Hello,

I want to construct an SVAR model including Gasoline shock --> Headline Inflation --> Core Inflation --> Inflation Expectations --> Interest Rate (Matrix 5x5).

I have a problem in programming this in R. See attached my model and restrictions. In my model all elements are asterisks, thus denote unrestricted elements except for the upper row (4 elements are 0) and the value of the lowest row/first column indicating no direct influence of gasoline to interest rate. In my model Gasoline prices influence all inflation indicators.

Now I do not understand how to model this in R. I set all the unrestricted elements to "NA", thus to be unrestricted; but I obviously made the mistake that the diagonal has to be 1. So I corrected it and got the next error.

As I know from Lütkepohl, a A-matrix can have n(n-1)/2 restrictions, thus did I enter too many restrictions?

As Lutz Kilian says: "The use of a partially identified model not only dispenses with the need for potentially controversial additional identifying assumptions, but also facilitates the construction of scenarios for the identified shock. [The model I want to construct] is an extension of a lower-dimensional model examined and validated in Kilian and Zhou (2021) who show that in their setting very similar results are obtained under a range of alternative, more restrictive identifying assumptions."

So do anyone know how to code this in R or any other programming language appropriately?

Thank you in advance and best regards,

Philipp