VECM Model
Posted: Tue Apr 13, 2010 10:07 am
Hello,
I have estimated a VECM model with three endogenous variables (log oil price, log gas price and log electricity price) and with the following exogenous variables: two dummies for 2006 and 2008, heating degree days (hdd) and cooling degree days. I got two cointegrating equations (from the Johansen procedure). However, I was confronted with two problems:
1) VEC specification imposes 1 unit root (implying that my model is not stable)
2)presence of heteroskdasticity (White heteroskedasticity test).
There is some way to solve these problems?
Thank you very much.
I have estimated a VECM model with three endogenous variables (log oil price, log gas price and log electricity price) and with the following exogenous variables: two dummies for 2006 and 2008, heating degree days (hdd) and cooling degree days. I got two cointegrating equations (from the Johansen procedure). However, I was confronted with two problems:
1) VEC specification imposes 1 unit root (implying that my model is not stable)
2)presence of heteroskdasticity (White heteroskedasticity test).
There is some way to solve these problems?
Thank you very much.