How Automation Speeds Up Loan Approvals

Personal loans were still a niche product when peer-to-peer (P2P) lending company LendingClub was founded in 2007. Today, PYMNTS research shows that 24% of U.S. consumers have used personal loans, the popularity of these loans increasing even more. during the pandemic.

As consumers faced the financial crisis of the pandemic, they were also increasingly seeking debt repayment relief through hardship plans, under which borrowers could make interest-only payments or completely ignore payments. These plans put additional pressure on systems that were not designed for a sudden influx of such requests, according to Anuj Nayar, financial health manager at LendingClub. Rather than letting borrowers spend hours on the phone, LendingClub has developed a digital solution.

“The first thing was just to make sure we could make it very easy for people to apply for a range of plans as their financial situation changed, to make sure they got what they needed with an online service. easy,” he said. “We were one of the first to implement the idea that you could request a difficulty plan online.”

Nayar said about 12% of the company’s loan base applied for hardship plans at the height of the pandemic. That number has now dropped to around 1%. At the same time, rather than consolidating credit card debt — the service LendingClub is best known for — many applicants seek personal loans for personal and business expenses, such as setting up home offices. This dynamic lending landscape requires data-driven solutions that can quickly adapt to new changes.

Speed ​​up the loan approval process

Even before the pandemic, online and mobile options changed the way people did business. The events of 2020 have accelerated this transition and the financial sector has been slow to catch up.

“Retail is now mostly online, especially during the pandemic,” Nayar said. “Booking a taxi is an online service. All of those things have changed, but the bank hasn’t been fundamentally rebuilt from the ground up until now.

With a Net Promoter Score (NPS) of 78, LendingClub attributes its customer satisfaction to helping customers eliminate credit card debt, as well as the fact that nearly half of its employees are in customer support. A survey showed that live personal service combined with digital self-service speeds up the loan approval process, for example. Nayar said about half of the company’s customers return to LendingClub for additional loans. Customer satisfaction brings borrowers back, but so does a faster approval process. With its digital lending approach, LendingClub can use data already collected on existing customers to approve returning candidates more quickly.

“Buzzwords are [artificial intelligence (AI) and machine learning (ML)] and all that, but we give $1 billion in loans a month,” he said. “You couldn’t make that amount if you weren’t using the latest technology. We have what we believe to be one of the largest data sets of all our peer sets. We have built something like 150 billion data cells on our customer base.

The data offers insight into consumer credit behavior across different credit environments. Coupled with constant testing and evaluation of everything from marketing to loan servicing and even fraud management, this data helps LendingClub meet customer needs while mitigating risk.

According to Nayar, the three main things borrowers want to know are: will they get the money, when will they get the money, and how much will the loan cost them? Whether applying for a car loan or a personal loan, receiving money quickly is important to all borrowers. Products such as balance transfer, in which money is electronically paid directly to a creditor, help speed up the process and facilitate approvals by mitigating the risk of sending money directly to a borrower.

With an unfamiliar risk environment at the start of the pandemic, LendingClub focused on its known customer base, reducing loan origination by approximately 90%. This slowdown gave the company time to dig into systems and data. Back to pre-pandemic lending levels, LendingClub has automated about 80% of its lending. Approvals for these loans can be done within hours, with disbursements being made to borrowers within a day or two.

Mitigating the disparate impact of automated approvals

Automation has brought speed and convenience, but it has also introduced potential pitfalls. In the past, minorities or other disadvantaged groups may have faced adverse decisions from loan officers. Now, lenders face the challenge of ensuring their algorithms don’t inadvertently discriminate.

“How do you deal with speed and simplicity ensuring that you are not, unfortunately, not by intent, but by actions, creating something that for some reason unfairly inhibits a particular protected class? ” Nayar said. “You can do things fast, but are you also making sure you’re doing it right?”

It’s possible, as technology advances, that a FinTech might not be able to explain why its automated process gave different rates to different borrowers, Nayar said. If a human made the decision, there would be a remedy for apparent harm, but a lender may not know where to start to ensure that an automated process is not discriminatory.

“It’s not OK,” he said. “We need to be able to make sure we’re not driving – even unintentionally – some of these things.”

LendingClub recently joined the National Community Reinvestment Coalition (NCRC) and other FinTech companies in calling on the Consumer Financial Protection Bureau (CFPB) to weigh in on lenders’ obligations to guard against disparate impacts.

“I think that’s an important point, because you’re thinking about speed and simplicity, to also consider fairness,” Nayar said.