Halvorsen & Reith

Articles, shareholder and joint venture agreements across 40

How articles, shareholder and joint venture agreements function across a cross-border group depends on which jurisdiction is asked, not on which template was used to draft them. The same clause that fixes a veto right in one register is treated as a private contractual promise in another, and that difference decides who can rely on the clause against a buyer, a liquidator or a co-investor. This page maps where the difference actually sits, jurisdiction by jurisdiction, and marks where the advisory perimeter around this work stops.

A holding company negotiates a joint venture agreement with a co-investor in a jurisdiction neither side has used before. Both legal teams assume the agreement itself settles who controls what. Six months later a dispute arises, and the answer turns out to depend on whether the relevant term was also carried into the articles filed at the local register – something nobody checked at signing, and something that is now difficult to fix without both parties agreeing to it in public.

The sections below set out what this practice covers, the axis along which jurisdictions genuinely differ, and how a group works out which axis applies to its own structure.

What this practice covers – and where it stops

This practice covers three related documents and the space between them: the articles of association (or the equivalent constitutional instrument, whatever it is called locally), the shareholders' agreement, and the joint venture agreement negotiated when the shareholders are not simply passive investors but active co-owners of a business. The work is drafting, reviewing and reconciling these documents, mapping which one governs which decision, and identifying where a term that the parties believe is settled is in fact only a private promise between two signatories. This is company law read as a working document, not as background reading before a transaction.

What this practice does not do follows from a licensing boundary, not from a choice about scope. The firm does not act as, supply, source or arrange a director, a company secretary, a nominee shareholder or a trustee for a client structure, and it does not undertake any activity for which a trust or corporate service provider licence is required in the jurisdiction concerned. Where the governance question a client raises turns out to require one of those roles to be filled, the practice can map exactly what the role needs to do and what standard the appointment has to meet – the articles review service is where that mapping usually starts – but filling the role itself sits outside this perimeter in every jurisdiction where it is regulated. That distinction matters more here than in most areas of governance work, because a shareholders' agreement that assumes a director will simply be found is a shareholders' agreement with a gap in it.

The axis on which articles, shareholder and joint venture agreements differ across a cross-border group

The axis is not civil law against common law, and it is not a matter of which jurisdiction is stricter. The axis that actually matters is this: what does a filed, public document control, and what can a private agreement only promise between the people who signed it. In some jurisdictions the articles are treated as the single instrument that binds the company and can be relied on by a third party who was never at the negotiating table – a buyer, a bank, a liquidator. A shareholders' agreement sits alongside them as a private contract, enforceable between its signatories but silent as far as anyone else is concerned. In other jurisdictions the constitution can absorb far more of the governance bargain directly, so a veto right or a transfer restriction written into the articles binds successors in a way the same clause in a side agreement never would. This is the single question a cross-border group has to answer before it relies on any term: does this clause bind the company, or does it only bind the two parties who signed the page it sits on. Corporate governance built on the wrong assumption survives quietly until the moment someone outside the original deal tests it.

The practical consequence is that the majority needed to amend the articles is not a detail to check once and forget. A shareholders' agreement that promises a minority investor a permanent veto is only as durable as the majority threshold the local constitution actually requires for the clause it depends on, and that threshold is set by the corporate law of the jurisdiction where the company is incorporated, not by the law the parties chose to govern their side agreement.

Board and director layer

Joint venture agreements routinely allocate board seats, set nomination rights for each co-investor, and list matters the board cannot decide without a specified majority or a specified shareholder's consent. None of that is self-executing. A nomination right written only into the joint venture agreement binds the shareholders who signed it to nominate a particular person; it does not, by itself, bind the company to appoint that person, and it does nothing to the duty the appointed director owes once in office. Directors owe their duties to the company, not to the shareholder who put them on the board, and a reserved-matters list drafted without acknowledging that split creates a conflict the joint venture agreement cannot resolve by contract alone. Shareholder rights of this kind need a route into the constitution itself if they are to bind the company rather than merely the shareholder who agreed to exercise them in a particular way.

