The idiosyncratic risk in Chinese stock market

Julia Darby, Hai Zhang, Jinkai Zhang

Research output: Contribution to conferencePaperpeer-review

85 Downloads (Pure)

Abstract

Using daily stock return data of all listed firms in Chinese stock market from 1998 to 2018, we disaggregate the volatility of common stocks at the market, industry and firm levels. We find market volatility, on average, is the highest while firm volatility tends to lead to market and industry volatility series. None long-term trend time series behaviour exists for all three volatility series and firm volatility is best described by an autoregressive process with regime shifts associated with the structural market reforms or volatile market movements. We further proceed to identify the source of volatility at the industry level and find the idiosyncratic volatility in the largest manufacturing industry not only accounts for the largest proportion in the aggregate firm volatility but also is the lead indicator for the idiosyncratic volatility of other industries. Finally, unlike Brandt et. al. [Review of Financial Studies 23(2): 863-899 (2010)], we find the idiosyncratic volatility in Chinese stock market is associated with high stock trading activities by institutional investors, the result of which is also robust when using other measures of idiosyncratic volatility.
Original languageEnglish
Number of pages36
Publication statusPublished - 9 Sept 2019
Event22nd Dynamic Econometrics Conference in Nuffield College Oxford, UK. -
Duration: 9 Sept 201910 Sept 2019
https://www.nuffield.ox.ac.uk/news-events/events-and-seminars/dynamic-econometrics-conference-2019

Conference

Conference22nd Dynamic Econometrics Conference in Nuffield College Oxford, UK.
Period9/09/1910/09/19
Internet address

Keywords

  • volatility decomposition
  • regime switching
  • volatility dynamics
  • idiosyncratic volatility
  • institutional trading behaviour

Fingerprint

Dive into the research topics of 'The idiosyncratic risk in Chinese stock market'. Together they form a unique fingerprint.

Cite this