LINE BK upgrades its K Line Score to be smarter, aiming to select "good people" for loan applications.

LINE BK is revamping its unsecured loan model, upgrading K Line Score to integrate financial and behavioral data for a smarter loan product. The goal is to help identify promising customers that the previous system missed, aiming to expand its portfolio to 50 billion baht within three years.

Thana Pothikamjorn, CEO of Kasikorn LINE Co., Ltd., gave an interview reflecting on the challenges of providing unsecured loans through digital channels over the past 5-6 years at LINE BK, amidst a volatile economic environment. These challenges include pressure from the cost of living, household debt, and the current slow recovery of purchasing power.

LINE BK is Thailand's first social banking service provider, a joint venture between LINE and Kasikornbank (KBank), operated by Kasikorn LINE Co., Ltd. It officially launched in 2563.

Unsecured digital microlending is a high-risk model because it lacks collateral and the face-to-face interaction typically seen in traditional vehicle registration loans.

In its early stages, LINE BK faced significant trial and error. In its first year of operation, the credit cost exceeded 20%.

Thana said that in the early stages of providing loans, the risk assessment model was not skilled enough, resulting in inaccurate loan approvals and a number of defaulting borrowers. This forced them to revise the approval model.

For example, initially, 10 people applied for loans, but only 3 were approved. This was then adjusted to 1.5 applicants for 10 loans each, in order to maintain portfolio quality. Gradually, this adjustment is then made to increase the overall approval rate.

“LINE BK took the time to maintain the quality of its loan portfolio and improve its algorithms, eventually becoming profitable in its fifth year. We were able to reduce our credit cost to around 10%, lower than in the first year. The longer we operate, the more we can disburse loans. LINE BK has been providing loans for six years now, serving millions of customers. The past five years have taught us valuable lessons and increased our confidence, but we don't claim to know everything. We still feel we can do even better and haven't yet found the right people.”

Thana also stated that loans are now not just a source of funding, but have become a tool to help consumers manage their liquidity during times of volatile income. In the first half of 2569, LINE BK had over 8.8 million total users (12% YoY growth), with 9.3 loan users (15% YoY growth) and outstanding loans exceeding 29,000 billion baht.

The core of LINE BK is to create opportunities for the underbanked, especially those without a credit history and who are not salaried employees, such as vendors, traders, or self-employed individuals, who are often rejected by traditional commercial banking systems.

While adjusting its debt collection methods, which initially involved sending LINE stickers as payment reminders but proved ineffective, LINE BK has shifted to more rigorous and stringent payment reminder methods, including the use of both internal and external (outsourced) debt collection teams.

Upgrade your K Line Score's intelligence to filter out deserving loans.

Thana further stated that LINE BK's risk assessment also utilizes an internal model called K Line Score , which integrates data from two sources: financial data from Kasikornbank (KBank) and behavioral data from the LINE platform. The financial data from KBank includes account activity and transaction data, which allows KBank to understand the financial behavior of over 70% of the Thai population.

Regarding social behavior data obtained from LINE Data, privacy is a top priority. LINE BK cannot read users' chat messages, but it will analyze behavioral patterns using analysis of factors such as usage consistency, response time, interaction, and daily lifestyles, such as the wake-up patterns and work schedules of office workers, which reflect discipline and consistency in life.

Furthermore, Network Affiliation analyzes the connections between groups and peer networks using relational statistics to assess the overall risk profile. A key principle is "correlation is not causation," meaning that finding certain characteristics in a borrower within a risk group doesn't necessarily mean those characteristics are the cause of default. When similar patterns are found across a large dataset, the system can use these relationships as one of the factors in assessing risk.

The K Line Score uses behavioral variables and data to compare statistics with a group of actual borrowers through Big Data Predictive Analytics (not Generative AI), allowing for more accurate risk assessment than using financial data alone. Credit assessment uses the K Line Score in conjunction with payment history data from the National Credit Bureau (NCB).

Although LINE BK has gone through 5-6 years of learning and the system has become stable and profitable, Thana says that future development goals remain focused on several aspects. This includes approving good customers in groups that were previously rejected; for example, out of over 10 million loan applicants who approached LINE BK, only just over 1 million were approved.

It is believed that among the 9 million people who were previously rejected, it doesn't mean that everyone is a bad borrower, but rather that current risk assessment tools are not sophisticated enough to filter out good borrowers without causing harm. The main goal in the future is to develop smarter models to unlock more opportunities for this group to access formal credit. If, in the future, there is an Open Data system that can link data from utility bills or other behavioral data, it will help to assess risk more effectively.

“LINE BK aims for cautious growth in line with the economic climate, targeting 2 million approved loan customers within the next three years, achieving a total loan volume of 50 billion baht, and maintaining an average annual growth rate of 15%.”

LINE BK continues to prioritize responsible lending, promote financial discipline, and help manage liquidity sustainably, while adapting to the entry of virtual bank players and expanding business collaborations, including features like "refer a friend" and digital health insurance products with partners.

Read all news related to the banking industry here.