Two founders, one idea, a capital to split. Then comes the question that seems trivial and never is: fifty-fifty, or does one of the two take a single share more? Many decide in thirty seconds, over a coffee. It is often the heaviest decision of the whole incorporation, and the one most regretted later.
Because the real issue is not the percentage. It is what the shareholders’ agreement — the convention d’actionnaires, in Swiss legal language — does with that percentage. A poorly built 51/49 protects less than a well-designed 50/50. Here is why.
50/50: perfect equality, perfect deadlock
On paper, 50/50 is the legal translation of trust. Two shareholders on equal footing, neither above the other. As long as both say yes, everything runs.
The trouble starts at the first no. At the general meeting, no majority emerges: resolutions are passed by the absolute majority of the votes attached to the shares represented (art. 703 CO) — those present or represented, not all issued shares. And 50 % against 50 % is not a majority. On the board of directors, if the two founders sit there alone, same impasse. Nothing moves. The company is alive but paralysed.
This is the deadlock. And it is not confined to major decisions: a disagreement over a key hire, a budget, the renewal of a lease, and the business stalls. I have seen two equal partners let a major contract slip because neither would yield on principle. The competitor signed it.
51/49: the illusion of control
Hence the reflex: give one of the two a single share more. 51/49. That way there is a boss, and the deadlock disappears. That is what most founders believe. It is wrong, or at least very incomplete.
The 51 % majority holder gains one thing, and one only: a simple majority at the general meeting (art. 703 CO). Enough to pass routine matters — approving the accounts, appointing a director. Useful, but far from everything.
First limit: the decisions that really matter escape the 51 %. Amending the corporate purpose, increasing capital, transforming or dissolving the company: the law requires a qualified majority, namely at least two thirds of the votes represented and the absolute majority of the nominal values represented (art. 704 CO). This double requirement — votes and capital — is a Swiss specificity. With 51 %, the majority holder does not reach it. On those subjects, the 49 % minority holds a de facto veto. He signed nothing special; the law gives it to him.
Second limit, more insidious. Operational power is not played out at the meeting, it is played out on the board. Yet if the board is made up of the two founders and its decisions require the presence — or the agreement — of both, the 51/49 of the general meeting is worthless at that level. The majority holder thinks he governs. In practice, the moment his partner objects on the board, we fall back into the 50/50 deadlock. The percentage has changed; the impasse has not. A casting vote granted to the chair in the organisational regulations can reverse this — provided the founders agreed, in calm times, on who chairs.
In short: 51/49 is not an improved 50/50. It is a different balance, with its own blind spots.
Three documents, three logics: articles, regulations, agreement
To see where the real protections sit, one must distinguish the layers of any Swiss company limited by shares.
The articles of association first. A notarised, public act, open to anyone at the Commercial Register (art. 626 CO). They set the architecture: capital, share classes, board organisation, quorums, transfer restrictions (art. 685a and 685b CO). What touches the very structure of the company lives here, under mandatory law — you do not put whatever you like in articles.
The organisational regulations next (art. 716b CO). An internal document adopted by the board, not public. This is where the operational detail is settled: allocation of tasks, signatory powers, the chair’s casting vote. It sits halfway between the public articles and the private agreement, and it is almost always forgotten.
The shareholders’ agreement last. A private, confidential contract between the partners. This is where the fine mechanisms live: voting undertakings, contractual veto rights, exit clauses, deadlock resolution, non-competition. Flexible, tailor-made, invisible to third parties.
The distinction has practical consequences. A pre-emption clause written into the articles binds everyone, erga omnes, because it is published; the same clause in the agreement alone binds only the signatories, inter partes. Conversely, a detailed voting pact has no place in public articles. Knowing what goes where is half the work — and the most common mistake in templates found online.
The deadlock: planning the divorce during the honeymoon
Whether the split is 50/50 or 51/49 with a parity board, the risk of deadlock exists. The question is not to eliminate it — one cannot always — but to plan its exit. In calm times, while the two partners still get along.
The mechanisms exist, and are chosen according to the degree of violence one accepts. A good-faith negotiation phase, then mediation, for ordinary disagreements. An expert-arbitrator who settles a specific decision, when the subject is technical. And for deadlocks that threaten the survival of the structure, the so-called buy-sell clauses, of which the ‘Russian roulette’ is only one variant: one partner names a price per share, the other chooses to buy or sell at that price. Brutal, but it forces an outcome rather than slow asphyxiation. These clauses form a family, and their exact terms must be drafted with care — a poorly calibrated roulette mechanically favours the partner with the deeper pockets.
