Behaviour predictions could assess loan risk

by Rebecca Pike26 Nov 2018

The possibilities of data sharing in Australia could include predicting consumer behaviour to assess loan risk, according to a firm in the United States.

A team from US credit scoring company FICO was recently in Brisbane at the Australian Retail Credit Association (ARCA) national conference, which focused on the transition to open data.

While new systems such as comprehensive credit reporting (CCR) and open banking are new to Australians, the United States has been doing it for years.

In Australia, CCR came into effect in July this year and is expected to provide extra credit information, such as credit limits and repayment history, to give banks a better picture on a potential borrower’s credit score.

Open banking comes into play in July 2019, when consumers will have the ability to share a wider range of financial data between financial institutions.

According to FICO, these two systems are working together in the United States and allowing people to build a credit score that can better assess risk.

FICO, which was established in the 1950s, created the well-known FICO Score to measure consumer credit risk.

It ensured there was one standardised credit score across the bureaus and lenders, unlike what we are currently seeing in Australia where each credit bureau has a different scoring methodology.

Sally Taylor, Scores vice president, at FICO, said they have recently launched something called an UltraFICO Score, which combines credit scoring with open banking.

Taylor and her colleague David Shellenberger, senior director, Scoring and Advanced Analytics, recently attended the ARCA conference.

Taylor said, “The standard credit bureau negative/positive data is the basis of the Ultra FICO score and then the consumer can contribute their current account, savings account type of information through open banking type of connections.

“So the typical case would be somebody who might otherwise be turned down for credit. Either their credit report isn’t scoreable, there’s not enough information or enough history, or they have a lower score and would have otherwise been declined.

“So, what does your inflow and outflow of funds look like? Do you ever have negative balances? Do you tend to save? That’s the information that’s reviewed there and then it adds to the FICO score.

“For the target customers, those who have not had a lot of credit or very negative credit, we see that about 80% then will actually get a higher score.”

Sharing this information could also mean predicting behaviour based on specific factors inside borrowers’ accounts.

Taylor said, “What we’re predicting with the scores is we’re predicting whether the consumer in the future like one year out two years out will default on any type of credit.

“It doesn’t just look at delinquency to predict future delinquencies. But we also look at how long you’ve had credit or picking up a lot of new credit recently. You see that those who have had credit a long time tend to be lower risk than those who haven’t.

“Those opening up a lot of new credit recently tend to be higher risk than those who aren’t and then in the US because we have balance information, we see people who have a very high balance relative to their limit are risky.”

Taylor said there was a specific job for brokers as more data sharing comes into play. Particularly without a credit score industry standard, brokers will have to educate customers on what their score means and how it might affect their ability to get credit.

She added, “I think brokers can explain how the credit process works and how the information is going to be used. But also some of the common sense tips around how to improve their scores.

“Don’t ever take more credit than they feel comfortable affording, make sure they pay all their bills on time and then make sure if they’re getting ready for a mortgage they don’t take any more credit than they need.

“It used to be only those priming themselves for a loan that would know their credit scores. But now accounts show their FICO score data to the consumer on a regular basis.

“So consumers are more and more aware of what their score is and they realise that they keep that score healthy not just right when they’re priming themselves for a loan but over time so that they can afford and get the best rates for a mortgage or a car loan or student loan.”