Application of Markov Model in Long-Term Care Insurance

Abstract

The Markov process is not only the actuarial basis of pricing of long term care (LTC) insurance but also the fundamental for predicting the future elder population and disabled population. This article aims to summarize how the Markov processes or Semi-Markov processes are used in the Long-term care risk and Long-term care insurance. We also discuss the models based on the time-homogeneous and time-inhomogeneous. Moreover, under the GLM framework, some studies show Tweedie-GLM would give more accurate predictions compared with other GLM models and additive models. However, these models, whether based on the Markov process or Semi-Markov process or GLM, have theoretical advantages due to the natural features, the researchers would quickly build the multi-state models, there are still exits many challenges, and they provoke the researchers into some tries of how to deal with the limitations of data, the development of medical technology, and the longer expectancy of life.