Looking backwards to move forwards

When you build your risk models, have you included Alistair Darling as a variable? If you are a lender, then you currently face a real difficulty – how to comply with the Government’s requirement for risk-based repricing of products, introduced in January, while at the same time meeting its demand for continued lending made this week.

One of the reasons why lenders are having such problems with bad debt is that they regularly ignored their own risk models. In the rush to get a slice of market growth, banks and building societies set aside the rules which were being developed for them by their internal risk departments.

Ironically, they were encouraged to do this by the previous Chancellor, Gordon Brown. Although he argued for a risk-based regulatory environment, he did not mean clamping down on borrowing by pricing people out of the market. He supported the extension of mortgages (and credit generally) into previously untapped groups in the population and product types, like buy-to-let, which have now become so toxic.

Reining in those risks through sharp repricing of products is most visible on credit cards, but is happening across all forms of lending. For the first time, financial services providers are actually applying the models, built at some cost by highly-skilled analysts, to their customer base.

As a result, the acceptable customer base is suddenly emerging as much smaller than anticipated. This is why the credit crunch is continuing because a limited proportion of businesses and consumers meet acceptable criteria. Risk-based pricing could remove as much as 30 per cent of the previous market if fully applied.

So Darling has stepped in to try to skew those models. At the same time as requiring lenders to maintain a larger, stable capital base by not having such high levels of exposure to bad debt, he wants them to keep more of that capital in the marketplace.

So can this seemingly impossible balancing act be achieved? What would be the impact of applying a Darling weighting to risk models?

One answer might be to go back to risk, credit and population data from the start of the decade, before the recent housing boom started. The profile of borrowers considered an acceptable risk back then provides a good indicator of the true market size. After all, if the same businesses and consumers are still around and looking for borrowing, that is a good sign that they can manage their own risk.

Data analysts are often accused of using a rear-view mirror when what the business wants is forward looking insights. But in this case, it may be the only way to meet contradictory demands by the Government.