The Application Of Wavelet In Analysing The Underwriting Cycles In China’s Property Insurance Market

Conning&Company once published an article entitled "A study of why underwritingcycles occur",which proved the existence of the underwriting cycle for the first time in1979.Subsequent studies on the underwriting cycle have sprung up like mushrooms after rain sincethen. Li Qian is the first domestic scholar to do researches on the underwriting cycle in1996.However,the studies on the field of research made little progress in the early2000, butafter the financial crisis in2008, the study of the underwriting cycle has become a hot,andthere are three fields that scholars have focused on,ie the existence of underwriting cycles,length measurements on the cycles, and the reasons for the underwriting cycle. Second-orderAutoregressive, Spectral Analysis, Filtering, Multiple Regression, Cointegration test andVAR model are widely used methods,and the studies focus on motor vehicle insurance,property insurance market, the non-life insurance market.Firstly, domestic scholars usually think that the existence of the underwriting cycle has adeep relationship with macroeconomic factors.This paper try to lay a solid theoreticalfoundation for further explanation of the existence of the underwriting cycle.Secondly, the phenomenon of foreign property and casualty insurance underwritingcycle industry has been widespread, but insurance industry practitioners at home are notfamiliar with this concept, only a few papers have fouced on the underwriting cycles. Whilethe cycle measurement methods are still followed the Western countries,such as traditionalthe regression model framework. We believe that, thanks to self-premise regression modelassumes the existence of the insurance industry profits fluctuate a fixed period, but frequentchanges in insurance industry since the reform and opening up policy, market volatility anddowns, this feature will inevitably lead to fluctuations in underwriting profit cycle is notfixed. Therefore, a simple regression model applied to China since the underwriting cyclestudies, often come to me there is no domestic market, insurance industry cycles conclusions.For these reasons, this paper studies on the underwriting cycle using non-stationary timeseries which can be processed directly by using wavelet analysis. It is proved that theunderwriting cycle of domestic non-life insurance market is exsiting,and the length is3-4years.Thirdly, most of our scholars study the existence of the underwriting cycle insuranceindustry and motor vehicle insurance industry.However,it is indicated that there may bedifferent types of insurance underwriting cycles of different lengths according to scholars Venezian (1985), Cummins et al (1987), insurance industry underwriting cycle and may alsohave different cycle lengths. In this paper, we do not only study the underwriting cycles ofproperty insurance,but also study the cycles of life accident insurance and health insuranceresearch and non-life insurance market.Finally, based on China’s insurance premiums, gross domestic product, consumer priceindex and fixed-asset investment data from1999Q1to2013Q1, we analysis the long-termand short-term relationship between them using the ARDL bounds testing approach. Theresults indicate that GDP is the major factor driving the growth of property insurancepremiums in China; fixed-asset investment has significant impact on China’s propertyinsurance premium income and they are adversely related; consumer price index has almostno effect on the premium income.