File(s) under permanent embargo
Reason: Restricted by author. A copy can be supplied under Section 51(2) of the Australian Copyright Act 1968 by submitting a document delivery request through your library or by emailing email@example.com
Aspects of volatility in the Chinese stock market
thesisposted on 21.02.2017, 01:37 by Chi, Wei
This thesis analyses three sets of issues: 1) the cyclical behaviour of the Chinese stock markets, 2) the fitness of using realized volatility (RV) in the generalized autoregressive conditional heteroskedasticity (GARCH) model, and 3) the volatility spillover between the Chinese and Australian stock markets. After conducting an extensive literature review, the thesis examines the three sets of issues seperately. First, a Markov regime switching model is applied to analyse the bull and bear cycles in the Chinese stock market, since the cycles of bull and bear markets can reflect economic development and investor confidence. Specifically, grouping stocks by industry and firm size, the results show the following: 1) Bear cycles between stocks and the index overlap heavily, indicating strong herding effects. A long bear market cycle is found and can be explained by widely diversified stock performance across the markets. 2) Certain shocks to one industry could have different impacts on the Shanghai and Shenzhen stock markets. 3) Firm size can have a significant impact on the performance of stocks in bull or bear cycles. The second topic focuses on estimating the RV of the Chinese stock markets and comparing it with the GARCH model. The actual volatility is inherently unobserved, while the RV could be treated as being directly observable and then be used to study time-varying behaviour and forecasting. Thus, a large number of studies use RV in GARCH models for volatility analysis. However, there is yet no study that discusses the correlation between RV and GARCH while using RV in GARCH models. This could lead to bias in estimation because of the different properties of RV and GARCH. The results show that GARCH models combined with RV could be more suitable for estimating volatility for large firms. When the firms are grouped in terms of positive/negative returns, similar results are found as when firms are grouped by firm size. The third topic estimates the volatility spillover between the Chinese and Australian stock markets, motivated by the lack of attention to spillover between these two markets in the literature. While economic interdependence between Australia and China has soared during the last two decades due to China’s tight reliance on Australia’s mining and resources, little research attention has been paid to these two countries. This study fills the literature gap and assesses the volatility spillover between the Chinese and Australian stock markets based on the CSI300 and ASX200 industry indices. To the best of my knowledge, this is the first study using Chinese industry data to discuss volatility spillover. The key findings of the thesis are that volatility spillover across these two markets is bidirectional, while there is one-sided or insignificant spillover across industries between these two countries. The findings of the thesis fill the literature gap, help clarify the debate about volatility spillover between the Chinese stock market and the world market, and provide a clearer idea of the channels through which volatility is transmitted across countries.