Setting credit limits and risk appetite: a practical method

A credit score tells you how likely a counterparty is to fail. A credit limit tells you how much that failure would cost you. Teams that conflate the two end up granting large limits to strong-looking companies with thin balance sheets, and refusing sound business because a score sits one band too low.

Start from what you can afford to lose

Risk appetite is an amount, not an adjective. Express it concretely: the maximum bad debt the business will absorb in a year, the maximum exposure to any single counterparty, and the maximum to any one sector or country. Those three numbers constrain every individual limit that follows and turn credit decisions into portfolio decisions.

Size the limit against the counterparty's capacity

A limit should be affordable for the customer as well as for you. Common anchors are a percentage of net worth, a multiple of monthly trading volume, or a share of available working capital — whichever is lowest. A limit larger than the customer's ability to settle in one cycle is a limit that will eventually be tested.

Adjust for mitigation, not for optimism

Prepayment, a parent guarantee, retention of title, a letter of credit or credit insurance each justify a higher limit because they reduce loss given default. A long relationship, a strong brand or a persuasive sales case do not. Keep the distinction explicit in the approval record.

Aggregate to the group

Limits granted to subsidiaries belong to the same parent's exposure. Aggregating to the ultimate owner is the single most common gap between a policy that looks disciplined on paper and a portfolio that is quietly concentrated.

Review on events, not on the calendar

An annual limit review is a formality; an event-driven review is a control. Score movement, a new charge, late filings, an ownership change or a sustained increase in days-to-pay should each trigger a re-look, with a documented outcome even when the outcome is 'no change'.

Measure whether the method works

Track two things: losses against appetite, and overrides against outcomes. If exceptions consistently perform better than the model, the policy is too tight and you are refusing profitable business. If they perform worse, the exception process is where your bad debt is being created.

— RisQo Research Team, Infocredit Group