One safeguard, not to be overlooked. An agreement cannot bind a shareholder indefinitely: an undertaking that restricts his freedom excessively, by its duration or scope, falls under art. 27 para. 2 CC. The Federal Supreme Court has held that an agreement heavily restricting, after thirty years, a shareholder’s freedom in his business-succession planning could be deemed excessive (ATF 143 III 480). Excessiveness is assessed at the moment the undertaking is invoked, not at signature. In other words: a clause locked too tightly may, when the day comes, no longer hold.
The hurried founder’s reflex is to wave all this away with a ‘we’ll see, we get along fine’. Yet no one signs an agreement thinking of conflict. One signs it precisely for the day when the getting-along stops — and on that day, it is too late to negotiate the rules.
The exit clauses you write before you need them
Beyond deadlock, a good agreement anticipates departures. They will happen, in one form or another.
The good leaver / bad leaver distinction structures everything. The legitimate departure — retirement, disability, death, dismissal without fault — opens a right to buy back the shares at market value, as assessed by an expert. The wrongful departure — serious breach of the agreement, unfair competition — triggers a buy-back at a discount, sometimes steep. The difference in treatment is not punitive in principle; it protects those who stay and deters conduct that endangers the company.
To these are added the share-circulation clauses. The tag-along, which lets the minority exit on the same terms as the majority if the latter sells to a third party — protection against ending up in business with a stranger. The drag-along, which lets one or more shareholders holding an agreed threshold — often a large majority, to be negotiated — pull the others into a 100 % sale, at a negotiated floor valuation: without it, a single share would block the sale of the whole company. The trigger threshold is the real point of negotiation here. A lock-up period, finally, during which no one sells, giving the project time to take hold.
None of these clauses is spectacular. Taken together, they head off the scenarios that send partners to court.
The minority is not defenceless — if it negotiates
A word for the 49 % partner, or the 30 %, or the 10 %. The percentage is not destiny. A minority that negotiates its agreement well obtains protections the majority cannot circumvent.
A contractual veto right over a list of strategic decisions, through a voting undertaking. Qualified quorums written into the articles, raising the threshold above the legal minimum and making its consent indispensable. An enhanced information right, beyond the minimum of art. 697 CO. A joint bank signature above a certain amount. A guaranteed seat on the board.
In a recent matter, a 49 % minority obtained the mandatory joint signature of any banking operation above a defined threshold. On paper, he did not control the company. In practice, nothing important happened without him. A contrario, a minority who signs the standard template without negotiating it has, in fine, only his eyes to count the decisions taken without him.
One safety net the law stretches both ways: the prohibition of abuse of rights (art. 2 CC). The majority, even at 51 %, cannot exercise its rights to harm the minority; the minority cannot divert its veto from its purpose either. An agreement compelling a director to act against the company’s interest would not bind that director, who would instead incur personal liability by complying (ATF 145 III 351). The contract stops where the legal duties of the corporate bodies begin.
So, 50/50 or 51/49?
The real answer disappoints those looking for a number: it depends, and above all, it matters less than one thinks. A 50/50 paired with a clear deadlock clause and exit mechanisms works better than a bare 51/49, where the majority holder discovers too late that he controls neither the qualified decisions nor the board.
The percentage sets a theoretical balance of power. The shareholders’ agreement decides what is made of it — who really decides, how a deadlock is broken, on what terms one enters and leaves the capital. It is the agreement, not the number, that protects the company and its founders.
The real cost was never that of a well-drafted agreement. It is the cost of its absence, the day two partners who once valued each other face a contractual silence and a court.
Legal basis and case law
Company limited by shares and incorporation: art. 620 et seq. and 626 CO. Ordinary majority at the general meeting: art. 703 CO. Qualified majority (votes and capital): art. 704 CO. Transfer restrictions: art. 685a and 685b CO. Organisational regulations: art. 716b CO. Shareholder’s information right: art. 697 CO. Non-transferable powers of the board: art. 716a CO. Dissolution for good cause: art. 736 CO. Good faith and abuse of rights: art. 2 CC; excessive commitment: art. 27 para. 2 CC. Federal Supreme Court: ATF 143 III 480 (excessive commitment of a shareholders’ agreement), ATF 145 III 351 (limits of contractual power and legal duties of corporate bodies), ATF 109 II 140 (dissolution for good cause).
About the author
Matthias Traussnig is an attorney at the Geneva Bar and founder of Sentinel Legal. Holder of the CAS in Digital Finance Law from the University of Geneva, he practises corporate law, economic crime and technology law.
On this type of matter
Sentinel Legal advises on company formation and the drafting of shareholders’ agreements in Geneva and French-speaking Switzerland, from capital structuring to exit clauses. Before fixing a split, a conversation beats a downloaded template: +41 22 512 76 00 or via the contact form.