(2). For companies, older treaties used a single automatic criterion (place of effective management), and post-2017 treaties use no automatic criterion at all: competent authorities decide by mutual agreement, weighing effective management, incorporation and other factors, with treaty benefits suspended in the meantime.
Does appointing local directors create residence where they sit?
Only if they genuinely exercise central management and control there — real deliberation, real authority to refuse, decisions formed rather than recorded. Professional directors who sign pre-decided transactions do not establish tax residence there: the residence follows the real decision-maker, as the bypassed-board cases show. And a genuinely active board split across two countries can produce the opposite problem — management and control exercised to a substantial degree in each, which is one recognised route into dual residence.
Our board meets by video from several countries — where is management and control?
Where the decisions actually abide, which a video meeting makes harder, not easier, to locate. The inquiry stays factual: where the participants who actually shape decisions sit, where papers are prepared, where the deliberation that matters happens over time. A company that wants a defensible answer concentrates the substance — the deciding directors, the preparatory work, the record of deliberation — in one place, rather than relying on the formal designation of a meeting's venue.
Working from a home abroad can raise a separate PE question. The 50% working-time threshold in the 2025 OECD home-office Commentary does not decide corporate residence: the company's residence still requires its own domestic-law and treaty analysis.
For a director who is the only or primary person conducting the enterprise's business, paragraph 44.20 requires separate consideration in the PE analysis; the ordinary working-time test is not a residence safeguard. Where key management and commercial decisions are made remains a separate question. The permanent establishment side of that boundary is mapped in permanent establishment, and the threshold itself, with the OECD's examples, in the home office abroad.
Our vehicle is a partnership, or an LLC we left transparent — where is it resident?
Usually nowhere, in the sense of corporate tax residence. A vehicle treated as transparent has no worldwide base of its own: the income is attributed to its members and taxed by reference to their residence, so the question moves up a level. Two complications follow. Classification is state-specific — the same LLC can be transparent at home and opaque abroad, which is how hybrid mismatches arise — and under the US check-the-box rules an eligible entity can elect the answer, subject to the default classifications for domestic and foreign entities. And a transparent vehicle is generally not itself "liable to tax," so it is generally not a resident of a Contracting State for treaty purposes; access then depends on the members and on whatever the particular treaty says about fiscally transparent entities. Start with entity classification, not with residence.
We hold a certificate of tax residence — does that guarantee the treaty rate?
No. A certificate records that the issuing authority regards the company as its resident under the treaty definition, for the Article and income source stated in the request. HMRC says in terms that a certificate will not guarantee that the claim succeeds: the other state decides, applying its own view of residence and beneficial ownership, any limitation-on-benefits article, and the principal purpose test, which denies the benefit where obtaining it was one of the principal purposes of the arrangement. In a dual-residence case the certificate is weaker still, because the other state can hold — and certify — the opposite view until the competent authorities agree.