Three founders. One million dollars. A shareholder agreement drafted in Mandarin around a kitchen-table joint venture — long on goodwill language, short on everything that matters. When the business failed and the money-lending founder came to collect, the ensuing litigation ran through winding-up proceedings, statutory demands, bankruptcy complaints, two High Court actions and a nine-day trial — and produced a reported judgment dismantling every recovery argument he advanced.
Salem Ibrahim LLC acted for the two founder-shareholders who were sued. In Ren Xin Wu v Lee Kuan Fung [2025] SGHC 63, the High Court dismissed both claims in their entirety, with costs of $320,000 across the two actions awarded to our clients — and the judgment reads as a checklist of everything a DIY shareholder agreement fails to prevent.
The firm is rated 5 stars for Alternative Dispute Resolution in The Straits Times/Statista Singapore’s Best Law Firms 2027 survey — and is ranked No. 1, the only 5-star firm, in Aviation, Aerospace and Space Law, with 5 stars also in Family Law.
Here are the lessons.
Lesson 1: An oral guarantee is worth nothing. Write it down.
After the company defaulted, the lending founder claimed the other shareholders had orally agreed — in recorded phone calls — to be personally responsible for the loan. The court’s answer was blunt: under s 6(b) of the Civil Law Act 1909, an agreement to answer for the debt of a third party “is required to be in writing and signed by the party in question.” An oral promise of this nature “simply would not suffice under law.” If you want a guarantee from your co-founders, it exists only on paper, signed — never in a phone call, never in a nod, never in a WeChat message.
Lesson 2: The words in the agreement mean what they say.
The agreement recorded the $990,000 loan as extended on a “no guarantee, interest-free basis.” The lender then spent years arguing the court should imply a term making the other shareholders procure repayment — effectively converting “no guarantee” into a full guarantee. The court refused, applying the three-step test in Sembcorp Marine v PPL Holdings: one “cannot have two contracts, one written and the other implied, contradicting each other.” An express term saying “no guarantee” is the end of the discussion. DIY drafters beware in both directions: the clause that protects your co-founder from liability is the same clause that leaves your loan unsecured.
Lesson 3: Missing stamps and seals won’t save you.
The defenders initially argued the agreement never took effect because cl 9.3 required signatures and affixed stamps — and no stamps were ever affixed. The court rejected this emphatically: contract validity is an objective inquiry, and “formalities do not overshadow the conduct underscoring the parties’ intentions.” Where the parties conducted themselves for three years as if bound, the agreement is binding — ceremony clauses like “affix stamps” are not escape hatches. Conversely, don’t rely on them as safety valves either. Substance wins.
Lesson 4: Directors owe duties to the company — not to you personally.
The lender argued the managing founder owed him — a specific shareholder — implied duties of honesty, care and loyalty, actionable by him personally. The court held there is no such thing: the duties under s 157 of the Companies Act are owed to the company, are “non-delegable and mandatory,” and any breach “must be the subject of an action by the company, as opposed to an action by a specific creditor.” A shareholder who feels wronged by a director cannot sidestep this by drafting a “duties” clause into the shareholders’ agreement. The remedies lie elsewhere — and a well-advised shareholder knows which doors they are.
Lesson 5: An agreement that ends with a shrug is no agreement.
The agreement was to be “valid for three years” — and said nothing about what happened after. The court’s dry observation: an agreement to agree “is, with respect, no agreement at all.” Every shareholder agreement needs an exit architecture: what happens on expiry, on deadlock, on a founder’s departure, on insolvency. This one had feel-good platitudes where those clauses should have been.
Lesson 6: Undocumented nominee arrangements poison everything.
One founder’s shares were held by a nominee; another transferred his shares quietly in 2018 without telling the lender, leaving him believing a former shareholder was still invested. The court found the transfer was real — and that the lender “appears to have been led, by way of omission of material facts, to wrongly assume” continued shareholding. A paper arrangement like that breeds exactly the suspicion that detonated into this litigation.
Lesson 7: Evidence discipline — half a recording is worse than none.
The lender supported his case with secretly recorded calls — but produced only carefully selected extracts. Ordered to disclose the full recordings, he did — and the full tapes destroyed his case, recording his co-founder saying, repeatedly and plainly: “My bottom line is that the company will bear these debts, and the company will repay these debts.” Selected disclosure in litigation is not strategy; it is self-destruction. And even at its highest, the court noted the partial material was “non-probative.”
