Credit scoring has been common practice for many years now among banks and other financial institutions. Other industries are just starting to realise its benefits.
Today, in our Big Data society, there is such an astronomical amount of data available to us that it can be mind-boggling to make sense of it all. The statistical techniques lying behind a typical scorecard are an ideal way to evaluate all the data available to you, understand what each data asset tells you and how it correlates with others, determine which specific data assets are the most valuable, and how exactly they should be used.
A score is simply another way of expressing a probability – a likelihood. If you had ten balls in a bag, nine red ones and one blue one, then you are 90% likely to pull out a red one if you stick your hand in blindly. Clearly though, there’s always a chance you could pick the blue one. Indeed, you should pick the blue one 10% of the time.
Credit scoring can be likened to this. If you look at the way people have repaid their debts in the past, those with lots of unsecured debt elsewhere would have struggled more than those with no other debt. In some cases, people with a recent County Court Judgment may be less likely to manage other payments than those with a clean record. But then of course there will be people with a lot of debt who manage their payments very well, even some who obtained a CCJ through unfortunate circumstances and have since turned things around. This is why a holistic view of all available data is required, and the more factors you consider together, the more likely you are to make the most responsible decisions for each customer, helping them to stay within the boundaries of their affordability.
Every day here in Decision Analytics at Experian, we are developing statistical models that take a vast number of data items and produce an overall score for each customer, giving the best possible assessment of their financial situation.
We let advanced statistical methods do the talking, and make sure we utilise all the data we have available to us. Risk managers can then overlay policy rules to define their overall strategies, and ensure they make the most appropriate and responsible lending decisions.
The economic environment we live in is constantly changing, and credit scoring strategies need to change with it. As economic conditions alter and more data becomes available, scorecards should be refreshed every 2-3 years, as the more data that can be utilised, the stronger your resulting strategy will be.
If you would like to talk to an expert about scorecards, and how they could help your business, please get in touch.
You may be interested in: