When a shareholder leaves and the protective clauses in a share sale agreement

When a co-shareholder tells you they are leaving, the first draft of the share sale agreement usually follows within the week. Most of it will be devoted to price, warranties and completion mechanics, but the clauses that decide what your business actually keeps sit quietly at the back: confidentiality, intellectual property and restraint of trade. This article walks through those protective clauses, the drafting choices behind each one, and the traps that make them fail.

A share sale agreement transfers the departing shareholder's shares to the remaining shareholders, a new investor or the company itself. Unlike an asset sale, the company's assets do not change hands. The goodwill, client relationships, trade secrets and intellectual property all stay with the company. What the agreement does is stop the seller from carrying the value of those assets out the door. It also has to work alongside the constitution and any shareholders' agreement, which operate as a contract between the company and its members under s 140 of the Corporations Act 2001 (Cth). If those documents contain preemptive rights or transfer restrictions, the sale process must follow them before any of the clauses below come into play.

The confidentiality clause

The confidentiality clause is the foundation of the whole document. It stops the departing shareholder from using or disclosing the business's confidential information after completion. A shareholder who also acted as a director or worked in the business will have seen customer lists, pricing, financial results, business plans and trade secrets, and that knowledge does not evaporate when the shares transfer. The clause has two jobs: defining what counts as confidential information, and setting out what the seller may and may not do with it. The drafting choices come down to three elements:

Definition: The clause should cover all information relating to the business, in any form and whether or not it is marked confidential, not just a list of examples. A definition that names specific documents and stops there invites the argument that anything not listed was never protected.
Obligation: The seller should be prohibited from using or disclosing the information, directly or indirectly, for any purpose other than completing the sale.
Exceptions: Every clause needs carve-outs for disclosure required by law, a court or a regulator, and for information already in the public domain through no fault of the seller. Courts use the same language; in Cactus Imaging Pty Ltd v Peters [2006] NSWSC 717 the injunction restraining disclosure of confidential information was expressly qualified "except as may be required by law".

The traps are mostly in the detail. The clause needs a survival period so the protection continues after completion. It should require the seller to return or destroy confidential material on exit, including copies on personal devices and in cloud accounts. And it should be drafted so the seller's new employer is caught too, otherwise a seller who joins a competitor can argue the obligation binds only them, not the people they now work with.

The intellectual property assignment clause

In a share sale the company keeps its assets, but ownership of intellectual property is not always clean. The departing shareholder may have written code, drafted branding, built the website or designed the product outside the scope of any employment or consultancy arrangement, which means the copyright sits with them personally rather than with the company. The assignment clause fixes that gap, and a well-drafted one covers four elements:

What it assigns: A full definition of intellectual property, covering copyright, trade marks, patents, designs, know-how and domain names, followed by an assignment of all of it to the company.
Future assignment: Not every piece of IP can be identified on signing day. The clause should also assign IP created before completion that has not yet come into existence or been identified, and require the seller to sign any further documents needed to perfect the transfer.
Formalities: An assignment of copyright does not take effect unless it is in writing signed by or on behalf of the assignor: s 196(3) of the Copyright Act 1968 (Cth). Registered rights such as trade marks and designs also need the assignment recorded with IP Australia, or a later dealing can defeat the company's ownership.
Warranties: The seller should warrant that no third party owns or claims an interest in the IP being assigned, and that the company has not previously licensed it away.

If the seller refuses to assign particular IP, the buyer might accept a licence instead. That is a poor substitute, because the seller keeps ownership and can dictate terms later, so it should only ever be a fallback for genuinely disputed assets.

Copyright carries more than economic rights. Under Part IX of the Copyright Act 1968 (Cth), an author has moral rights: the right to be attributed as the author, the right of integrity (not to have the work subjected to derogatory treatment) and the right not to have authorship falsely attributed. These rights stay with the author even after the copyright itself is assigned, so the assignment clause alone does not neutralise them. A departing shareholder who created the company's logo or wrote its software can later object to how the company uses that work.

The solution is a written consent. Under s 195AWA of the Copyright Act, it is not an infringement of a moral right to do something within the scope of a written consent genuinely given by the author. The share sale agreement should contain that consent, wide enough to cover all acts the company might do with the IP, now and in the future. Two drafting points matter: the consent must be in writing, and it must be genuinely given. A consent obtained by duress, or by false or misleading statements, has no effect: s 195AWB. This is the clause most often missing from a first draft, and the easiest for a departing shareholder to use as leverage later.

The restraint of trade clause

The restraint of trade clause stops the seller walking out and setting up, or joining, a competing business. It is also the clause most likely to end up in court, because at common law a restraint of trade is prima facie void and unenforceable unless the party seeking to enforce it proves the restraint is reasonable in the interests of the parties and the public. That means the restraint must protect a legitimate interest of the business, such as goodwill, client connections or confidential information, and go no further than is reasonably necessary to protect it.

The drafting choices come down to three dimensions:

Activity: What the seller cannot do, such as carry on, or be involved in, a business that competes with the company.
Geography: Where the restraint applies, such as Australia, New South Wales or a particular city or region.
Duration: How long it lasts, such as three, two or one years.

