| After more than 30 years of rapid development,China stock market has great potential for further growth,as access conditions gradually relax and a comprehensive stock issuance registration system is implemented.However,due to its late start,compared with more mature stock markets,China stock market exhibits characteristics such as high volatility of the stock index and high percentage of private individual investors,indicating its immaturity.Therefore,exploring the factors that affect China stock index is of great significance for the sustained and healthy development of China stock market.Forecasting stock returns is one of the core of financial statistics research.Due to the close connection between macroeconomics and stock market,the predictability of stock returns exhibits an identifiable pattern related to the dynamics of economic cycles.Many scholars at home and abroad have also explored the impact of macroeconomic fluctuations on stock returns from different perspectives and using different methods,however,their studies have some problems such as focusing on individual stocks,weak representativeness of indicators,and short data periods.This paper thus attempts to improve on the above issues based on a macro perspective.This paper selects six monthly macroeconomic indicators of economic growth,inflation,fiscal policy,monetary policy,interest rate and exchange rate,and uses the CSI300,SSE50and CSI500 closing price indices to represent A-share main board,large-cap and small-cap stocks respectively.Then explores the correlation between macroeconomic and stock market through vector autoregression.The results show that:(i)there is a significant correlation between macroeconomic indicators and the three stock indices;(ii)the impact and direction of each economic indicator on the CSI300 and SSE50 indices are similar;(iii)the commodity retail price index,which measures inflation,is the main reason for changes in the three indices.On this basis,expand the selection range of macroeconomic indicators,and the comparative evaluation of the horizontal and vertical forecasting effects of Chinese macroeconomic indicators on different stock price indices is examined under different quantile points using 161 macroeconomic indicators of 13 categories and stock price indices of three different sectors from February 2005 to February 2022.The initial estimation period is set from 2005M2 to 2015M3(122 periods),and the forecast evaluation period is set from2015M4 to 2022M2(83 periods),which is split into the pre-crisis period(2015M4-2018M3)and the financial crisis(2018M4-2022M2)subpe-riods.The results show that:(i)the quantile factors around q=0.5 can be well explained by the PCA mean factors;(ii)the R~2 at the upper and lower tail quantiles are lower,and the R~2 decrease as the quantile moves towards both ends,indicating that these additional factors will help predict the variables;(iii)adding upper tail quantile factors to the original forecasting model during the estimation period and economic mild period set by us will improve the predictive performance of the model;(ⅳ)the quantile factors can provide hidden information for the density forecasts of macroeconomic indicators on the stock price index.Based on the above empirical research results,this paper summarizes the main findings and proposes relevant policy suggestions from the perspectives of investors,market,and government.(i)In terms of persuading investors,the government needs to educate investors not to follow blindly through various channels.Meanwhile,strengthen education,supervision and punishment of financiers;(ii)In terms of regulating market operations,it is necessary to further improve China stock market information disclosure system and impose heavy penalties for fraudulent activities in the market;(iii)On the government side,the government needs to improve and introduce laws targeting China securities market to ensure the normal operation of the stock market and ensure that the policies launched are in line with the market reality. |
PDF Full Text Request