Lin, C. H., Hsu, S. F., and Chen, S. C. 2012. An Analysis of the Underwriting Cycle in Taiwan's Private Health Insurance: The Second-Order Autoregressive Model vs. the Markov Switching Model. NTU Management Review, 22 (2): 221-252. https://doi.org/10.6226/2012.(22-2).08
Chao-Hsin Lin, Associate Professor, Department of Risk Management and Insurance, National Kaohsiung First University of Science and Technology
Shuo-Fen Hsu, Professor, Department of Risk Management and Insurance, National Kaohsiung First University of Science and Technology
Su-Chu Chen, Ph.D. Candidate, Graduate School of Management, National Kaohsiung First University of Science and Technology
Abstract
The main purpose of this paper was to validate the underwriting cycle in Taiwan's private health insurance, and to discuss the variables that influence it. The results of the AR(2) model and the Markov switching model are compared to determine the cyclical pattern and to verify if the characteristics of the underwriting cycle remain the same before/after the structure changes and under different regimes. The empirical results show that there is no AR(2) cycle. However, they also show that the cyclical behaviour can be captured by the Markov switching model with the two regimes representing the soft and hard markets of Taiwan's private health insurance market. In addition, we confirmed our hypothesis that the variables affecting the underwriting profit margin differ substantially across the regime switch. The hypotheses regarding the underwriting cycle in the literature can not be fully explained by the linear model because the latent variable of state is omitted in regression. We demonstrated that the Markov switching regression model is well suited to capture the distinct characteristics of the loss ratio under different cycle stage with estimated time varying parameters.
Keywords
private health insurance underwriting cycle Markov switching model