The notice went up on a Tuesday morning, taped to the door before the first shift arrived. Within the week, every outlet in the group was closed — including the ones that had been profitable every single month.
No calls the night before. No meeting. Staff found out they no longer had a job the same way the group's customers found out the outlet was gone — by standing in front of a locked door.
Eighteen months earlier, the same group's investors had approved an aggressive expansion plan with real enthusiasm. Double the footprint. Prove the concept could scale beyond its first few outlets. The growth story was genuinely good. Nobody in that boardroom asked what would happen to the profitable outlets if the unprofitable ones went wrong — and the answer, when it arrived, was that the whole structure came down together.
The financial problem: growth built on a margin nobody stress-tested.
Singapore's hospitality and F&B economics are unforgiving in a way that's easy to underweight when you're only looking at topline growth. Typical margins for an established concept here run somewhere in the range of five to seven percent — thin enough that a single bad quarter of rent increases, staffing gaps, or a slow tourist season can erase a year of profit. The same concept, expanded into a regional market with lower rent and labour costs, can plausibly run margins of twenty to thirty percent. That gap is exactly why so many operators eventually look outward. This group did the opposite first: it expanded harder domestically to prove scale, in a market where the underlying unit economics were already the thinnest in the room. Worse, the group reported consolidated numbers to its board — one blended topline, one blended margin — which meant two or three structurally unprofitable outlets could hide comfortably inside the average of several genuinely good ones, for a long time, right up until they couldn't.
The human capital problem: the cost curve nobody priced into the expansion plan.
Singapore's Dependency Ratio Ceiling caps foreign workers at 35% of headcount in the services sector, and the levy on each Work Permit holder isn't flat — it rises in tiers as a company's foreign-worker share climbs toward that ceiling. A basic-skilled worker inside the first 10% of headcount costs a few hundred dollars a month in levy. The same worker, hired once the company is already sitting in the 25–35% band, costs meaningfully more — the marginal hire gets more expensive precisely when a fast-expanding operator needs to hire the most. Each new outlet needed staff faster than local hiring could realistically supply them, which pushed the group's foreign-worker share upward outlet by outlet, quietly sliding the whole group into a higher levy tier without anyone tracking it as a line item tied to the expansion plan itself. Layer the Progressive Wage Model's mandatory wage floor and the Local Qualifying Salary increase on top, and the true, fully-loaded staffing cost of outlet number twelve was materially higher than outlet number three — a cost curve baked into national policy, fully knowable in advance, and entirely absent from the financial model the board had approved.
The legal problem: one security structure, engineered to fail as a group.
The expansion had been funded through secured bank facilities, cross-collateralized across the group's outlets and its holding entities, with a personal guarantee from the founding director. Cross-collateralization is a completely standard feature of SME lending in Singapore — it's often the only way a growing hospitality group gets the facility at all. What it also means, and what rarely gets explained in board terms, is that a default anywhere in the group gives the lender rights across the whole security package, not just against the specific outlet that missed its numbers. When two underperforming outlets fell behind on rent and breached a loan covenant, the bank's position wasn't limited to those two leases. Real cases in this exact market show what that actually looks like in practice: letters of demand with repayment windows measured in days, the right to sell shares in subsidiary entities, and the ability to enforce against mortgaged assets held anywhere in the structure — a chain reaction that can turn one bad outlet into a group-wide liquidation within a matter of weeks. Landlords, seeing the same signal, tend to file their own claims at the same time rather than wait their turn, and staff wage claims — while they do rank ahead of unsecured creditors under Singapore's insolvency framework — still mean employees are last to actually see the money, discovering the closure by locked door long before anyone tells them their claim is even in the queue.
The governance problem: the board approved a number, not a structure.
Nobody on the board had asked to see the security structure underneath the expansion facility, because nobody had framed it as something the board needed to approve. The growth roadmap got board sign-off. The financing that made it possible was treated as a management-level, banking-relationship detail — exactly the kind of decision that feels too operational for a board agenda, right up until it turns out to be the single most consequential structural risk the company carries.
The expansion problem: racing to prove scale in the wrong market first.
The deeper strategic error sat underneath all of it. Singapore's F&B closure rate has been running at its fastest pace in roughly two decades, driven by exactly this margin compression — rising rents, tightening labour policy, and thin local economics that reward operators who expand regionally rather than operators who expand harder at home. Several of this group's more disciplined competitors had already begun opening outlets in neighbouring markets specifically to escape these constraints, chasing the twenty-to-thirty percent margins available there while Singapore operations stayed lean and steady. This group did the reverse, doubling down domestically to build a growth story for its investors before ever testing whether the concept traveled — leaving it maximally exposed, on maximally thin margins, in exactly the market where the ceiling was lowest and the downside risk of a cross-collateralized default was highest.
What actually got fixed.
Brought in the week the first covenant breach surfaced, before the cascading closures began, the mandate was to stop the structure from taking down outlets that had never done anything wrong.
The first move was renegotiating directly with the lender to unwind the cross-collateralization — ring-fencing security by outlet rather than across the group, so that a failure in one location could no longer trigger enforcement rights against the others. Banks will generally engage seriously on this once a business shows it understands its own numbers well enough to propose a workable alternative; this one did, largely because the second move happened at the same time. Outlet-level profit and loss reporting went to the board monthly for the first time, ending the consolidated-average reporting that had let the underperforming locations hide inside the group's blended numbers. Third, the foreign-worker cost curve got built directly into the model for any future outlet: before signing a new lease, the plan now had to show what tier of levy that outlet's staffing need would push the group into, and whether the unit economics still worked once that real number was in the model rather than assumed away.
The two structurally unprofitable outlets were closed anyway — but on notice, in an orderly wind-down that paid staff through their statutory entitlements and gave landlords a real handover rather than a locked door. The remaining outlets, no longer carrying two loss-making locations inside a blended average nobody could see clearly, showed a healthier margin within two quarters than the group's full consolidated numbers had ever displayed.
The expansion wasn't the mistake. Financing it as one indivisible structure, and reporting it as one indivisible number, was — and by the time anyone asked to see the security package, the answer was already written into contracts nobody at the board table had read.