0
$\begingroup$

I'm in the midst of preparing data to test the relationship between stock index and macroeconomic variables. The variables include m1, cpi, ipi, interest rate, exchange rate. I have some question below:

  1. Is seasonality adjustment on the data required? If one variable is seasonality adjusted, is that all variables in the models have to be seasonality adjusted as well? If all are not seasonality adjusted, will there be a problem?

  2. Would it be alright if all data are not inflation adjusted, that is nominal data? And if one variable are inflation adjusted, is that all must be so to be consistent?

  3. The base period for CPI and IPI is not the same. Is it a requirement for the base period to be the same in the model ?

$\endgroup$
  • $\begingroup$ Would appreciate it very much if someone can give your comments .. : ( $\endgroup$ – Josephine PM Dec 25 '17 at 10:11
1
$\begingroup$
  1. Is seasonality adjustment on the data required? If one variable is seasonality adjusted, is that all variables in the models have to be seasonality adjusted as well? If all are not seasonality adjusted, will there be a problem?

Seasonal adjustment of data is required in your VAR/VECM, you can do this by employing seasonal dummies (dummies 11 if you have monthly data and 4 if you have quaterly data).

All your data has to initially be seasonally unadjusted, because the inclusion of dummies will affect all the data that you include in your model. You will end up having skewed estimates of your already adjusted variables influence on a given process.

  1. Would it be alright if all data are not inflation adjusted, that is nominal data? And if one variable are inflation adjusted, is that all must be so to be consistent?

Again if you want your estimates to be consistent you have to make sure all of your monetary variables are inflation adjusted. practically you may want to use real exchange rate instead of nominal exchange rate data for this.

  1. The base period for CPI and IPI is not the same. Is it a requirement for the base period to be the same in the model ?

No this is not necessary and your estimates will be fine. Just remember to keep this in mind when your run your model.

Hope this helps

$\endgroup$
  • $\begingroup$ Thanks for your comment! For the seasonal adjustment, can i just adjust the seasonally unadjusted data using the X-13 available in Eviews and do nothing for those which is already adjusted, Regarding the exchange rate, I currently have the nominal exchange rate data. To convert it to real exchange rate, is that i can just divide it by the CPI of the respective year and multiply it with the CPI of the base year? Alternatively, I have a set of Real Effective Exchange Rate Index (CPI Based). Which one is better to be used? $\endgroup$ – Josephine PM Dec 28 '17 at 3:41
  • $\begingroup$ Seasonally adjusted or unadjusted CPI should be used to adjust the exchange rate? $\endgroup$ – Josephine PM Dec 28 '17 at 3:50
  • $\begingroup$ @JosephinePM why dont you use the formula for real exchange rate? ie subjectmoney.com/… $\endgroup$ – EconJohn Dec 29 '17 at 2:51

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.