The Cross-Subsidization from Imperfect Risk-Classification─ Imperfectly Risk Classification of Gender

• Main Authors: Shu-Rong Shie, 謝淑榕
• Other Authors: Kili C. Wang
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/82016368027062947529

碩士 === 淡江大學 === 保險學系保險經營碩士班 === 98 === It is the fact that the insureds of automobile insurance are different from real drivers in Taiwan. In intuition, female and the married were less involved in an accident. However, the Taiwan Insurance Institute shows, over the age of 25 married women who have a higher percentage of loss frequency and severity compared with unmarried women. In addition, the Ministry of Transportation and Communications Department of Statistics indicates that only a vehicle is used usually by the male in the family. Thus, the purpose of this paper is to show whether the insurers can’t observe the hidden information that the insured is different from the real driver, and focus on the variable of gender to find the empirical evidence of the cross-subsidization from imperfect risk-classification. We employ the OLS regression method and measure the loss ratio in terms of loss severity. As a result, we find the evidence that many variables have not be sorted out completely. Futhermore, the findings that cross subsidization with the policies of the female insured is different from the real driver, especially in collision automobile insurance and non-dealer of car associated with the sale of automobile insurance.