The shift in the assessment from traditional credit scores to assessment of “digital footprints” has enabled access to institutional credit to the hundreds of millions of underserved.
As credit gets integrated with the lifestyle of consumers, traditional ways of credit scoring are gradually turning insignificant due to the large amounts of customer-specific data that is unaccounted for. It’s alarming to know that 50% of first-time loan applicants are still denied a loan and are forced to resort to informal channels of credit, such as moneylenders, at exorbitant rates. Credit bureaus such as CIBIL are still dependent on an individual’s financial history even though highly insightful data can be found through other sources, i.e. through their digital presence.
The shift in the assessment from traditional credit scores to assessment of “digital footprints” has enabled access to institutional credit to the hundreds of millions of underserved, thus transforming the retail credit landscape. Alternative data collected from an individual’s digital presence is now facilitating the issuance of small ticket-size loans. The introduction of new ways of assessing data by leveraging big data and AI technology is increasingly being seen as a means to an end — the end being a country where the majority of the population has access to institutional credit.
Shift From Conventional to Unconventional
While traditional scores focused largely on long-term behavior indicated by factors such as payment history, total amount owed, length of credit history, the unconventional data assessed these days includes things like payment history for telecom bills, online shopping history and even social media profiles. Modern credit scoring exposes banks to a dataset that is largely inaccessible but can prove highly beneficial in determining the creditworthiness of an individual. For example, if there is a 21-day period to settle a bill payment and as a practice, a person settles it in the first two days of the amount becoming due, it can be said that this person will not be likely to default, if given a loan.
Interestingly, social media has emerged as a crucial factor in alternative lending. Social networking websites provide great insights into how an individual spends their time and their network of friends. An individual’s personal preferences may possess hints about individual traits and personal backgrounds that relate to creditworthiness. Also, the analysis of a person’s social ties may help improve credit scoring as it carries important unobservable information about the borrowers. Getting the “behind the scenes” information such as where the borrower shops or went for his/her last vacation can be beneficial in drawing up conclusions.
While the statistical definition of a credit score is based on the probability of 12 months of default, in totality credit score measures discipline. For example: If two people earn an income of Rs 30,000 per month and one of them ends up spending all of it in a week and the other person saves Rs 2,000 at the end of a month – person A will be more likely to default if given a bank loan. Other factors such as when cheques get bounced or ATM cards are declined — a part of a person’s financial discipline – are also taken into account. As a part of digital footprints, an individual’s SMS history helps ascertain income, timeline for payment of electricity bills, uber payments etc.
Role of Artificial Intelligence
There are enough proof points to substantiate that the future for credit scoring in India is going to witness greater evolution in the three ABC’s of Fin-tech — Artificial Intelligence, Big Data and Cloud Computing. In recent years, there have been huge efforts towards developing comprehensive scoring mechanisms. Banks and NBFCs are far more open to innovation and the adoption of new technologies, which is being complemented with an evolving policy framework around lending.
Technological advancements, through AI and smartphone penetration, have built capacity in the alternative credit scoring and data analytics space. Big data, for example, has helped in assessment; it helps identify certain geographical locations wherein the percentage of defaulters spike. The density and frequency of interpersonal relationships can also be identified across geographies, which facilitates the estimation of an individual’s income bracket.
AI algorithms churn currently available data to predict future outcomes with 90 percent accuracy. This has been a big improvement from the earlier model, wherein banks assessed an individual based on limited data points due to dispersion of data. Modern credit risk assessment models rely on ML techniques that can help find hidden traits in a dataset, which leads to more accurate predictions. By using AI, we have drastically reduced the time taken to approve a loan request as compared to traditional lenders.
Evolution of Credit Scoring in India
India is currently home to over 350 million first-time borrowers and the evolution in fin-tech has been successful in including a large segment of the population into the credit ecosystem. We still have a long way to go before social media alone can compute credit scores, but we believe that alternate credit scoring can definitely pave way to affordable credit accessible to more than 3 billion underserved people across the world. With an increasing focus on data security, fin-tech firms are now taking measures to improve data security by hiring system hackers in their own organizations. In the future, availability of infinite data will lead to newer ways of harnessing it. Integration of traditional and digital data to form powerhouses of information – available through one source can also be seen as an upcoming trend in this space. Along with this, an individual’s digital presence will also strengthen with technological innovation and can ensure 33% lower delinquencies for the same level of risk. Banks and NBFCs are increasingly looking for alternate sources of data to reach out to India’s rising youth population with an intention to enhance the potential of perhaps the most productive set of consumers. Alternative lending will, therefore, continue to grow as a new wave with hope that credit will reach the last mile consumer in the most convenient and affordable manner.
(By Abhishek Agarwal, Co-founder and CEO, CreditVidya)