Sandeep Anandampillai, Co-Founder & CPO, Crediwatch explains how alternative data and insights could revolutionize the entire credit ecosystem
There are approximately 63 million MSMEs in India, a majority of whom have sparse data and an almost negligible digital footprint. Does that mean that a lack of substantial digital data will mar their chances to credit? Well, it would become the most logical outcome if traditional data points are made the only go-to parameters. Today, the advancement in the technology and finance ecosystem has opened up new avenues for the financial inclusion of small businesses.
Financial institutions are leveraging alternative data to serve digitally thin-file borrowers, who may otherwise not be approved for loans. This information is further working as a robust mechanism in screening out fraudulent practices and better servicing the customers as well as loan repayments. Banks and other traditional lending institutions have taken after the experiences of fintech lenders and are fast adopting new technologies and methods to expand their lending base.
Alternative data and insights to the rescue
Alternative data and insights could be extracted from unconventional sources, including a customer’s payment history such as rent, cell telecom bills, other utility bills, transactional details against a personal bank account, a customer’s use and repayment of certain alternative loans, etc. Moreover, data that may not necessarily reflect financial behavior, at first sight, might contain clues about their attitude towards economic commitments. Hence, monitoring social media and online activity can be considered as well. This data can be extensively applied in lending for fraud detection, loan decisioning, account monitoring, and collections. This has become even more important during the COVID-induced uncertainties in the business landscape. Knowing the overall health of the business in such trying times can save banks from falling for bad loans and NPAs later in the loan cycle.
Undoubtedly, this calls for pinning a lot of consent-based data that needs the approval of the applicant. Alternate data is not available in the public domain and needs a customer’s consent that his/her data be submitted to the third-party systems for processing. It is important to have a consent framework in place to ensure data privacy.
- Get consent from the user to analyze their personal data such as utility bills, digital footprint, etc.
- Based on the credentials shared, the system fetches the data from the service provider.
- Look at details to extract relevant information such as due amount, payment, info, delayed/on-time payment, etc. For instance, data points from the telecom operator can be analysed to interpret the usage and payment patterns to understand the user’s credit profile. A long trail of heavy call time and the same mode of payment, bank account, and timely payments before the due date can give an estimate about the stability of the profile. Where inconsistencies in the modes of payment can act as a red flag. Some of the use cases of telecom data include analysing default probability, operations of the company, network size, telecom bill efficiency, chances of using black money, and even knowing the technology-savvy quotient of the borrower.
Overall, default probability, operations of the company, network size, telecom bill efficiency, and transactions that involve illegitimate monies, are some of the aspects that a lender can analyse with alternative data. And, technology plays an indispensable role in it.
The underlying technology
A mechanism backed by technologies such as Machine Learning will be required to extract only sensible and usable information from the sea of data. With large, unstructured data sets, the right use of ML and artificial intelligence can help identify data patterns that relate specifically to credit risk. AI-driven credit intelligence platforms can track thousands of data sources in real-time to build comprehensive credit reports. These systems are scalable with no human intervention. This process and technology increase the likelihood of decisions being free of cognitive biases.
This could revolutionise the entire credit ecosystem. The benefits of a regulator-driven consent framework help in analysing the risk associated with entities that have less public data. This enables the lender to have a holistic view of the entity instead of being completely dependent on the financial statements and IT returns.