Photo: Hemant Mishra/Mint
Photo: Hemant Mishra/Mint

Private agencies devise new credit score models for first-time loans

With older methods falling short in case of new borrowers, firms eye Internet footprint, social media behaviour

The inability of traditional credit bureaus to assign credit scores to many loan applicants who haven’t borrowed earlier has spawned a variety of private agencies using novel techniques to fill the gap.

With industrial credit growth slowing, banks are focusing more on retail customers, where first-time borrowers make up nearly half of all new retail applications.

However, such applications are approved only when they are backed by a credit bureau score and a report supporting the customer’s income claim and repayment record.

Yet, traditional credit bureaus are unable to determine an ideal score for many borrowers who are young, and have few transactions with banks.

Credit scorers such as CreditVidya and Lenddo use alternative measures to assess credit risk, typically analysing a borrower’s Internet footprint, social media behaviour and employing other means such as psychometric tests.

CreditVidya says it looks at data sources ranging from behavioural, transactional, location and social profiles to assess the risk of an individual. This could be, for instance, a check on whether a loan applicant checks office e-mail from a place where she says her office is located. The bureau says its engine runs basic searches around the Internet to verify the applicant’s employment, spending habits and cash withdrawal rates from ATMs.

“Consistency of a particular behaviour among these data points demonstrates stronger stability of a customer," said Abhishek Agarwal, co-founder and chief executive officer, CreditVidya.

Hong Kong-headquartered Lenddo, another credit assessment firm, says it uses psychometric tests and application form analysis as tools to check credit risk. The company, which works in 20 countries around the world, is focusing purely on small-value loans between 1 lakh to 8.5 lakh in India.

To be sure, traditional credit bureaus are upping their game, too.

Experian, which started in 2010, has introduced a technique that compares a first-time customer with statistically similar customers who have borrowed before to see the trend in repayment.

However, credit bureaus are not excited by non-traditional tools such as social media data.

“It has to be noted that social media data is very subjective and prone to manipulation," said Harshala Chandorkar, chief operating officer, CIBIL, which is India’s largest credit bureau.

“So, although there’s a buzz over it at present, one really has to wait and watch if such systems work. On the other hand, credit information solutions are tried and tested models that have proved their utility and benefits for the credit sector in our country," she said.

Credit bureaus grew in prominence after many retail loans that weren’t backed by collateral went unpaid during the financial crisis of 2008. Large private sector lenders and foreign banks suffered heavy defaults due to lack of proper customer data and liberal lending policies. Now, most retail lending decisions have shifted to a credit score-based lending system.

The new-age credit scorers believe the coverage that credit bureaus provide is not enough, as banks have little information on first-time borrowers.

“Bureaus have introduced a model that uses variables like name, age, location and marital status to give credit scores to a new borrower. This model is limited in its power, as the bureaus segment customers based on demographic information, which is a biased approach. Had traditional credit bureaus been enough, banks would have catered to new customers in much larger numbers," Agarwal of CreditVidya argues.

While lenders see the logic behind acquiring younger customers to expand their retail base and the need to use alternative credit assessment tools to ensure credit worthiness, they are yet to fully warm up to them.

“We are surely looking at this, but as of now, these are only additional parameters over and above the traditional credit score. However, in lending opportunities like micro-finance, we are closely looking at newer methods of assessment," said Rajiv Anand, executive director-retail, Axis Bank.

According to Anand, since the bank is currently focusing on its own customers to push smaller loans, traditional methods of credit assessment seem to be enough. “This is how things stand right now. But in the future, we will see these things change," he added.

According to Sumit Bali, senior executive vice-president and head-personal assets at Kotak Mahindra Bank, the main issue with using third-party services for credit assessment seems to be the issue of invasion of privacy.

“There is a thin line between using social media or SMS-based information for credit assessment and invading a customer’s privacy. We have been meeting with a few companies that offer these services over the last two months, but are yet to take a firm call on it," said Bali.

Kotak Mahindra Bank has been using surrogates such as educational background, duration of employment and financial stability to assess younger customers who may not have a credit history.

“This customer base is large enough for us to not ignore it and the surrogates have helped us in assessing them better. On an average we have found these customers to be better performing than those who have a low credit score," Bali added.

On the other hand, peer-to-peer (P2P) lenders, who have become active over the last two years, have embraced these alternative methods of credit appraisal.

For instance, LenDenClub, a P2P lender, has been looking at a customer’s income data, equated monthly instalments (EMIs) repayment record and credit bureau data and pairing it with behavioural patterns on consumer loans, family details and employer track record to take lending calls.

“Most credit bureaus do not take into account the repayment record on smaller loans such as those taken to buy electronic appliances or for short-term requirements, as they are small and repaid in a couple of months. For our customer base, this data point becomes important and so we keep a close tab on these transactions," says Bhavin Patel, co-founder and CEO of LenDenClub.

According to Patel, his company has seen a default rate of only 0.3% in about 18 months of active lending of loans with a duration of six to 12 months.

Close