| The risk and risk volatility of the stock market has always been one of the issues of concern in the field of finance.Only by mastering the risk characteristics of the stock market can we make the right decisions and thus obtain stable returns.However as globalization trends sweep across the globe,interactions between countries and regions become more frequent,The impact of financial markets in other countries on the domestic stock market has become one of the external factors.Therefore,this paper studies the existence and transmission of volatility spillovers between different stock markets.But,with the rapid development of hardware and algorithms,financial high-frequency data has become available,so many statistical methods for modeling high-frequency data have emerged.Among them,the functional data analysis method has the characteristics of modeling for function objects and the advantages of adapting to high frequency and mixed data processing.Therefore,this paper uses the functional linear regression model which is one of the functional data analysis method to study the volatility spillover between different stock markets.Different from the traditional econometric model which use scalar data to build models,this paper uses the whole daily stock market yield density curve as a sample instead of scalar data to model.Apart from this,it is the first in China to apply the functional autoregressive model and its moment correlation test to the study of stock market volatility spillovers.This paper first uses the kernel density estimation method to fit the high-frequency intraday stock returns to the density function values at each quantile level.Then fit the discrete density function values into functional data.using data reconstruction method in the function analysis method.Then,using the intraday return rate distribution of the Shanghai Composite Index as a research object to construct a functional autoregressive model to verify the feasibility of the model.Finally,using the functional linear regression model to discuss whether the US stock market represented by the Standard & Poor's 500 Index has obvious volatility spillover effects on the Chinese stock market represented by the above-mentioned KLCI.The main conclusions of this paper are as follows:(1)The distribution of the yield of the Shanghai Composite Index on the day is indeed affected by the distribution of the yield of the lag phase,and the mean and variance of the current yield distribution are simultaneously affected by the mean variance of the yield distribution of the lag phase,while the current yield distribution Higher order moments such as kurtosis and skewness are only affected by the moments of the corresponding order of the lag phase of the yield distribution.It is also confirmed that the hypothesis test of the regression operator of the functional regression model can indeed obtain the influence of the distribution on the distribution.(2)The functional regression model has better prediction effect,and can better predict the distribution of the yield of the Shanghai Composite Index in the next few days,indicating that the functional regression model does extract the characteristics of the training data.However,when the stock encounters large fluctuations,the prediction effect will have some distortion.(3)Within the whole sample interval,the intraday return rate distribution of the S&P 500 index lags in the first period has a significant spillover effect on the intraday yield distribution of the Shanghai Composite Index,and this spillover effect is reflected in the mean,variance,kurtosis and Skewness on each moment.(4)During the different time periods,the intraday return rate distribution of the S&P 500 index lags in the first period,which has different spillover effects on the intraday yield distribution of the Shanghai Composite Index.The spillover effect of kurtosis and skewness always exists.It is affected by the moment of the corresponding order.Mean and variance,which show strong and weak spillover effects over time. |
PDF Full Text Request