Shareholders' agreement in an SME: content and key clauses

A shareholders' agreement is a confidential contract between shareholders that organises governance, share transfers and exit. It is not mandatory, but no fund comes into the capital without one. The Companies and Associations Code allows it, within limits: voting agreements must be limited in time and consistent with the company's interest.

My thesis fits in one sentence. After a funding round, your power is not written in the share register; it is written in the agreement.

What does a shareholders' agreement contain?

Seven chapters, almost always in this order. Each one shifts part of the power your shares gave you before the fund came in.

1. Governance. Board composition, the fund's right to nominate one or more directors, an observer, frequency of meetings.
2. Reserved matters. Decisions requiring the fund's consent or a qualified majority: budget, investments above a threshold, borrowing, asset sales, management pay.
3. Information. Content and frequency of reporting.
4. Share transfers. Temporary lock-up, pre-emption, approval.
5. Exit. Tag-along, drag-along, intended timetable, buy-back procedure. The mechanics are set out in tag-along, drag-along and leaver clauses.
6. Managers. Exclusivity commitment, non-compete, good and bad leaver clauses.
7. Dilution protection. Right to take part in future rounds, anti-dilution and ratchet clauses.

The agreement does not stand alone. It fits together with the articles of association, and the boundary between the two documents decides what can be enforced against whom.

Transfer restrictions can sit in either. Written into duly published articles, they make a sale in breach unenforceable against the company and third parties, whether the buyer acted in good or bad faith (art. 5:67 for an SRL, art. 7:78 for an SA). In the agreement alone, they only give you a claim against the signatory who sold.

The thesis: 75% of the shares no longer means 75% of the power

Take the most common case in the files I have read: a fund comes in at 25%, the founder keeps 75%. On paper, the founder controls every general meeting decision requiring a three-quarters majority. In practice, the agreement removes from that scope everything it labels a reserved matter.

As a percentage, that gives the following table, built from the articles of the Companies and Associations Code and applied to a fictitious SME held 75/25.

The first line almost always comes as a surprise. In an SRL with no contrary provision in the articles, if you want to sell your 75% to a third party, the approval under article 5:63 is calculated excluding the shares you are selling. Only the fund is left, and in principle it alone makes up half the shareholders and three-quarters of the remaining shares. Your exit then depends on its consent, before a single line of the agreement has been signed.

The rule works both ways, which is the good news: the fund cannot sell its 25% to a third party without you either. Approval is not required, however, for a transfer to another shareholder, a spouse or legal cohabitant, or to direct ascendants and descendants.

Is it mandatory?

No. Nothing in the Companies and Associations Code requires a shareholders' agreement, and an SME can run for years on its articles alone. That is the serious objection to my thesis: if the law already governs approval, pre-emption and majorities, why negotiate a second, confidential document that binds only its signatories?

The objection holds for a stable family company with no outside investor. It falls away as soon as a fund comes in. A fund commits money it will need to get back, and the articles say nothing about how it will exit, what it can block in the meantime or what happens if you leave. Those are precisely the subjects of the agreement. In the files I have read, no fund entered the capital without one.

The nuance lies elsewhere. Part of what the agreement provides must also appear in the articles to be effective. A pre-emption right lodged only in the agreement can be breached at the cost of damages; written into published articles, it makes a sale in breach unenforceable. For every point you negotiate, check which document it ends up in.

Three legal safeguards, and everything else left to contract

The framework is federal: the Companies and Associations Code, consolidated on ejustice, applies in the same way to a Charleroi SME bringing in Sambrinvest, a Brussels company raising with finance&invest.brussels (the Brussels-Capital Region's public financing body) and a Flemish business dealing with PMV. From one region to the next, the counterpart changes, not the rule.

Three safeguards matter to the founder:

• Voting agreements. Shareholders can undertake to vote a certain way, for instance to appoint the director the fund nominates. They must be limited in time and not contrary to the company's interest. Void in particular are those in which a shareholder undertakes to vote as instructed by the company itself (art. 5:46 SRL, art. 7:56 SA).
• Lock-ups. In an SA, a lock-up must be justified by a legitimate interest, particularly as to its duration; if indefinite, it can be terminated with reasonable notice (art. 7:78).
• Approval and pre-emption in an SA. They cannot block a sale for more than six months from the approval request or the invitation to exercise pre-emption (art. 7:78).

The rest is contractual: the list of reserved matters, the price under the leaver clauses, the drag-along threshold. That is where the negotiation happens, and where the law does not protect you.

Which clauses protect the founder?

The ones you ask for. The fund arrives with its own template, designed to protect its investment; clauses favouring the founder only appear if you have them added. The ones that change your position most:

• a closed list of reserved matters, with thresholds in euros, and no wording such as "any significant decision";
• a fixed term for the agreement and for each voting agreement;
• a tag-along letting you sell at the same price as the fund if it sells its stake;
• a drag-along with a trigger threshold and a floor price;
• a narrow definition of bad leaver that excludes illness, disability and dismissal without cause;
• a deadlock procedure: discussion between senior people on both sides, mediation, then as a last resort a buy-out mechanism;
• the death or incapacity of a shareholder-manager, provided for in plain words.

The fund's horizon shapes the exit clause, and it is published. Checked on each fund's website on 15.10.2026: ORSA (formerly Investsud, a Walloon evergreen investment fund) imposes no exit horizon, as an open-ended evergreen fund; Invest.BW (the Walloon Brabant provincial investment company) states interventions of typically 5 to 10 years; Theodorus II and III are closed-end funds with a 12-to-14-year life. A closed-end fund has an end date, and its agreement will almost always include an exit timetable and a tighter drag-along. ORSA also presents shareholders' agreements as an area of expertise in its own right for SMEs (ORSA's shareholders' agreement page). The ranking of SME funds sets these horizons side by side with the tickets.

Three mistakes I have seen reading agreements

The first: copying a start-up template for a family SME, or the reverse. A start-up agreement anticipates successive rounds and an exit through a sale; a family SME needs rules for succession and buy-back.

The second: letting the articles and the agreement contradict each other. On the day of the dispute, each side brandishes the document that suits it, and the enterprise court decides. Nobody wins.

The third, and the most expensive: signing without a deadlock procedure. The founder and the fund will one day disagree on a reserved matter. Without a mechanism in place, the company stands still until judgment.