How HDFC recast systems to offer loan approval, disbursement in 10 seconds

In the initial stage of the digital journey, the target was the urban market but later HDFC Bank decided to attack the semi-urban and rural pockets as well. Behind all these, the objective was to expand the business and bust the myth of the so-called base effect, which slows down a big bank’s growth rate. 

The bank’s employees took a while to understand what Aditya Puri wanted. When the boss told them to bring down the cost-to-revenue by 5 percentage points, the team came back with an estimate of 0.25 per cent. But Aditya wasn’t looking for a marginal change — they weren’t really getting the point. He wanted nothing less than transformation through digital.

For instance, if he was talking about a 10-second loan, it meant the entire process being completed in that time—the customer’s application, the approval (with the help of an in-built credit scoring model) and money credited into the account. It took six months after Aditya’s first meeting on his return from Silicon Valley to bring about fundamental changes in the way HDFC Bank was working.

It was not an easy task as there was complacency after creating what the employees believed was the best bank in India—it had the least bad assets and enjoyed steady growth that outstripped all its rivals comfortably. Aditya kept saying that whatever they had created was history and now they would have to reinvent the bank and themselves. Targets were set daily with three goals—customer convenience, competitive pricing and reduced costs.

The exercise involved an intense effort with most senior managers working without a break between September 2014–February 2015. These six months marked the beginning of HDFC Bank 2.0. Aditya was ready to take the responsibility for anything that went wrong but anyone who didn’t respond to the call of duty would eventually look for other opportunities although the bank never explicitly asked them to do so….

The 10-second push

Arvind Kapil was the first to come up with this innovative product. He introduced the concept of a 10-second loan for the bank’s existing customers that was launched in April 2015. Existing customers (about 30 million) were given a credit score based on the transactions, payment records and internal and external credit history. The credit team said about one-third of customers could be given pre-approved loans of up to Rs 1.5 million.

For a seamless experience, they integrated the front end and back end. The selected customers would be given an opportunity to fill in a few details, get an OTP and the money (loan) would be transferred to their accounts within 10 seconds — the world’s fastest loan approval and disbursement.

Let’s hear the story from Arvind, country head, unsecured, home, mortgage and working capital loans.

According to him, there were constant debates in the bank about how to get the product to consumers in the fastest and most convenient way—how to empower them through the do-it-yourself route. 

HDFC Bank 2.0 cover | Photo: Amazon

Arvind recalls a meeting in early January 2015 with a customer in Chennai who was running a large agro products supply chain. The customer was convinced about the bank’s products but was reluctant to seek a loan because of the vast quantity of documentation and multiple visits to bank executives involved. He asked Arvind whether he could get a loan sitting at home without having to call the bank. 

That was the story of most customers—happy with product offerings but not comfortable with the paperwork and multiple executives being involved in the process that starts from sanction of a loan to its disbursement. That was the inspiration for the 10-second personal loan product.

A senior IT executive in Delhi who had been banking with HDFC Bank for a long time once asked Arvind why would a customer seeking a loan need to provide so many documents—salary slips, PAN, address proof, and so on. “If you already have access to my salary account and all other cash flow statements, why do I need to prove it with paperwork?”

An unchartered territory

That set the ball rolling. Arvind started connecting the dots and working towards a seamless product for both the customer and the bank. It was unchartered territory. Initially, the process would be limited to 30 minutes. They were following a bottom-up approach and started thinking of ways to tackle the challenges on loan approval as well as disbursal.

The bank used to undertake income-based lending that required income tax returns. How would they do this online? Another important aspect was handling the risk. This meant that they had to be super careful but, at the same time, come up with an innovative product. Since the customer would not provide any information, they had to draw it from data they already had such as bank statements, credit/debit card transactions, information about funds flowing into the account, etc. Of course, this could be possible only if HDFC Bank was the customer’s primary bank….

In the initial phase, it was meant only for select salary account holders where the risk team had conducted various algorithmic and analytical tests. The product was launched on the bank’s net banking website on the evening of 30 April 2015. 

The first loan

The first loan was made at 4.18 a m  on 1 May 2015, to a 25-year-old process associate in a BPO in Jaipur who was working a 10 pm–5 am shift. “Talking to him we found out that as his working day never matched banking hours, so he found it very difficult to apply for a personal loan. While checking his balance online, he found the offer, clicked hoping to see a form and was surprised to get money in his account almost instantly,” Arvind says.

 Day One saw 215 customers availing of the offer for a 10-second personal loan online. The loans were generally Rs 0.3–0.4 million (capped at Rs 1.5 million). The team clocked a loan disbursal of almost Rs 3.5 billion in the first month without any advertisement.

The initial customer response was encouraging. During the early days of the product launch, Arvind’s team frequently called up customers to understand their experience on whether the product was meeting requirements and living up to expectations. As the bank gained confidence, it kept customising and personalising the product offering and ticket size, keeping in mind its relationships.

How does it work? 

How does the interface work? A customer logs onto the bank’s net banking website. If the customer is eligible, there is a loan option available at the top. The customer fills in some information (some fields are pre-filled), authenticates it using two levels of verification and the amount is credited to the customer’s account within 10 seconds. 

Arvind claims that in reality, the amount is sent to the account within 5–6 seconds. The USP is that for a customer, a loan cheque is available and encashable 365 days a year, 24 hours a day. By November 2018, 34 per cent of the bank’s personal loans were booked during non-banking hours.

How does it happen so fast? The credit team of the bank considers every customer of the bank as a potential loan customer and, after analysing data (which is continuously updated), awards a credit score on the basis of an algorithm. All such customers are “pre-approved” for a specific loan amount at a certain rate. The bank then informs the customer of this loan availability online. Once the customer confirms the loan requirement at a click, it gets disbursed immediately.

For the customer, the entire process is online but there is a lot of work at the back end where credit and risk teams are involved. 
 

The product is a culmination of the bank’s underwriting prowess and risk understanding, buttressed by a mass of customer data stored in a data warehouse explaining transaction and payment pattern and credit behaviour. There is no paperwork; the documentation is done online via OTP. …

The personal loan segment of the bank clocks around  Rs 60 billion in disbursements every month, the 10-second loans have one-third share of this. Customers are mostly 30–40 years old. The delinquency rate is on a par with loans disbursed physically.

The bank’s paranoia for risks has increased as it treads the digital path. For risk management, it has created a new credit risk underwriting framework christened P27, or the power of 27, as it is fondly called inside the bank. It incorporates the “ability to pay” income score and cascades it with the score of the credit bureau as well as a behavioural score of spending habits or past repayments and then mashes it up with the social score, catching an individual’s internet consumption patterns.

The campaign, Kil-Bil (Kill the Bad, Build the Good—inspired by the American martial arts film written and directed by Quentin Tarantino), is another interesting initiative introduced to engage the front line teams to use data and insights to improve credit underwriting and portfolio quality.

(Excerpted with permission)