Watch List: How a Bank Flags a Loan Before It Fails

A bank reads most of its loans once a year. It sends a small number to a specialist workout team. Between the two is a band of files where something has moved and nobody yet knows what it means. Left to the annual cycle, a slipping borrower can lose a year. Sent to workout too soon, a recoverable borrower gets a treatment it does not need. A watch list gives that band an owner and a shorter clock.

The list itself is a management tool, and its label and rules are the bank's own. Regulators ask for the outcome. In Australia, APRA's credit risk standard requires deposit-takers to have policies for the early identification and management of problem exposures and a system for monitoring individual exposures, without prescribing a list. US supervisors have a formal grade for this band, called special mention. A New Zealand or UK bank is more likely to use its own label.

Triggers differ by bank. Some are automatic, set off by a grade or a policy test. Others are one person's judgment, and a good process lets a banker or credit officer add a name without waiting for a breach. The signals that lead a file there are often the early warning indicators, and several moving the same way carry more weight than one. A credit function or committee then reviews the list at set intervals, often monthly or quarterly.

The exit deserves as much attention as the entry. Without written tests, a name tends to stay until it either recovers beyond doubt or fails, and a list that only grows stops telling anyone anything.

A specialist cheese maker, constructed for this page like every figure in it, has a term loan of US$4.2m and an overdraft limit of US$1.2m, so its exposure is US$5.4m. Debt service is US$1.0m a year and a covenant requires cash cover of 1.20 times. It sits at grade 6 on the bank's ten-point scale. The scale, the covenant and the probabilities below are the example's and are not offered as a market norm.

No covenant has failed and no payment has been missed. But three lines are moving together, and the cover test has about US$40,000 of cash to spare. A credit officer adds the name at Q3 and moves the grade to 7. The bank's estimate of expected loss changes with it, here using a 35% loss given default:

The file now carries monthly accounts within twenty days, a monthly call, and a plan built around its largest customer's renewal. Any change to margin comes only as the facility allows. The exit tests are written down: cover of 1.40 times for two consecutive quarters, the overdraft back to nil at its seasonal low, and accounts within twenty days. Cover then comes in at 1.31, 1.43 and 1.46 times. The overdraft clears, and the name comes off after the third of those quarters. Had the first of them come in at 1.08 times, a breached covenant would have started the reservation-of-rights letter and the file's move toward workout. Either route is the list doing its job.

DEFINED IN ONE PARAGRAPH → · HOW BANKS MANAGE CREDIT RISK → · ISSUE 09 · THE TUESDAY LIST →

A risk grade is a measurement; the watch list is what the bank does about it. The two usually move together, but a bank can list a file without changing its grade, while it finds out more. A non-performing loan is a later category again, generally defined by ninety days past due or an unlikeliness to pay, and most watch-listed loans are still performing.

The list also sits close to, but is not the same as, the accounting stages behind a loan loss provision. Staging is a test written into the accounting standard. The list is a management decision. A listed loan will often have had its credit risk reassessed and may be in a higher stage, and the Basel Committee's guidance on expected credit losses names watch-list loans among those where individual assessments are commonly made.

Listing rarely affects the facility in place, but it changes everything around it. New limits are harder to get approved. A renewal coming up is read with the list in mind. Requests that would have been routine, such as a dividend, a covenant reset or a new guarantee, now go to a more senior reader. In Issue 09, the chair of a portfolio review with eight names and two intensive-care places has to decide which two names take them, and the answer turns on behaviour more than on ratios.

From the borrower's side the list mostly arrives as extra work: more frequent requests, sharper interest in the cash forecast, a banker who has started copying in someone from credit. The borrower cannot control the entry, but it can control what the bank learns next. A business that brings its own figures and its own explanation early gives the bank something to weigh against the trend.