Risk Joins the Top Table

As a credit risk professional I must admit to enjoying the challenges the credit crisis and subsequent recession have created. Prior to 2008, the focus was on lending strategies, bad debt forecasting and Basel modelling. Ever since then, the pendulum has swung, because today the focus has shifted towards collections and recoveries, leaving the days of arguing the toss with marketing folk about accepting business below scorecard cut-offs a long way behind. Well, almost.

Behind the scenes, many of the leading debt collection agencies, including ourselves, have spent the past few years battening down the hatches, investing significantly in credit risk with clear purpose of mitigating our losses in the immediate term and increasing collections. This has meant looking at the key areas of data enhancement scoring and segmentation, re-examining, revising and refreshing the risk models that will inform future decisions. But it has not been an easy task.

As every one knows, credit risk analysts take information from various feeds – including credit reference agencies – and then look to refine and enhance the data that they have. Such information is then, in turn, fed into a predictive model, a piece of logic, an algorithm or a scorecard that takes a piece of information today and predicts a defined outcome in the future.

Customer detail
The role of risk in collections is not just focused on whether a particular customer will repay their debt. It is about getting to the detail of that particular customer, through insight, analysis and modelling, to proactively determine what collections strategy will achieve the best result for both that customer and the debt collection agency. It is also about uncovering whether they could be paying more, or less, whether their account should be reviewed, a new payment plan reached or the debt settled outright. Certainly, it helps in both the need and desire to treat customers fairly but it has also improved performance.

Operational teams have been crying out for analytics to support their decisions for many years and especially since the 2008 crisis. Indeed, there is a certain irony here. Modelling requires data, a history and a period of stability in order to predict a future outcome – and this all takes time. In the turmoil of the past two years in particular, agencies have had to rely on the skills and intuition of the operational management teams and trust their experience.