Empirical Analysis Of Market Risk Of Real Estate Trust Fund In Hong Kong Based On ARMA-GARCH Mode Family

In recent years,China’s real estate credit policy has been tightening,forcing real estate to find new financing channels.Real estate trust,a new financing mode,has emerged as required,providing a new way of thinking for the financing of housing enterprises.By the end of March 2019,49 real estate trust fund products had been issued in China,with the issuance scale reaching 98.964 billion yuan.Firstly,this paper combs the relevant research of real estate trust funds at home and abroad,then expounds the relevant concepts and characteristics of real estate trust funds,risk management and market risk.This paper combines theoretical analysis with empirical research,adopts VaR and CVaR risk measurement methods and ARMA-GARCH model family to conduct empirical research on the market risk of Hong Kong Hang Seng REITs.In the specific empirical analysis,first,the logarithmic return series of the selected funds are tested by unit root test and ARCH-LM test to judge the stability and heteroscedasticity of each series.According to the empirical test results,the selected series of Hong Kong real estate trust funds in this paper has better stability and heteroscedasticity.It is suitable to construct ARMA-GARCH model to predict the logarithmic return rate of Hong Kong real estate trust fund.Through GARCH model,EGARCH model and TGARCH model,the fitting effects of different models are analyzed under different distribution assumptions.It is found that EGARCH model can better reflect the return of Hong Kong real estate trust fund.The peak and fat tail characteristics of rate series,asymmetry of return rate and information impact are asymmetry.On the basis of GARCH model family fitting,under different confidence levels of 95%and 99%,VaR measurement model is used to measure market risk under different residual distribution.The results are compared and Kupice test is adopted.In order to measure the tail market risk better,CVaR model is also introduced based on VaR model.The results of VaR model based on ARMA-GARCH model family and CVaR model are compared under different residual distributions at different confidence levels,so as to measure and screen out the best risk measurement model.Empirical analysis shows that CVaR is better than VaR to measure the extreme market risk of Hong Kong Real Estate Trust Fund.It also verifies that the CVaR model of EGARCH model under GED distribution can better measure the market risk of Hong Kong Real Estate Trust Fund.