# Vector error correction model

I am working on finding the relationship between equity, gold, crude oil, and currency value. I used the Johanson cointegration test. Trace test indicates 4 cointegrating equations at the 0.05 level. For using VECM the rank for cointegration in Eviews has a maximum of 3. What is the problem in this model? How do you do the analysis when both number of variables and cointegration are same?

Suppose you have two variables. With one cointegrating relationship, it describes the long-run relationship between the two variables. Something is still left to vary in the short-run. That would not be the case for two cointegrating relationships. For the same reason, you can only have 3 cointegrating relationships with four variables. The Spatial Econometrics toolbox for Matlab will only test $r<=m-1$ where $r$ is the number of cointegrating relationships and $m$ is the number of variables. I imagine your package is the same.