Two companies, same sector, opposite risk — which one gets the tighter limit?
Two customers, same sector, same limit request. One is strengthening, one is weakening — and the sector view gives you one answer for both.
Two customers sit in the same sector and ask for the same limit. One is strengthening. One is weakening. The sector view gives you the same answer for both — same code, same rating band, same assumptions priced into the line. Your ledger will not.
Sector risk is the default lens because it is cheap and everywhere. Every bureau report, every trade-credit outlook, every insurer bulletin arrives pre-sorted by sector, so “construction” or “wholesale” quietly becomes the unit you make decisions on. As a weather report, that is fine. As the basis for a credit limit, it fails — because a sector number is an average, and an average sits precisely between companies moving in opposite directions.
This spring made the point in plain sight. Inside the same construction-supply channel, one merchant business kept growing while another’s like-for-like sales fell 5%, with builders’ merchant volumes down across the board even as individual names pulled apart. A credit team pricing both off “construction is elevated risk” gets both calls wrong at once: too cautious on the one that is compounding, too loose on the one drifting toward a default. The sector told you it was raining. It never told you which building was on fire.
The same sector, two different credit decisions
What separates those two customers is not their sector; it is their behaviour. One is paying to terms and holding its debtor days steady. The other is paying you slower than it did 90 days ago, stretching its own creditors, and edging toward a county court judgment that will land well before any insolvency notice. A sector rating sees none of this because it was never measuring the account. It was measuring the category the account happens to file under.
Why sector-level risk fails when precision matters
The problem is not that sector data is wrong. It is that it is blunt at the exact moment you need precision — setting or revising a limit on a live account. A sector rating tells you what is happening to a category of companies. It does not tell you which customer is strengthening, which is weakening, or whether your exposure to that account is still appropriate. Two companies in the same sector can carry opposite credit risk because risk lives in the individual account’s payment behaviour and trajectory, not in the sector it belongs to.
And the error is asymmetric. Price is too tight, and you lose margin, or lose the customer to a competitor who looked closer. Price is too loose, and you carry exposure straight into a deterioration you never saw coming. Both mistakes come from the same habit: making the decision at the level of the category instead of the account.
What account-level monitoring shows that sector ratings miss
This is where onboarding ends and monitoring begins. Onboarding tells you who you are dealing with; it is necessary, and it is a snapshot. Monitoring tells you what that customer is doing with the exposure you have already extended — the slowing payments, the stretching debtor days, the director change, the judgment that arrives before the headline does. When a sector turns, account-level monitoring is what tells you which of your customers is the exception and which is the rule.
The practical shift is to stop treating the sector rating as a decision and start treating it as a prompt. When “construction” deteriorates, that is the signal to look — not the answer. The answer is in the account: is this specific customer paying you slower than last quarter, and is their limit still appropriate for what they are doing now?
You can’t reprice a sector. You can only watch an account.
Grand exists for that second job — the continuous, account-level view of your live customers that a sector rating and an annual review simply cannot provide. The sector will always tell you the weather. Grand tells you which building is on fire.
See how it works at heygrand.com.
