What is digital lending?

Lending is simply the act of giving money on credit to another person called the borrower. The borrower repays the money to the lender with interest over a defined time-period.

Lending is a widely understood concept - banks lend to people like you and me for buying homes, cars, education or even for personal use. Banks and other financial institutions also lend money to corporates or institutions so that they can undertake projects or build new products.

The lending process begins with bringing a borrower on-board through various sales channels. The next step is to collect information about the borrower such as his identity, financial history, income, etc (remember, even big corporates have a financial history) along with document proofs. This is called Know Your Customer or KYC. Once the lender has collected all required information, the lender has to make a decision if he would like to give a loan to the customer, the amount of the loan, interest rate and time period to repay - this is called underwriting. After the loan is disbursed, the borrower is supposed to make periodic and timely repayments to complete the loan. If the borrower defaults (does not repay), the lender may have the authority to recover the loan by taking over the borrower’s assets, but this is not always the case.

Digital lending attempts to perform each step of this process through paperless or electronic means. For example, banks sell home loans, auto loans, personal loans or credit cards through digital channels such as email, Facebook, Google, etc.

India has recently introduced e-KYC through which a bank can use an individual’s Aadhaar information to verify his/her identity. Almost all banks have internet-banking or mobile banking facilities through which a user can quickly download his financial history and demonstrate his/her ability to repay a loan. Traditionally, ability to pay was demonstrated by submitting copies of bank passbooks and income proofs.

Certain banks and startups also use digitized modes of paperwork - that is accept scans of these documents from the customer. This reduces the step of visiting a bank branch and provides the convenience of doing everything from home at your convenience.

The most important step of underwriting a customer can be completely automated by a machine or an algorithm which crunches this data, learns from it’s own underwriting history and makes a decision to lend. Many start-ups are now betting on this approach to disrupt traditional models of lending.

These algorithms attempt to improve on existing models of underwriting by constantly learning from data and reduce Type I errors in hypothesis testing. In other words, they try to approve customers who should really be eligible to take loans but cannot through the traditional lending models.

Disclaimer: I work at a digital lending startup called Paysense