var with exog. feedback

Aug 22, 2014 at 5:36 AM
Hi, please do you know to help with any of these questions?

1 Testing VAR stability when there is feedback from an exogenous variable which evolves acc. equation that is added to the system as identity

2 Historical decomposition (again VAR with exog. f) :
2-1 Contribution of some variable x to the HD of a variable y is shock scenario minus baseline scenario, where the shocks are the estimated structural shocks to x?
2-2 What should contributions of all variables sum up to?
2-3 Sources mention HD would be sensitive to initial conditions etc., suggesting there is no one ultimate HD. How to recognise a good one?

Thank you.