Managing for Change in Credit Risk

Scott Dawson

The future will be like the past

The principle of good underwriting is to not just be a responsible lender, but to manage to the company’s appetite for risk. This is often easier said than done since many companies look at historical performance and decide whether the profitability is acceptable at the level of bad debts experienced. Of course historical performance doesn’t mean optimisation and nor does it mean that it will always perform in the same way.

There are three big issues to managing based on past experience: The competition, the economy and the natural propensity for models to degrade. Fundamentally the future will not be like the past and yet this is the main tenet of scorecards. A sample is taken from the past and used to predict performance. But scorecards degrade – fairly quickly sometimes and it is rare that a model will reasonably survive beyond three years. Ignoring completion and the economy for now, why does this degradation occur? The driver is socio-demographic shifts and human behaviour. The products and attitude to credit is changing. The Millennials prefer different credit tools to the ones of prior generations. They are relaxed about privacy and intense users of social media. But the next generation are likely to be different. They’ll probably be more sensitive to personal data and will want their communications delivered in new ways. Historically, these trends were slow, like the take up of home ownership and mobile phones. Then employment shifted towards contracts and self-employment.

Credit bureau data usually contributes between 80 and 100 (if you let it) per cent of the predictive power of a scorecard. That’s OK then, right? This concern over socio-demographic shifts and behaviour doesn’t matter. Well it does. The credit bureaus data isn’t the same. On the one hand they are improving their services by adding data and models, but they still reflect the behaviour of the consumers and the way data is used and reported. Let’s take a simple example: Many years ago, the quotation searches were very predictive. More searches indicted that an applicant may be shopping around for credit. Now quotation searches are no longer available for credit risk assessment, but there has been the secondary development that lenders offer quotations to check whether a prospect would be eligible for their product. This means that previous declines have become hidden searches.

The competition impact the performance of your portfolio in a number of ways. Firstly, let’s look at underwriting. If your biggest competitor with the same product offers a significantly better rate, what will happen? The applicants who previously came to your company will switch to the competitor – at least the ones who are most likely to secure the best rates will. This leaves you with an adverse applicant population. They’ll have lower scores, the acceptance rate will be lower and the subsequent performance will be worse. They’ll have higher defaults than better profile application months – assuming your models haven’t degraded because of the socio-economic shift issue I’ve already discussed.         

The second issue is that your existing portfolio of customers is going to notice that your rates are out of line with the competition. As a result you should expect an increased attrition rate and customers close their accounts and switch to the better interest rate. This can also happen based on product rather than rate. A slick new one with desirable features (mentioning the instant transaction information via apps as an example) may leave you looking at a portfolio with fewer transactors – and worse still: accounts with low balances because the customer hasn’t bothered to tell you they’re not using your product any more.

And then we have the economy, the likely focus of concern if there’s a No Deal Brexit. The economy impacts your customers’ income and ultimately their jobs. There’s plenty of evidence from past recessions to show that portfolio credit losses can go up by between two and three times that of a benign economic environment.

So, for all these reasons, we know the future is not like the past. It never has been and it never will be.

So what can we do?

Companies talk about “Managing through the Cycle” and it’s a requirement under capital adequacy regulations, but it’s often treated as being an estimate of credit losses throughout a whole economic cycle. It’s a forecast to work to, not a solution.

People, Systems and Strategy are the three components for successful management of risk but for the rest of this article, I’ll consider the one most often underrated: your people.

Your staff, both underwriters and collectors, cannot work in isolation of the company’s appetite for risk. They need clear guidance on how to operate. What is acceptable and what is not? In collections, your agents are your best tool for recovering money and so it’s important that they know which accounts to work and when. This means segmenting your arrears portfolio based on stage and issue. Without this, you cannot expect to be efficient or effective. This approach also means, knowing what your end game is and when to use a DCA or other partner.

When new business volumes increase, it is not good enough to look for short cuts. The time to look for efficiencies in underwriting is not when the dam has already burst. Don’t cut out verifications just because the backlog is too great. I hope that many of you are shaking your heads at the very thought of doing this, but some will have seen it happen. Some of you will have seen or heard of companies suddenly automatically declining or approving more or changing the way a bank statement is reviewed for example.

Staff in Collections are just as important and it’s not just about their training and deployment. Having the right number of agents is vital if actions are to be executed in accordance with the strategy. It’s no good expecting that customers in a certain segment will receive three attempted phone calls in a month if that objective isn’t achieved.

So managing through the cycle should also mean, having robust capacity plans in place. And if staff can’t be recruited (or flexed) as quickly as you’d like, then maybe it’s time to look at outsourcing.