Dear All
I am currently conducting research where I estimate real quarterly GDP with high frequency variables using an AR-MIDAS regression model in EViews. In the attachment, I have results for two separate AR-MIDAS models. In the first one, I have results of the AR-MIDAS where real quarterly GDP is regressed on monthly global oil prices. The second set of results show real quarterly GDP regressed on daily globla oil prices. They each give different results. Why would this be the case. In addition, I have had some challenge in interpreting the results from the MIDAS model where the high frequency lags (say, 8) show a positive relationship for the first four lags but a negative for the last four, therefore a declining shape of the polynomial. How would I interpret such a result?
Dear Ajit Kumar Roy
Thank you so much for the response. I appreciate you pointing me in the direction of the Adjusted R-sq. I will certainly take this into consideration. What I still do not understand is how to interpret the coefficients of the high frequency lags if throughout the lag length I have an equal mix of negative and positive coefficients. Any help in this respect will be greatly appreciated. Thank you.
- Badges
- Science topic
- Similar topics
- Engineering
- Petroleum Engineering
- Oil