0
$\begingroup$

Can someone walk me through how to manually calculate the IRF values that Eviews outputs for the effect of a nonfactorized one unit innovation.

I have my VAR equation and coefficients:

VAR Model - Substituted Coefficients:
===============================
MARKET_A = 1.38531644845*MARKET_A(-1) - 0.456533910287*MARKET_A(-2) + 0.0908884378218*MARKET_B(-1) - 0.0116001711851*MARKET_B(-2) - 0.0458678142475

MARKET_B = 0.0934164283416*MARKET_A(-1) - 0.0678660464848*MARKET_A(-2) + 1.34042022538*MARKET_B(-1) - 0.372260999882*MARKET_B(-2) + 0.0341759469926

The Eviews output for the effect of a nonfactorized one unit innovation from MARKET_A.

 Period MARKET_B    MARKET_A

 1   0.000000    1.000000
 2   0.093416    1.385316
 3   0.186763    1.471058
 4   0.258970    1.421328
 5   0.310545    1.318772
 6   0.346591    1.203253
 7   0.371878    1.092721
 8   0.389868    0.994217
 9   0.402869    0.909562
 10  0.412376    0.838231

Nonfactorized One Unit      

To calculate period 2, I have tried plugging in the values in the equation i.e

  • 0 for MARKET_B(-1),
  • 0 for MARKET_B(-2)
  • 1 for MARKET_A(-1)
  • 0 for MARKET_A(-2)

but the output is not the same for the first period. Obviously, I'm missing something. Could someone point me in the right direction?

$\endgroup$
0
$\begingroup$

I ended up figuring it out. I was including the estimated constant in my IRF equation/calculation in Excel, which shouldn't have been there and was throwing my calculation off slightly. This obviously compounded with each subsequent period.

I found this resource provided the eureka moment for me: https://www3.nd.edu/~esims1/arp_companion.pdf

| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.