Lesson 8: Even a winning argument needs a provable loss.
Finally, the court observed that even had the lender won on the term, damages failed: he led no evidence of what he would actually have recovered from an insolvent company. Suing for the full $990,000 ignores that the remedy is worth what the company can pay — and proving that requires evidence, not arithmetic.
The pattern beneath the lessons
Read together, the judgment identifies the disease: an agreement “long on language but… short of clauses that add value” — clauses that “proffer feel-good platitudes that add nothing to its legal contours.” Three friends could draft it; no lawyer did. When the money was lost, there was no machinery to handle it, so the parties built the machinery in court, at ruinous cost.
The fix is not more paper — it is the right paper. Founder loans documented with security or guarantees, properly executed. Exit and deadlock provisions that actually operate. Nominee arrangements with proper declarations. And when the relationship sours, remedies pursued through the correct doors, with complete evidence.
We know because we defended this case — and won.
What a trial lawyer actually brings
Clients often ask what a trial lawyer actually brings. Here is the honest answer, from a case we ran: the outcome turned less on who argued louder in court than on three engineered concessions — exposed selective disclosure, a pinned-down legal theory, and an abandoned remedy. Every one was foreseeable, prepared for, and executed on. The submissions that followed were, by then, almost clerical.
Choosing litigation counsel is choosing whether your dispute will be conducted or designed. Ask prospective counsel a simple question: “What admission are you trying to get in cross-examination, and what does it unlock?” If there’s no crisp answer, there’s no strategy — there’s only a proceeding.
We are, in the words of our website, Counsel and Strategy with Good results. This case shows what that means: the record built before trial, the admissions engineered during it, and the law applied once the groundwork made it decisive.
The ending both sides deserved less
The claim was dismissed in its entirety, and the court ordered the claimant to pay $320,000 in costs to the defendants collectively across the two actions — $200,000 for the main trial action and $120,000 for the related proceedings. The court declined to award indemnity costs, observing that neither side was entirely blameless in the events leading up to the suits — a reminder that even a clear winner rarely walks away clean.
The lesson is not that litigation always ends well. It is that this litigation should never have been necessary. Three friends could draft that agreement. No lawyer did. When the money was lost, there was no machinery to handle it, so the parties built the machinery in court — at a cost measured in hundreds of thousands.
A shareholder dispute brewing, or a shareholder agreement that deserves a professional eye before it’s tested? Speak to the team that litigated this one to judgment — call (65) 6226 1233 or request a callback.
FAQ
Can a shareholder be personally liable for a company's debts in Singapore?
Generally no. Shareholders benefit from limited liability, and courts will not imply personal responsibility for company debts unless it is expressly agreed — and any personal guarantee must be in writing and signed under s 6(b) of the Civil Law Act 1909 to be enforceable.
Is an oral guarantee enforceable in Singapore?
No. Under s 6(b) of the Civil Law Act 1909, an agreement to answer for the debt of a third party must be in writing and signed by the party giving the guarantee. Oral promises of personal liability are unenforceable, as confirmed in Ren Xin Wu v Lee Kuan Fung [2025] SGHC 63.
Do directors owe duties to individual shareholders in Singapore?
No. Directors’ statutory and fiduciary duties under s 157 of the Companies Act are owed to the company itself, and breaches must be pursued by the company — not by individual shareholders or creditors through private claims.
Are secret audio recordings admissible in Singapore courts?
They can be, but courts weigh them critically. In Ren Xin Wu v Lee Kuan Fung [2025] SGHC 63, selected extracts of secretly recorded calls were found to have been taken out of context, and the full recordings — ordered disclosed — undermined the recording party’s own case.
Is a shareholder agreement valid if stamps or seals were never affixed?
Yes, if the parties’ conduct shows they treated it as binding. Singapore courts apply an objective, substance-over-form inquiry — a ceremony clause requiring affixed stamps will not invalidate an agreement the parties performed for years.
What does a shareholder dispute cost in Singapore?
Costs vary with complexity and duration, but contested shareholder litigation is expensive. In Ren Xin Wu v Lee Kuan Fung [2025] SGHC 63, the court awarded the defendants $320,000 in costs collectively across the two actions — $200,000 in the main action and $120,000 in the related action — illustrating how quickly a dispute over a poorly drafted shareholder agreement escalates.