This is where the board and director layer of governance work earns its place in the same review as the constitutional documents. A reserved-matters clause that looks identical on the page can be fully binding on the company in one jurisdiction and purely contractual, and therefore only as good as the signatory's willingness to comply, in another.

Records, registers and what becomes public

What becomes public and what stays private is set locally, and it is set separately from what the parties intended. The articles of association are, in almost every jurisdiction covered here, filed at a public register and available to anyone who asks – that is close to universal. The shareholders' agreement and the joint venture agreement are, in almost every jurisdiction, private contracts held by the parties and disclosed only if a party chooses to disclose them, or a court orders it. The registered office and the filed constitutional documents sit at the point where regulatory filing turns a private bargain into a public fact, and that is exactly where the risk in this practice concentrates.

An amendment lodged at the company's register cannot be reversed by agreement between the parties. It can only be corrected on the record, and the correction sits next to the original entry rather than replacing it, so anyone running due diligence sees both. A transfer restriction written into the articles becomes visible on the register the moment it is filed, and from that point it is a fact any counterparty is taken to have had the chance to check, whether they actually checked it or not.

Substance and where decisions are actually taken

A reserved-matters list is only as useful as the statutory floor beneath it. Some jurisdictions allow the shareholders considerable freedom to set their own majority thresholds and veto arrangements by contract; others impose a statutory minimum that no private agreement can reduce, whatever the parties intended. A joint venture agreement that promises a forty per cent shareholder a blocking right on capital increases is worth exactly nothing if the jurisdiction's corporate law fixes the relevant threshold below that level and treats any attempt to raise it by private contract as unenforceable against the company. Working out where a jurisdiction sits on that spectrum, for the specific decision the parties actually care about, is the substantive core of this practice – more central to the outcome than the drafting style of either document. Good corporate governance design starts from the statutory floor and builds the contractual layer on top of it, not the other way round.

The same logic applies to deadlock. A well-drafted deadlock mechanism assumes a route exists to force a resolution if the board or the shareholders cannot agree. Whether that route is a contractual buy-out clause, a statutory unfair prejudice remedy, or a petition to wind the company up on just and equitable grounds depends entirely on the jurisdiction, and a joint venture agreement silent on which route actually applies has not solved the deadlock problem – it has only described it.

Constitutional documents and shareholder arrangements

Articles, shareholder and joint venture agreements are not three independent documents that happen to be negotiated together. They are three layers of the same governance bargain, and the order of precedence between them matters more than the content of any single clause. Where the articles and the shareholders' agreement conflict, most jurisdictions give the articles priority as against the company and as against anyone dealing with it, while treating the conflict between the shareholders themselves as a separate contractual question that may still leave one party with a claim in damages, even if the clause they relied on never binds the company. A well-structured set of constitutional documents and shareholder arrangements is deliberate about which layer carries which promise, rather than repeating the same clause in both and hoping the repetition settles the priority question. Silence in the articles about a matter the shareholders' agreement covers in detail is sometimes the correct drafting choice, and sometimes the gap that undoes the whole arrangement – the distinction turns on whether the matter needs to bind the company or only the signatories.

A useful discipline is to ask, clause by clause, whether the parties actually need the term to survive a share transfer to someone who never read the shareholders' agreement. If the answer is yes, the term belongs in the constitution, or nowhere that matters.

Transactions and the governance conditions attached to them

Constitutional documents attach conditions to transactions long before anyone thinks of the joint venture agreement as a transaction document. Pre-emption rights, transfer restrictions, drag-along and tag-along clauses, and consent requirements tied to a change of control are all, in most structures, anchored in the articles rather than in the side agreement, precisely because they need to bind a buyer who has not yet appeared. A due diligence exercise that reviews the shareholders' agreement without checking whether the same restriction is mirrored in the filed constitution is reviewing the wrong document for the question it is trying to answer.

