Ai is a KK coefficient matrix and ut is a
K-dimensional white noise process. To obtain stable
results, the underlying time series has to be stationary.
In the following subsection, the underlying
VAR(p) model is exploited to test Granger causality
relationships and orthogonalized impulse response
functions. In the second part of the study, trading
strategy based on the estimated VAR(p) framework is
employed to economically exploit the relationship
between the markets.