Volatility And Dependence Analysis Of BRICS Countries' Stock Markets Using Semi-parametric CARR Models

The BRIC,first proposed by Jim O Neill of Goldman Sachs in 2001,is the acronym of the four member countries—Brazil,Russia,India and China.In 2011,South Africa became a member of the organization,makes the name of the organization to be BRICS.So far,BRICS has become an important force to push the development of the global economy.With the development of financial globalization,the relationships between the members of BRICS become increasingly closed.It is important to research the volatility and dependence of BRICS countries' stock markets.By using the weekly range data of IBOVESPA index of Brazil,RTS index of Russia,Bombay SENSEX 30 of India,Shanghai SSEC index of China and SBK index of South Africa,we establish the parametric and semi-parametric CARR models to study volatility in these countries' stock markets.Then based on semi-parametric CARR model,we establish static state Copulas and time-varying Copulas models to study the relationship and contagion effect between Chinese stock market and others.Finally,we establish Vine models to study on the dependence between the stock markets of the five BRICS countries.The key points and main achievements of this paper are listed as follows:1.This paper adopts a semi-parametric CARR model estimated by local linear least squares method to describe the volatility in the stock markets of the five BRICS countries.The results show that semi-parametric CARR has a higher fitting ability than parametric CARR model.2.Here we select semi-parametric CARR models to replace GARCH models to describe the marginal distribution when establish Copulas models.As for the relationship between China and others,we use four kinds of static state and time-varying Copulas models to study them.3.This paper uses a semi-parametric CARR model to describe the marginal distributions of all the weekly ranges time series in the stock markets of the five BRICS countries when establish Vine models.Also we provide 14 kinds of copulas to select as the pair-copulas.Along with C-Vine Copulas and D-Vine Copulas to describe the dependence between the five stock markets.