Analyze and Forecast Stock Market Volatility

Research on conditional volatility of asset prices
has been a topic of expanding interest in the field
of finance. It is known that volatility is
inherently unobservable, thus the selection of
models and how to define them is crucial for
financial research.
This book attempts to analyze and forecast stock
market volatility by both parametric and non-
parametric approaches. Augmented GARCH models with
an investor sentiment effect derived from trading
volume are compared with conventional GARCH models.
Furthermore, A Monte Carlo experiment is adopted to
generate stock-return data and a neural network
approach is applied to forecast Value-at-Risk of the
stock market. Results suggest that accuracy of GARCH
models is improved by accounting for the volume
effect and non-parametric neural network technique
can be a good alternative to forecasting stock
market volatility.

Autorentext

Qianru LiPh.D. Utah State University, USA, 2008

Klappentext

Research on conditional volatility of asset prices has been a topic of expanding interest in the field of finance. It is known that volatility is inherently unobservable, thus the selection of models and how to define them is crucial for financial research. This book attempts to analyze and forecast stock market volatility by both parametric and non-parametric approaches. Augmented GARCH models with an investor sentiment effect derived from trading volume are compared with conventional GARCH models. Furthermore, A Monte Carlo experiment is adopted to generate stock-return data and a neural network approach is applied to forecast Value-at-Risk of the stock market. Results suggest that accuracy of GARCH models is improved by accounting for the volume effect and non-parametric neural network technique can be a good alternative to forecasting stock market volatility.