This paper has examined the dynamics of price discovery between different Taiwan stock index and derivative products
(the ETF50, the ETF51, and ETF52) by applying the vector autoregressive (VAR) model. It is easy to show that
four series variables are in positive correlation relationships. The finding suggests that ETF51 has the largest return and
volatility. Further, it explores at least three number of co-integrating vector among the variables. The stock index and
derivative products share co-integration relationship so that they will not wander arbitrarily far from each other. On the
other hand, it has demonstrated the methodology of Granger causality to examine the causality linkage among the variables.
The leading relation exists in stock index with stronger evidence that stock index leads derivative products.
Moreover, according to the decomposition of forecast error variance, stock index is the least influenced by outside
force among the four series variables. In other words, stock index is mostly influenced by its own shock, but is less by
other variables shock. The stock index variance decomposition can explain more except its own influence. Secondly,
they cannot trace consistently out the time path of impulse response. Consequently, investors manage trading strategies
in information spillover