The lowest tender price is paid for later, in shortages
Southeast Asian governments buy most of their generic medicines through tenders that reward the lowest compliant price. A manufacturer that follows the price down to its marginal cost wins the lot and starves the plant that was meant to supply it.
Every year Lumaren submits a few hundred bids to public buyers in Southeast Asia. Some go to central agencies running national rounds, some to health-ministry tenders, some to provincial and hospital purchasing committees, and in Indonesia a large share of our public volume runs through the electronic catalogue that supplies Indonesia's national health insurance scheme. The rules differ in almost every detail. They share one habit: once a product meets the technical requirements, price decides.
I understand why. Public health budgets in the region are stretched, generic medicines are meant to be interchangeable once approved, and a ministry that pays more than it must for metformin has less to spend on nurses. I have no quarrel with buyers who press hard on price. My argument is narrower. When a tender treats the lowest price as the only thing worth scoring, it does not buy medicine cheaply. It buys a promise cheaply, and the cost of broken promises lands later, on hospital pharmacists, on patients and eventually on the same budget.
How a price race empties the shelf
A generics plant has high fixed costs and modest variable ones. Once the building, the laboratory and the quality system exist, an extra batch costs little more than its materials, packaging and testing. That arithmetic shapes how manufacturers bid. A company with spare capacity can bid near its variable cost and still feel it has gained something, because the batch contributes a little towards overheads that would otherwise sit idle. Repeat that across enough bidders and the winning price settles near the variable cost of whoever has the emptiest plant.
The trouble is that a plant cannot live on variable cost. Laboratory staff, stability studies, equipment qualification, inspections and the renewal of registrations in several countries are fixed in the short run and unavoidable in the long run. A supplier that wins at variable cost has two choices when the contract runs for two or three years. It can subsidise the contract from other business, which works while the other business lasts. Or it can let the costs that do not show up in any single batch drift downwards: fewer analysts, deferred maintenance, a cheaper source of starting material qualified in a hurry. Neither choice is visible on the day the award is published.
What becomes visible, usually ten to eighteen months later, is a late delivery. The winner misses a shipment because a batch failed or a material arrived late. Hospitals ration their stock, then buy from whoever has it, at short notice and higher prices, often under emergency rules that were never meant to be the normal channel. Some patients are switched between brands in the middle of treatment. The ministry's record shows a saving on the award and nothing about what the shortage cost.
I have watched this cycle from both sides of the counter. In Vietnam's 2021 central round, prices on several high-volume lots fell by more than a third. Two of the companies that won lots we had held fell behind on deliveries within the year, and the hospitals that depended on them bought from us through local procurement at prices close to where they had been before the round. Nobody saved what the award had promised.
Bidding at a price we can keep
After that round our board approved a rule I had argued for: Lumaren does not bid below the full cost of supplying a lot for its whole term. Full cost means the plant's fixed cost allocated by the hours a product uses, the laboratory's testing load, the cost of holding safety stock to the contract's service level, the registration and pharmacovigilance cost of keeping the product licensed, and a modest return. A committee of commercial, supply, quality and finance leaders signs each floor. Only I can approve an exception, in writing, for a defined reason such as a launch or the protection of a product prone to shortage.
The rule costs us volume. We lose lots we could have won, and our commercial teams find it painful to explain to a hospital pharmacist why the company that served them for years has been outbid. Plant utilisation falls in the year after a bad round, and fixed cost is spread over fewer batches. Our Vietnam business lost about S$31 million of annual revenue in 2021 before it recovered.
The rule also changes what we can offer. Because we know what a lot really costs us, we can commit to things a price-only bidder cannot: safety stock held in the country, shorter lead times, a qualified second source of the active ingredient. Where a buyer's rules let us show those commitments, we do. Where they do not, we bid at our floor and accept the result.
I should be honest about the limits of this argument. A floor price does not make a company reliable; plenty of firms charge more and still deliver late. Some competitors bid low because their costs are genuinely lower, and they supply well. And there are tenders where our floor is simply above the market, which tells us something about our own costs that we would rather not hear. The rule is a discipline, not a virtue.
What a ministry could ask instead
Buyers do not need to abandon price competition to buy more reliably. Several changes are within reach of any procurement authority, and versions of some already appear in parts of the region.
The first is to score supply history. A supplier that delivered on time and in full under the previous contract has offered evidence that a new entrant has not. Weighting that evidence, even modestly, changes what bidders compete on.
The second is to split large lots between two or more suppliers, so that one failure does not empty the shelf. Splitting raises prices a little, because each supplier gets less volume. It buys a resilience that a single award cannot.
The third is to make penalties for failing to supply real, predictable and published. Many tender documents contain penalties that are seldom enforced. A bidder that knows it will pay for a failure prices that risk into its bid; a bidder that expects leniency does not.
The fourth is to compare the winning price with an independent estimate of what supply actually costs. When the gap is large, the buyer is not getting a bargain. It is extending credit to a supplier that will repay it in shortages.
None of these ideas needs new legislation in most countries, and none asks a ministry to trust manufacturers more than it should. Each asks the buyer to price a risk it already carries without recording it.
For boards of generics companies the point is just as practical. A tender book that grows by winning at marginal cost is borrowing from the plant. Directors should ask what floor the company bids against, who may break it and how often they have. A management team whose answer is that it bids to win has not yet decided what it is willing to lose, and in this business that decision is the one that keeps medicines on the shelf.