Why credit data must feature in your plans for 2019

The use of credit data in insurance is well established. And to some degree, it’s been an understandable development. It can send signals about the risk being insured. It tells a story.

Yet stories can be told in more than one way. That credit data may tell an insurer one thing, but it can tell a debt charity something quite different. This points to there being validity in both stories, but that the context within which you ‘read’ them matters hugely. Indeed, data can only be ‘understood’ if that context is as clear to the reader as the numbers themselves.

So what do senior executives need to understand about the use of credit data in insurance? They need to understand how it influences underwriting and claims decisions of course, but they need also to understand it within other, more customer-centric contexts. Credit data can send signals about vulnerability.

Credit Data and Vulnerability

That credit data could well have a story to tell about vulnerability is not something new. It’s pretty much self evident. So why pay more attention to it now? It’s because there have been changes around important ethical questions relating to underwriting, claims and governance. In underwriting, how credit data is collected and used is highly pertinent to the current pricing review. In claims, the trialling of its use in settlements sets all sorts of warning lights flashing. I’ve written on both those topics here and here, so it’s on governance that I’d like to focus in this post.

Because credit data can send such strong signals about vulnerability, senior insurance people need to think carefully and comprehensively about the questions they need to ask of their underwriting, claims and data analytics people. And they need to be sufficiently informed so as to understand the answers to those questions, for they may need to challenge them for completeness or accuracy.

I’m not going to list the questions that could be asked about the use of credit data (it’s a long list), but suffice to say, they should cover the suppliers from whom it is obtained, the scope of what it covers, and the analysis being carried out with it.

Consumer Outcomes too

Remember though that these are input questions. Equally important are questions about the impact it is having on consumer outcomes. So how is it influencing the availability of cover, or the composition or price of the cover on offer?

Remember as well that these questions matter to all of us. As what is considered ‘credit data’ broadens, and as analytics drives ever more personalised decisions, most of us will find the cover on offer and the premium quoted being nudged one way or other depending on our near real time spending patterns.

There’s also nothing inevitable or obvious about the conclusions that insurance people can reach. The US states of California and Massachusetts have banned credit based scoring for motor and household covers, while Maryland bans it for household and Hawaii for motor. They deem it to be unfair and discriminatory. There is more than one story around the use of credit data in insurance.

Test, Now

So where to start? Try running some specially designed test datasets through your algorithms to establish how credit data is influencing decisions and the extent to which those decisions are falling within the bounds of fairness and non-discrimination. A third party free of conflicts of interest helping you with this would add to its credibility.

And the time to do this? Now is a good time, certainly for UK insurers. 2019 will see the Financial Conduct Authority (FCA) launch a market study into the “access to, and use of, credit information including the coverage, timeliness and consistency of data and products provided by credit reference agencies”. It would have happened sooner if the FCA had not given priority to the pricing review.

Given that the use of credit data in insurance is likely to feature in the FCA’s pricing review, it makes sense for insurers to kill two birds with one stone and really explore the influence credit data is having on outcomes for customers. And one final point: don’t just reach a conclusion; “show me your workings” could become the regulator’s favourite refrain.