"Alternative Credit Scoring Must Overcome Privacy and Transparency Issues"

Hana Financial Management Research Institute Report Presentation
"Many Challenges to Address but the Path Forward"

As the U.S. central bank, the Federal Reserve (Fed), hinted at additional interest rate hikes, attention is increasing on whether the Bank of Korea's Monetary Policy Committee will raise rates at its meeting scheduled for the 25th. The photo shows a loan counter at a commercial bank in downtown Seoul on the 19th. Photo by Kim Hyun-min kimhyun81@

[Asia Economy Reporter Minwoo Lee] Alternative credit scoring, which evaluates borrowers' creditworthiness based on non-financial information, is emerging. As the related market is expected to grow further, analyses suggest that issues such as personal data protection and transparency must be overcome and long-term strategies established.

On the 23rd, Hana Financial Management Research Institute released a report titled "Alternative Credit Scoring, Can It Be an Alternative?" Alternative credit scoring evaluates borrowers' creditworthiness using non-financial information. With various information being digitally stored and data processing speeds increasing, it has become possible to assess various non-financial data such as rent and utility payment histories almost in real-time, enabling the evaluation of creditworthiness for 'thin-file' borrowers who lack sufficient financial transaction history.

Nevertheless, there is still a long way to go. First, there is the issue of personal privacy infringement. Legal problems may arise if personal information is included in alternative data. It is also crucial to supervise whether the system transparently utilizes the information. It is difficult to manage fairly and transparently which information is used and how this information affects loan interest rate decisions. There is a risk of 'malicious' consumers emerging. If it becomes known that certain information significantly impacts credit scores, borrowers might exploit this. Yoonyoung Jung, Senior Researcher at Hana Financial Management Research Institute, analyzed, "Individuals might deliberately delete information on social network services (SNS) or post favorable information to manipulate the system."

Technological capability is another hurdle. Unlike traditional financial information, which can be relatively easily obtained in a structured form, alternative information requires a higher level of technology. Building AI-based machine learning models using alternative data requires a sufficient amount of data, and since most of it is unstructured data, the processing is more complex than with financial information. Researcher Park explained, "To improve the suitability of alternative credit scoring models, it is important to use accurate and homogeneous information. However, alternative data may include highly variable or worthless data, which can reduce model accuracy," adding, "Securing specialized personnel capable of utilizing alternative data is also essential but challenging."

Despite these challenges, the report emphasizes that this is a 'path to take.' The alternative credit scoring market is already opening not only overseas in countries like the U.S. and China but also domestically. Since the implementation of the so-called 'Data 3 Laws' (amendments to the Personal Information Protection Act, Credit Information Act, and Information and Communications Network Act) in August 2020, which allow the combination of non-financial information across different industries, internet-only banks, big tech companies, and card companies have been entering the alternative credit scoring market. Naver Financial launched a smart store business loan product exclusively for online merchants using non-financial data such as the proportion of regular customers, customer reviews, and return rates. Kakao Bank also developed a new credit scoring model using telecommunications payment information in June last year and applies it to loans. Additionally, Toss Bank and K Bank have prepared their own credit scoring models using similar methods.

Researcher Park advised, "Since alternative credit scoring can expand into new areas such as mid-interest loans and Buy Now Pay Later (BNPL), the market will grow despite various challenges," adding, "Financial companies need to establish long-term strategies for applying alternative data, and financial authorities should consider measures to prevent side effects."

This content was produced with the assistance of AI translation services.