statistic is substantially less than 2, there is evidence of positive serial correlation. If Durbin–Watson is less than 1.0, there may be values of d indicate successive error terms are, on average, close in value to one another, or positively correlated. If d > 2 successive error terms are, on average, much different in value to one another.
7.Cointegrating Regression
If these variables under the study are cointegrated. We only focus on the classical analysis of I (1) and I (0) systems and estimate the Vector Autoregressive (VAR) model as the equations 2.19 and 2.20.
∆y = {3 ∆y + {3 ∆x + v∆y
∆xt = {321∆yt-1 + {322∆xt-1 + v∆X
As we can see that lag length in these equations are t and t- 1. But in time series data of limited length, this assumption of errors is violated if a relationship between and is insignificant; that is if lag length is 0. Vector Autoregression (VAR) model would not fit the estimation anymore. We should estimate the cointegration regression to study the relationship in this case.
Engle and Granger (1987) note that a linear combination of two or more I (1) series may be stationary, or I (0), in which case we say the series are cointegrated. Such a linear combination defines a cointegrating equation with cointegrating vector of weights characterizing the long-run relationship between the
variables. Consider the n+ 1 dimensional time series process, with cointegrating equation:′1
yt = xt {3+ D1t y1 + u1
where