DOI QR코드

DOI QR Code

Survival analysis for contract maintenance period using life insurance data

생명보험자료를 이용한 계약유지기간에 대한 생존분석

  • Yang, Dae Geon (Department of Statistics, Pukyong National University) ;
  • Ha, Il Do (Department of Statistics, Pukyong National University) ;
  • Cho, Geon Ho (Division of Cosmetic Science and Technology, Industrial Quality Engineering, Daegu Haany University)
  • 양대건 (부경대학교 통계학과) ;
  • 하일도 (부경대학교 통계학과) ;
  • 조건호 (대구한의대학교 화장품공학부 산업품질공학전공)
  • Received : 2018.09.18
  • Accepted : 2018.10.05
  • Published : 2018.12.31

Abstract

The life insurance industry is interested in various factors that influence the long-term extensions of insurance contracts such as the necessity for the advisors' long-term management of consumers, product consulting, and improvement of the investment aspects. This paper investigates important factors leading to a long-term contract that forms an important part of the life insurance industry in Korea. For this purpose we used the data of contents (i.e., data from Jan 1, 2011 to Dec 31, 2016) of the contracts of xxx insurance company. In this paper, we present how to select important variables to influence the duration of the contract maintenance via a penalized Cox's proportional hazards (PH) modelling approach using insurance life data. As the result of analysis, we found that the selected important factors were the advisor's status, the reward type 2 (annuity insurance) and tendency 4 (safety-pursuing type).

최근에 생명보험 산업은 보험계약의 장기 연장에 영향을 미치는 다양한 요인들에 관심을 두고 있다. 예를 들어 모집 설계사의 장기간 고객관리의 필요성, 상품상담, 투자측면의 개선 등이다. 따라서 본 연구에서는 우리나라 생명보험사의 장기계약을 유지하는 중요한 요인들을 조사하고자 한다. 이를 위해 우리나라의 모 생명보험사의 2011년 1월 1일부터 2016년 12월 31일까지의 계약건의 내용에 대한 데이터를 사용하였다. 본 논문에서는 이러한 데이터를 사용하여 벌점화 콕스 비례위험모형 접근법을 통해 계약유지기간에 중요한 영향을 미치는 변수를 선택하는 방법을 제시한다. 분석결과 설계사의 변경 유무, 연금 상품군, 그리고 안정적 투자성향과 같은 세 가지 변수가 계약건 유지에 주요한 요인으로 선택되었다.

Keywords

GCGHDE_2018_v31n6_771_f0001.png 이미지

Figure 3.1. Plot of survival curves of Kaplan-Meier.

Table 1.1. Explanation for variables of the life insurance data

GCGHDE_2018_v31n6_771_t0001.png 이미지

Table 3.1. Basic statistics for the life insurance data

GCGHDE_2018_v31n6_771_t0002.png 이미지

Table 3.2. Comparison of survival rates among groups (P : p-value)

GCGHDE_2018_v31n6_771_t0003.png 이미지

Table 3.3. Results of fitting Cox’s proportional hazards model for life insurance data

GCGHDE_2018_v31n6_771_t0004.png 이미지

Table 3.4. Variable selection and estimation using Cox’s proportional hazards model for life insurance data:estimates (SE)

GCGHDE_2018_v31n6_771_t0005.png 이미지

References

  1. Choi, Y. M. and Choi, W. (2008). The effect of economic variables on lapse and surrender rate in life insurance, Journal of Insurance and Finance, 55, 3-36.
  2. Cox, D. R. (1972). Regression models and life tables (with Discussion), Journal of the Royal Statistical Society B, 74, 187-220.
  3. Fan, J. and Li, R. (2001). Variable selection via nonconcave penalized likelihood and its oracle properties, Journal of the American Statistical Association, 96, 1348-1360. https://doi.org/10.1198/016214501753382273
  4. Fan, J. and Li, R. (2002). Variable selection for Cox's proportional hazards model and frailty model, The Annals of Statistics, 30, 74-99. https://doi.org/10.1214/aos/1015362185
  5. Ha, I. D., Jeong, J. H., and Lee, Y. (2017). Statistical Modelling of Survival Data with Random Effects: H-likelihood, Springer, Singapore.
  6. Ha, I. D., Noh, M., Kim, J., and Lee, Y. (2018) frailtyHL: frailty models using h-likelihood, From: http://CRAN.Rproject.org/package=frailtyHL.Rpackageversion 2.1.
  7. Ha, I. D., Noh, M., and Lee, Y.(2012). frailtyHL: a package for fitting frailty models with h-likelihood, The R Journal, 4, 307-320.
  8. Ha, I. D., Pan, J., Oh, S., and Lee, Y. (2014). Variable selection in general frailty models using penalized h-likelihood, Journal of Computational and Graphical Statistics, 23, 1044-1060. https://doi.org/10.1080/10618600.2013.842489
  9. Kaplan, E. L. and Meier, P. (1958). Nonparametric estimation from incomplete observations, Journal of the American Statistical Association, 53, 457-481. https://doi.org/10.1080/01621459.1958.10501452
  10. Kim, B., Ha, I. D., Noh, M., Na, M. H., Song, H. C., and Kim, J. (2015). Variable selection in frailty models using frailtyHL R package: breast cancer survival data, The Korea Journal of Applied Statistics, 28, 965-976. https://doi.org/10.5351/KJAS.2015.28.5.965
  11. Kim, S. Y. and Lee, J. J. (2012). A comparative study on the effects of macroeconomic variables and consumer's dissatisfaction on the lapses of life insurance contracts, Korean Journal of Insurance, 93, 237-269.
  12. Lee, Y. and Oh, H. S. (2014). A new sparse variable selection via random-effect model, Journal of Multivariate Analysis, 125, 89-99. https://doi.org/10.1016/j.jmva.2013.11.016
  13. Tibshirani, R. (1996). Regression shrinkage and selection via the Lasso, Journal of the Royal Statistical Society Series B, 58, 267-288.
  14. Tibshirani, R. (1997). The Lasso method for variable selection in the Cox model, Statistics in Medicine, 16, 385-395. https://doi.org/10.1002/(SICI)1097-0258(19970228)16:4<385::AID-SIM380>3.0.CO;2-3