An IIT Madras alumni, Rahul expertises in use of data modeling to assess risk and evaluate credit worthiness

India always had relationship with driven banking and lending. To open a savings account in the State Bank of India, one still needs `an introduction from a respectable person / a customer of the branch'. The local auto garage only gets loans from the neigh-borhood bank, if the branch manager knows the proprietor personally. This seems like a great mitigant of risk. With NPAs soaring, this seems like a reasonable strategy to control bad loans. However, this is only a false comfort. Most of the NPAs come from large accounts greater than INR 1 Crore. What this relationship driven finance has done is that it has made credit very exclusive. Credit cards are seen as a luxury when it needs to be a necessity like mobile phone and internet. Credit is essential to break out of the poverty cycle.

Technology is yet to disrupt the lending space as it has done in entertainment, e-commerce, payments and to some extent in Banking. Everyone is watching videos & live events on their mobile phones. Amazon delivered packages to 99 percent of India's pincodes. Tea vendors on bicycles are accepting PayTM. Banks no longer want customers walking into their branches. However, lending is lagging behind. While the loan or the credit card application process has been digitized, there is a lack of awareness among users. Many are still apprehensive of using lending apps. Submitting sensitive information like ID proofs, Bank Statements through a mobile app is still a challenge for most. The underwriting policies are still archaic in most lending institutions. They are yet to embrace data science when it comes to the lending decisions. Post approval formalities are still largely

paper-based. The customer still has to sign agreements and provide postdated security cheques.

Disruptions happen on a strong backbone of internet, GPS and mobile penetration. For lending, in addition to mobile and internet, the disruption enabling platform is India Stack. The stack consists of different layers including Aadhaar eKYC, eSigns, UPI and consent driven data sharing. The process from identifying a user's intent to borrow all the way through to repayments will happen paperless and presence less. In the next five years we will see all users authenticate them-selves with Aadhaar and sign agreements with eSigns. UPI 2.0 will automate payments and remove the need for physically signed cheques. A consent driven data sharing layer will make the process of submitting bank statements, IT returns and utility bills seamless and secure for the users.

In the next five years we will see all users authenticate themselves with Aadhaar and sign agreements with eSigns. UPI 2.0 will automate payments and remove the need for physically signed cheques

An important piece that needs to mature more is underwriting pro-cess. While there are many smart NBFCs that have automated the underwriting process, the larger ones still rely on human intervention and subjectivity. This is due to the myriad forms in which an applicant's data could come. Address proof, for example, can be one of 20 plus document types each of which has no standard format. Data science, ML and AI are ripe to start experimentations in this space. Once the data formats are standardized, algorithms can munch through personal documents and financial statements in no time and without human errors or fatigue. These algorithms can learn over time to discern between good and bad borrowers. There is ample research to prove how psychometric traits like impulsiveness and delayed gratifications are strongly related to a person's financial discipline. We will see more this in the coming years. One word of caution though - AI and ML can go horribly wrong. While we avoid human errors, computer bugs and faulty mathematical models will crop up.

Data privacy is an area for regulation to catch-up. Lending organizations will have access to private data of potential and existing borrowers. The need of the hour is to make it is easy for the user to know what data is being collected and constraint the sharing this data with 3rd parties. European Union has just taken a bold step in this direction with General Data Protection Regulation (GDPR). Something along similar lines needs to emerge in India also.

Credit has not reached people who need it the most. Technology is yet to disrupt lending like some of the other industries. Mobile and internet penetration along with India Stack is a transformative platform that will launch the lending industry. What we require is careful, innovative experimentation and data privacy regulations.