How close a relationship does a capital market have with other markets? A reexamination based on the equal variance test

Can Stock Index Future Really Stabilize Stock Market? The Evidence from Chinese Stock Market

“…But the unequal variance test (Lee et al, 2012) assumed that data of cross section are independent. Yang et al (2014) proved that the unequal variance test could reject wrongly the true null hypothesis when the assumption of independent cross section was violated. Then under the assumption of correlated cross section, the paper, following Yang et al (2014) shows a new equal variance test which has better finite sample performance.…”

“…Yang et al (2014) proved that the unequal variance test could reject wrongly the true null hypothesis when the assumption of independent cross section was violated. Then under the assumption of correlated cross section, the paper, following Yang et al (2014) shows a new equal variance test which has better finite sample performance. Assume , t t x y are detrended time series: Before constructing the test and performing statistical inference, we must be precise about the dependence structure of the squared errors (5) and (6) to ensure the validity of a functional central limit theorem.…”

“…(Note 1) However, martingale differences are sequences of a rather special kind and many important dependent processes are not mixing, e.g., infinite moving average (MA(  )) under general conditions and stable first-order linear autoregressive models with Bernoulli-distributed i.i.d. shocks (Yang et al, 2014). The mixing condition is also difficult or impossible to check because establishing conditions where uniform mixing holds is harder, while the best known sufficient condition (e.g., Davidson 1994, Theorem 14.4) requires the shock process to be bounded with probability 1, ruling out normality for example.…”

“…The performance of the equal variance test determines the validity of the empirical results. Therefore, we next examine the equal variance tests proposed by Lee et al (2012) and Yang et al (2014). Consider the following data generating processes: …”

“…This paper investigates whether the Chinese CSI300 index futures can stabilize the Chinese stock market. By using the econometric approach of Bai (2003) and Stock and Watson (2011) to estimate the latent global factor from panel data of international stock markets, this paper controls the economic fundamentals and employs the equal variance test of Yang et al (2014) to compare the volatility of the CSI300 stock index before and after the appearance of CSI300 index futures. We find that CSI300 index futures can help to stabilize the Chinese stock market significantly.…”