Only some suppliers got the letter.

Two builders' merchants supply the same housebuilder. Same terms, similar balances, the same exposure. In the second week of August, one of them received a letter about that customer and the other did not.

When a trade credit insurer reduces cover on a buyer, it notifies the suppliers who hold a policy on that name. It does not notify the market, the buyer's other suppliers, or anyone reading the buyer's credit file. Cover can be reduced for new trading while existing arrangements are untouched, and the supplier's practical choice becomes whether to keep shipping on terms, hold the limit, or tighten. The exposure is shared across everyone selling to that buyer. The warning is not.

What was reported, and what it was

On 10 August, trade press reported that Allianz Trade had told suppliers it would reduce credit limits on new trading with Vistry, one of the UK's largest housebuilders, by as much as 70% on some accounts. The reduction was reported as applying to new agreements rather than retrospectively, with the eventual level tied to the housebuilder's performance over the following weeks. Vistry said its cover more than meets its requirements and reported no interruption to supply.

Four days later, the Builders Merchants Federation described the effect on its members: falling volumes and reduced insurance arriving together, a “double impact”, against early figures showing June volume sales down around 10% year on year.

That statement is the only reason anyone outside the affected policies can discuss this at all. Which is the subject of this article — not the housebuilder, and not the insurer.

A cover decision is a judgement, not a verdict

Allianz Trade publishes how the mechanism works, and it is worth reading before drawing conclusions from it. Cover requires a valid credit limit against each customer. Where a supplier holds an approved limit, the insurer monitors that buyer continuously. On negative information, it may ask the supplier to complete a buyer review form setting out amounts owed and orders in hand. Many reviews result in no change. Some result in reduced or withdrawn cover, and where a limit is withdrawn the insurer states that it considers the risk to have increased significantly, with cover ceasing after a delayed effect period — usually 30 days — during which goods already supplied remain covered.

Two things follow, and both matter. The decision is a judgement about the insurer's own exposure, formed from the information available to it and its appetite for that risk. It is not an audit of the customer, and it is not a finding about solvency. At the same time it is genuinely informative: an insurer carrying a live position on a buyer and reviewing it continuously is watching that buyer more closely than many creditors watch their own ledgers.

The grace period tells you something too. It exists so a policyholder can go back to the customer for information that might change the decision — which is an acknowledgement, built into the process, that the insurer's picture is incomplete. Treating a cover reduction as noise is a mistake. Treating it as proof is a different mistake.

The warnings that matter most stay private

Trade credit insurance is a bilateral contract. The assessment belongs to the policyholder who paid for it. There is no register of reduced limits, no public feed of withdrawn cover, and no obligation on anyone to tell the rest of a supply chain what an insurer has decided. That is not a defect in insurance. It is what a private contract is.

It does, however, produce a specific market condition, and the cover decision is not the only example. A bank covenant waiver is in a facility agreement nobody files. An HMRC time-to-pay arrangement is not published. Changes to an invoice finance facility — reserves, concentration caps, termination — are private between lender and borrower. A statutory demand is served, not registered. A notice of intention to appoint administrators sits on a court file, is not filed at Companies House and is not gazetted.

So the information most likely to change a credit decision is, as a class, the information least likely to be public. By contrast, a credit file records things that have already become visible.

The three channels a warning travels

• The policy channel. Your insurer tells you if you hold cover on that name. Precise, timely, and available only to policyholders.
• The public channel. A trade body, a regulator or the press makes something visible. Broad reach, but it depends on somebody choosing to speak, and it arrives after the decision.
• The observation channel. What you watch yourself. Slower than the first, earlier than the second, and the only one that is always available to you.

Most credit functions are equipped for the first channel or resigned to the second. The third is where the gap sits, and it is the only one a supplier controls.

What the supplier without the letter can still watch

The uninsured merchant's problem on 15 August was not philosophical. It was an order in hand and a delivery booked. Stop trading with a customer that is still paying and you lose the account and the margin, often to a competitor who kept supplying. Keep shipping at a limit set when conditions were different and you finance the loss yourself. Doing nothing is not neutral; it is choosing the second option without saying so.

Without a policy on the name, these remain observable:

• Registered charges at Companies House, including new security and the gap between the date a charge was created and the date it was delivered.
• Filing behaviour — accounts or confirmation statements running late. This is a behavioural signal rather than a financial one, and it is often earlier than the financial one.
• Officer changes, particularly a finance director leaving without a replacement.
• County court judgments through Registry Trust, and winding-up petitions advertised in The Gazette. Both sit late in the sequence, but both are public.
• Payment behaviour: the payment practices register for larger businesses, and your own ledger for everyone else.
• Concentration — who pays your customer, and whether that payer has changed.

And it is worth being equally clear about what cannot be seen at any price: another supplier's cover position, a covenant waiver, a time-to-pay arrangement, a stop applied by a competitor. No UK register of credit holds exists. That absence is the structural reason the observation channel matters.

The trade outcome is the ordinary one. A merchant on 45-day terms shipping into an account whose limit was set in a better year does not discover the problem at the point of decision. It discovers it in the receivable, one or two quarters later, at a balance nobody re-underwrote.

Shared exposure needs independent visibility

Insurance did its job here. An insurer watched a buyer, formed a view, and told the people it had a contract with. The difficulty is not that the insurer acted. It is that a supply chain's picture of a shared risk is assembled from private notices that reach different creditors at different times, and reach some creditors never.

A creditor whose first sign of trouble is a change to someone else's policy has outsourced its early warning to a contract it is not party to. That works until the day the letter goes to a different address.

Every creditor extending terms to a customer needs its own continuing view of that customer — not because the insurer's view is wrong, but because a credit decision cannot rest on a warning that may never arrive. The risk is shared. The warning isn't.

Grand keeps a live view of the companies you trade with, so a change reaches you whoever else it reaches. Free checks every month. No subscription required.