0

I am doing a multivariate time series analysis. However, one of my time series is the Productivity Index. The time series consist of quarterly data from 1971 to 2015. The index is such that 2010 = 1, meaming that the average value of all four quarters in 2010 is equal to 1. The remaining values for every quarter are expressed according to the index value.

enter image description here

Can I proceed with the time series?

What implications do the index values have on my results?

Richard Hardy
  • 67,272
user1607
  • 909
  • Why did you include the [tag:prediction], [tag:garch] and [tag:cointegratoin] tags? Your question does not seem to indicate they would be relevant. – Richard Hardy Dec 10 '17 at 20:12
  • because I found that these two time series are cointegrated, there is at most one cointegrating equation. I am planing to use use these time series for prediction. I did a univariate prediction on GDP using the GARCH ARMA model. I am sorry if i did something wrong, i am still figuring out how it works here... – user1607 Dec 10 '17 at 20:17
  • You might want to see https://stats.stackexchange.com/questions/315206/cross-correlation-of-two-non-stationary-time-series/315350#315350 – IrishStat Dec 10 '17 at 20:18
  • @Dana, no worries, I just wanted to find out why you included the tags and whether they were relevant. I think the connection with GARCH and cointegration is perhaps too weak to include these tags, but you can edit your post to say you are going to forecast your series, then the [tag:forecasting] tag is fine. – Richard Hardy Dec 10 '17 at 20:35

0 Answers0