My name is Simon and currently I'm workin on a Vector Autoregressive model to simulate interest rates.

The inputs are the parameters from the Nelson-Siegel forumula from 1980 untill now, on a quarterly basis. For more info check: Fixed-income attribution - Wikipedia, the free encyclopedia

So the VAR model looks like this:

Yt=V + A*Y(t-1) + et


V = vector with n variables (the constants)
A= vector with n variables
et= vector with n random numbers

I have estimated the VAR model using Stata.

However because "I" believe that the long-term interest rate will be lower than the average interest rate from my dataset, I want this to also show in my model.

For what I know the averages of the input variables are somehow related to the constants in vector V. Therefore I would like to know if there is a way of incorporating these beliefs I have about lower interest rates in my model. For example by changing the constants.

Maybe my idea that the constant from the model is related to the average of the inputs is totally wrong. In that case I want to ask if there is some other way of incorporating my beliefs about the fuutre interest rates in the model.

Thanks in advance for everyones help.

Kind regards,