understanding model equation output of VEC model

[khalid]

Hi all,

I am trying to use VEC model on 2 time series. But I am not able to interpret the vec model equations. What does d() stand for here.
The equation looks like as follows:

d(A) = 0.1123522698*( B(-1) - 0.09145065747*A(-1) + 272.7504444 ) - 0.5778918819*B(-1)) + 0.8630636894*B(-2)) + 0.01105613981*d(B(-1)) - 0.06809563181*d(A(-2)) - 0.5700930685


d(B) = 0.005507195627*( B(-1) - 0.09145065747*A(-1) + 272.7504444 ) - 0.2300183269*d(B(-1)) + 0.07976447043*d(B(-2)) + 0.01779323877*d(A(-1)) - 0.005012549952*d(A(-2)) - 0.05370207617


Regards

[trubador]

d() is the difference operator and takes the first difference of the series in your case (i.e. d(A)=A-A(-1)). It seems, you have estimated a two variable model with 2 lags and no exogenous variables. Each equation has two different parts, associated with long term and short term dynamics, respectively. The first term (B(-1)-0.09*A(-1) + 272.75) is called the cointegration (error correction) relationship and is same for both equations, since it is the equilibrium. The coefficients of this term measure the speed of adjustment for each variable towards the long term equilibrium in case of a deviation. Given this restriction, variables are allowed to display short term adjustment dynamics via the second term.

I strongly recommend you to dip into an econometrics book on the subject before going any further, since error correction models are more complicated than they seem...