What is risk weighting?

A bank does not hold capital against the raw dollar value of its loan book. It first multiplies each exposure by a weight that reflects the chance of loss, then adds the results together to get risk-weighted assets. Capital is then measured as a ratio against that total.

The weight depends on what kind of exposure it is, who the borrower is and what secures it. Cash and government debt sit at the low end. Residential mortgages sit above that, unsecured business lending higher again, and property development, defaulted exposures and equity holdings at the top. Some banks use supervisor-approved internal models instead of the standard table.

Capital is expensive, so a loan that soaks up more of it has to earn more to be worth writing. That feeds through into how keenly a bank chases certain lending, how it structures deals and where it sets its appetite, rather than into any single quoted number.

It also explains why lenders ask for the things they ask for. A lower loan-to-value ratio, registered security over the collateral and a clean repayment history all reduce expected loss, and lower expected loss generally means a lighter capital charge.

Because the weights differ by asset class, two lenders can look at the same deal and reach different answers. A bank funded by deposits and bound by prudential capital rules weighs it one way. A non-bank lender, funded through securitisation or wholesale lines, works to its own funder's criteria instead.

That is one reason a broker will place similar customers with different lenders. It is also why exposures that turn into a non-performing loan hurt twice: the loan itself is at risk, and its weight rises, pulling more capital in behind it.