On the surface, evaluating consumer risk may seem simple – we’re assessing the creditworthiness of our customers. An accurate assessment will help deliver a more predictive picture, so where we concentrate our focus on what really matters. But why do we assess consumer credit risk, and how might this change in the future? It turns out, the answer lies in alternative data.
Not all data is created equal – we need quality over quantity.
CCR (Comprehensive Credit Reporting) data is currently used by approximately 50 financial organisations in Australia, allowing these institutions to use both positive and negative information to assess credit risk. With this in mind, it’s more important than even to have the relevant data for the specific type of customer risk your business needs to analyse.
Let’s take a look at the traditional – and alternative – data sets that make up consumer credit risk.
What are traditional data sets?
Credit reports usually include the following customer data:
- Credit Enquiries – previous enquiries for credit products such as credit cards, home loans, personal loans and business loans. They typically include the credit provider and loan amount.
- Defaults on credit cards, loans and utility bills.
- Bankruptcies, court judgements, personal insolvencies.
- CCR Data –
- Accounts – Information about accounts held with CCR data providers.
- Repayment history – a history of up to two years of repayments, whether these repayments were made or not.
Credit data is fantastic to have, but it only includes credit-related behaviours, so it doesn’t give us the complete picture of a customer’s credit risk, or their likelihood of going into hardship. This is because it excludes details that can tell us a lot about someone’s risk of defaulting – income, Buy Now Pay Later (BNPL) payments, rent, insurance and subscriptions like Netflix.
Alternative data sets look beyond demographics to deliver a more complete picture of risk.
In contrast, alternative data sets give us a granular view of expenditure, and high visibility of income. This includes things like:
- Employment insights – see employer and industry details, income amount, any increases/decreases in income and payment frequency.
- Rent recognition – amount of rent, on-time or late payments.
- Welfare information – for example Centrelink payments.
- Significant transaction alerts
While traditional banking information is readily available on most bureaus, access to unique data can be more relevant when predicting credit risk.
Blending traditional and alternative data sets helps see the full picture
When combined with existing traditional data sources, alternative data sets can provide us with a host of new insights. This has become even more apparent since the advent of the Covid-19 pandemic – we’ve seen that customers may look squeaky clean due to current stimulus payments or mortgage holidays, but there may be more under the surface that we’re not seeing. Using alternative data can actually deliver deeper insight into people’s actual financial situation, clarifying those market distortions.
Our customers crave digitisation, automation and personalisation
Time is our customer’s most precious commodity, and in finance, we haven’t always respected this with lengthy application forms and loan approval rates that can stretch into weeks.
We’ve all seen the impact of the BNPL on the traditional finance market – customers expect to be able to complete the onboarding journey wherever they are, ideally on their smartphone, and quickly. Covid-19 has pushed the need to digitise the customer journey to the next level.
A step change in onboarding can only be achieved through automation. For example, when it comes to customer onboarding, rather than asking someone to fill out a form, could you get them to take a photo of it with their phone? A streamlined journey is likely to win more customers.
Our Open Banking future
With open banking very much on the way, alternative data sets will have an even bigger role to play in credit risk. At illion, we’re strong supporters of Open Banking, especially as a driver of increased consumer confidence. By February 2022, we expect all financial institutions to be on board, which increasing both the diversity of data available and the ease with which customers can control, and share it.
The future of credit scoring
Over the next couple of years, we’ll be feeding in employment insights and transaction data to give a more accurate score that’s even more predictive of consumer risk. With geo-Centrelink data as a predictor of future hardship, employment and industry data as a prediction of hardship and of credit risk and bank statement, BNPL, income and ATM use as predictors of credit risk, we’ll have a more accurate and predictive score than ever before.