Lecture 9 Structural VAR models

Question 1

Sims recursive ordering

Answer 1

Order it to e1t=….

e2t=….

Question 2

Sims Bernanke

Answer 2

Seems to be a matrix summary of sims recursive?

Question 3

Sims structual VAR

Answer 3

You estimate the whole system and then do restrictions on the residual. Simple Choles. decomp. result depends on the ordering. Has to be argued by the author why something stand about the other.

Question 4

SVAR or VAR?

Answer 4

VAR: often difficult to draw any conclusion from the large number of coefficient estimates in the VAR system. vector and autoregressions have the status of “reduced form” and so they are merely vehicles to summarize the dynamic properties of the data. Parameters dont have an economic meaning and are subject to the Lucas critique.

SVAR: Dont contain fixed coefficient expectational rules.
(Give linear approx. to the behaviour of the private sector and monetary authorities). Includes dynamics arising from revision in forecasting rules as well as other sources of dynamics.

Question 5

Parameter restriction:

Answer 5

Cholesky decomp: gives an exactly identified system, implies a recursive structure among the variables of the system.
SVAR: allows us to use a non recursive structure , we identify the model by imposing short run restrictions on B0 or long run restrictions on B1…. at least n(n+1)/2 restrictions on B0.

Question 6

Determining the set of restrictions on B0:

2 approaches

Answer 6

i) explicit macroeconomic model
ii) Choosing restrictions based on the structure of the economy, which arent rejected by data, which produce the results consistent with econ theory.

Question 7

What means overidentified?

Answer 7

More restrictions than necessary, perfectly identified (lower triangle matrix)