observation equation

[sujee2sujee]

Hi,

I have a log-linearized model to estimate. I have a problem with observing inflation. The data is the deviation of inflation from the target ( which is 5% on annual basis adjusted for quarter). But the mean of the data is higher ( which is 9% on annual basis adjusted for quarter). I observed data is not demeaned. There is no explicit inflation target in practice for the selected country. But, some studies on this country suggest to consider 5% as the implicit target. I have mentioned the observation equation as follows.

pi_obs=pi_hat;

The target of 5% is given as a parameter pi_bar. I think there is something going wrong.

What happens to the mean in the observed data. Should I adjusted the observation equation? or demean the deviation of inflation from target data?

Please respond. The model didn't generate any error message and the model also converges in the current set up. Please let me know whether I am missing something.

Thank you very much for any suggestion.

Sujee

[jpfeifer]

If both your model variable and the data are deviations from the 5 percent target, you should be fine. What happens if the mean is positive is you force shocks to account for the on average above target inflation. You will only get trouble if the reason for the higher inflation is a different target of the central bank, because then forcing shocks to account for this is wrong.

[sujee2sujee]

Thank you for the thought. The policy documents of the Central bank also says that they will be targeting to 5% when the country shifts to an explicit inflation target framework. Then I think I am doing is correct.

Sujee