Political uncertainty, realized volatility and jumps in the Chinese stock market

Published: 10 March 2020| Version 1 | DOI: 10.17632/f3frtvt6f4.1
Yun Hong


Our initial sample consists of all the Chinese A-share listed firm samples spans from January 1, 2014 to December 31. The 5-min frequency price data and the single trading data containing the price, volume and the buying/selling initiation information are collected from the CSMAR high-frequency database; the former data are used to calculate the RV. Following Wang et al. (2016), we use the local official replacement to measure the LPU, that is, for the firm i in day t (in month m), LPUi,t takes the value of 1 if observing a mayor or the party head turnover during the (m-1, m+1) months in the firms’ registered prefecture city, and 0 otherwise. The detailed information of local government official replacements is collected from China Economic Net (http://www.ce.cn/) and other authoritative government website manually. The political sensitively data include the “innate” state ownership (SOE) and the “postnatal” Chairman/CEO political connection (PC), which are collected from the CSMAR database and compiled as Fan et al. (2007). We delete the sample-firms going public and the firms with trading days less than two years, and the samples with missing variables. Finally we get 2048676 daily samples comprising 2318 firms, with the maximum length of 1219 trading days.



Hunan University Business School


Financial Economics