Academic SeminarYour Movement in a City Tells Your Credit: Credit Default Prediction based on Geolocation Information
- 일시
- 2019-05-02 ~ 2019-05-02
IT경영 분야 세미나를 아래와 같이 개최하오니, 관심 있는 분들의
많은 참석 부탁 드립니다.
1. 일시: 2019년 5월 2일 (목), 16:00~17:20
2. 장소: 수펙스 경영관 101 강의실
3. 강사: 방영석 교수 (홍콩중문대학교)
4. 주제: Your Movement in a City Tells Your Credit: Credit Default Prediction based on Geolocation Information
5. 연구분야: IT경영 분야
* Lecture will be delivered in Korean.
* Seminar materials: Abstract
[Abstract]
Do people with a high credit default risk visit different locations in a city, compared to where people with a low credit default risk visit? We propose that geosimilarity risk and the geolocation network size of people comprise a critical classifier that predicts their credit default. Two people are defined as geosimilarity network (GSN) neighbors to each other if they share a visited location during a specific period. Based on the consumer-location and loan-repayment data from a leading FinTech company, we found that the GSN neighbors of a person who defaulted on a loan are approximately three times more likely to default compared to the average default rate, and are approximately 4.5 times more likely to default compared to the GSN neighbors of a person who has not defaulted. The geosimilarity risk and geolocation network size significantly explain credit defaults after controlling for traditional factors such as demographics, financial ability measures, and loan characteristics. In addition, incorporating these measures into the traditional model improves the prediction accuracy of credit default by approximately 9%.