Initial Values in Shock_decomposition

Hi,
I found a large share of initial values in shock decomposition. Is this normal? Or does it imply some problems?

You can clearly see from the model that your data shows a strong trend. This suggest your observation equation is wrong.