The RATS Software Forum

This is a replication file for Grier, Henry, Olekalns & Shields(2004), "The asymmetric effects of uncertainty on inflation and output growth," Journal of Applied Econometrics, vol 19, no. 5, 551-565.

This does a VARMA(2,1)-GARCHM mean model with an asymmetric BEKK GARCH on inflation and growth. Because BEKK asymmetry is sign-dependent, this uses a feature added with RATS 8.2 to allow an adjustment to the sign of the asymmetric effect.

Note that the VARMA(2,1) mean model is specific to this data set. Similar data for a different country may require a different model for the mean. The complexity of the mean model causes a few problems with the estimation. What this does is to estimate by the typical combination of simplex + BFGS, but then switches to BHHH to get the standard errors as BFGS doesn't work well in this case. If I were applying this to a different data set, I would start with simpler VAR mean model and check the diagnostics to see if that's enough to get rid of the serial correlation.

Can you show us how to obtain the Generalized Impulse Response Analysis of the paper?

Dear TomDoan,
Can you write code for generalized impluse response function of Grier et al.(2004)?
many rats users want to use it, but we may not have the ability of writing the code.
Thanks very much!

Hi Tom

I am a bit confused as to how to interpret the coefficient on the asymmetry terms in this model, more specifically is it sign dependent? For e.g.- if I get a negative significant coefficient on the asymmetry term in the output growth variance equation D(1,1), does that mean negative innovations to output growth REDUCE output growth variability?

No. The non-standard asymmetry applies to inflation, not output growth.

So, its not sign dependent for output growth but it is for inflation? e.g.- a negative significant coefficient on asymmetry term in the inflation variance equation D(2,2) mean that positive innovations in inflation reduce inflation variance?

Sorry for too many questions. I read the paper to figure out how the coefficients are interpreted but in the paper coefficients on all asymmetry terms are positive.

Thanks a lot.

In the BEKK asymmetry term, only residuals of a certain sign enter the calculation. In most cases, those are negative shocks, which, in most applications would be the ones that are "bad". In this case, it's negative shocks on output and positive shocks on inflation. That has nothing to do with the signs of the coefficients. As with any other BEKK terms, the coefficients aren't identified by the model (multiply the D matrix by -1 and nothing changes).

Dear Tom,

Can you help me with the code to get the impulse-response functions in Kevin B. Grier, Ólan T. Henry, Nilss Olekalns and Kalvinder Shields (2004). The asymmetric effects of uncertainty on inflation and product growth. Journal of Applied Econometrics, vol. 19, No. 5, pp. 551-565 ?

In specific I would like to obtain graphs 4 to 6.

Best regards