Because no one can predict what a court will find reasonable, well-drafted clauses use cascading options: separate restraints for each combination of activity, area and period, each expressed to stand alone if the others fail. If a court strikes down the three-year nationwide restraint, the one-year Sydney restraint survives.

The position is more forgiving in New South Wales. Under s 4(1) of the Restraints of Trade Act 1976 (NSW), a restraint is valid to the extent to which it is not against public policy, whether or not it is in severable terms, so a court can read an over-broad restraint down rather than strike it out entirely. The same Act penalises lazy drafting: s 4(3) allows a court to declare a restraint invalid where there has been a manifest failure by the party who created it to attempt to make it a reasonable restraint. In practice that means the wider the clause, the more carefully it has to be justified on the face of the document.

The non-solicitation clause

A full non-compete is a blunt instrument, and courts are more likely to uphold a clause that targets specific harm rather than competition in general. The non-solicitation clause stops the seller poaching clients, staff and suppliers:

Clients and suppliers: The seller should not approach, solicit or deal with clients or suppliers they dealt with during their involvement, often limited to those with whom they had dealings in the year or two before exit.
Staff: The seller should not induce, or attempt to induce, employees to leave the company, and should not employ or engage anyone who has left.
Indirect conduct: The clause should catch acting through others. In Provida Pty Ltd v Sharpe [2012] NSWSC 1041 the court held that a restraint on acting "directly or indirectly" covered a former employee providing services through a newly established one-person company, and that a prohibition on "dealing with" clients can extend to accepting their business rather than merely soliciting it.

Non-solicitation clauses are often the most enforceable part of the suite because they map directly onto identifiable interests: the client connection the seller built up, and the confidential staff and client information they carried out.

Survival, enforcement and acknowledgements

None of this protects the business if the clauses die with the agreement. A survival clause states that the confidentiality, IP assignment, restraint and non-solicitation obligations continue after completion. The enforcement machinery matters just as much:

Remedies: Damages are hard to quantify for a leaked client list or a poached team, so the agreement should record that the parties accept an injunction is an appropriate remedy and that the business may seek urgent interlocutory relief without proving damages.
Acknowledgements: The seller commonly acknowledges that the restraints are reasonable and necessary to protect the business and that they have received independent advice. That does not bind a court, but it is evidence the court can weigh in the reasonableness assessment.
Costs: A clause entitling the business to recover its enforcement costs discourages a seller from treating the restraint as optional.

Optional clauses worth considering

Depending on how the exit came about, a few more clauses are worth raising:

Preemptive rights and transfer restrictions: If the constitution or shareholders' agreement gives remaining shareholders first refusal over the departing shareholder's shares, the sale has to follow that process before any third-party buyer can be introduced.
Drag-along and tag-along rights: Useful where the exit is part of a sale of the whole company, forcing or protecting minority sellers in the same deal.
Price adjustment and earn-out: Where part of the price depends on the business's performance after completion, usually paired with a warranty that the seller will not act to depress that performance.
Escrow and indemnity caps: A slice of the price held back for a period to back the warranties, with a cap on the seller's total liability.
Board and authority changes: The seller's resignation as director, removal from bank mandates, signing authorities and company registers, so a departing shareholder cannot keep operating the business by accident.

An experienced commercial lawyer will not start with the price. We would map the departing shareholder's actual role in the business against the confidentiality definition, checking whether it captures the information they genuinely hold. We would trace the chain of title to the IP, including work created by contractors and consultants, and flag anything that needs a separate assignment before completion. We would stress-test the restraint against what the business truly needs: a restraint protecting client connections looks different from one protecting a trade secret, and a court will treat them differently.

The order of negotiation matters too. The price can be argued later; the protective clauses are harder to strengthen once the seller has signed and the goodwill has started walking out the door. We would insist on the survival clause before anything else, push the restraint down to the narrowest defensible scope, and make sure the moral rights consent is in writing before completion, because it cannot be cured after a dispute starts. If the document has been drafted by the buyer's lawyers, the seller's interests in the same clauses are different and just as important to protect, which is where independent advice earns its fee.

Where the exit turns into a dispute

If the exit goes sour, the dispute will almost always turn on the restraint of trade clause. The confidentiality clause defines the asset, the IP clause fixes ownership, but the restraint clause is the one a court is asked to enforce, and its enforceability was settled the day the agreement was drafted. A restraint that is calibrated to a genuine legitimate interest, expressed in cascading alternatives and documented as reasonable on its face, gives the business a remedy it can actually use. One drafted by copying a template, with no thought to the seller's role or the business's real exposure, gives the seller's lawyer an argument that it is void and a departing shareholder a licence to compete.

The key points to take away: the confidentiality definition should be broad and survive completion; the IP assignment must be in writing and backed by a moral rights consent; the restraint should be no wider than the legitimate interest it protects, in cascading form, and read down under the Restraints of Trade Act 1976 (NSW) where that applies; and the non-solicitation clause is often the most enforceable weapon in the suite. If you are drafting, or have just received, a share sale agreement for a departing shareholder, have a lawyer review those clauses before anyone signs.