Once a share transfer is registered, a pre-emption right that should have been exercised first ceases to be available to the shareholder who was bypassed. The remedy that shareholder is left with – typically a claim for damages or, in some jurisdictions, an order requiring the transfer to be unwound – runs from the date of registration, not from the date the bypassed shareholder actually found out. That gap between the two dates is where claims are lost, not on the merits. Comparing how firmly a given jurisdiction actually enforces a shareholders' agreement against a non-signatory is therefore not an academic exercise; it is the difference between a transaction condition that holds and one that is decorative.

Duties, liability and disputes

A director appointed under a joint venture nomination right still owes duties to the company that appointed them, and those duties do not bend to accommodate the shareholder who nominated them. Where the reserved-matters list in the shareholders' agreement asks a nominated director to vote a particular way regardless of the company's interests, the director is caught between an instruction they may not be able to follow and a shareholder who expects loyalty. Company law in every jurisdiction covered here resolves that tension in favour of the company, though the mechanics of how a breach is pursued – derivative claim, statutory remedy, contractual indemnity – differ sharply.

Dispute resolution clauses in joint venture agreements are drafted as if the forum chosen will decide every question that arises. It will decide the contractual questions. It will not, on its own, decide whether a clause binds the company, because that question is usually governed by the law of incorporation regardless of what the parties chose to govern their contract. A joint venture agreement that sends every dispute to arbitration in one jurisdiction while the constitutional documents sit under the law of another is not wrong to do that, but the parties need to know, before a dispute arises, which forum actually has the power to grant the remedy they will want.

How to identify which of the above applies to your structure

The questions below are the ones worth answering before relying on any clause in an existing set of documents, not after a dispute has already started.

A group that already has a joint venture agreement in place, or that inherited one through an acquisition, usually finds the appointment terms buried inside it – nomination rights, removal rights, the conditions attached to a director's tenure. Those terms are frequently drafted once, at signing, and never looked at again, even as the shareholding or the board composition around them changes. Confirming what those terms actually say, and whether the constitution still supports them, is worth doing before a disagreement forces the question.

Review your appointment terms. Write to info@hreithlaw.com with the jurisdiction and the structure.

Services in this practice

Jurisdictions with verified requirements

Frequently asked questions

What happens if articles, shareholder and joint venture agreements are not reconciled with each other?
The documents keep operating separately, and the gap between them stays invisible until a transaction, a dispute or a due diligence exercise tests a clause the parties assumed was binding on the company. By then the fix usually requires the cooperation of parties who no longer agree on much.
How often should articles, shareholder and joint venture agreements be reviewed?
On any change to the shareholding, the board composition or the applicable corporate law, and otherwise on a fixed cycle rather than only when a transaction forces the question. A constitution that has not been checked since the last shareholding change is a constitution nobody has actually confirmed still matches the deal.
Does the position change for a foreign-owned company?
Yes, in two respects. The local corporate law still governs what the articles can do, regardless of who owns the shares, but a foreign parent often assumes its home jurisdiction's drafting conventions carry over, and they frequently do not. The gap shows up first in whichever clause the parent's own template treats as standard.
What does this work require in practice, beyond drafting?
Reading the articles and the shareholders' agreement side by side against the actual decisions a board needs to make, not against the clauses either document uses as headings. Most of the substantive risk sits in what the two documents do not say to each other, rather than in what either says on its own.
Who inside the company is actually responsible for this?
The board is responsible for the constitution the company operates under, and the shareholders are responsible for the private agreement between themselves, but in practice neither group monitors the interaction between the two unless someone is asked to. A director is not a formality that signs where indicated; the office carries a duty to know which document actually governs a given decision.
What evidence should the board keep on this?
A current, consolidated record of which version of the articles and which version of the shareholders' agreement is in force, together with a note of every point where the two diverge and why. That record is what makes a later dispute a matter of checking a file rather than reconstructing a negotiation from memory.

Make an enquiry

Or email us directly: info@hreithlaw.com