Page 1 of 1

Gaps in time series with N/a values-Macro Announcements/

Posted: Wed May 23, 2018 8:29 am
by Jakem
Dear Eviews forum,

I want to research the effect of Chinese macro announcements (using the surprise effect) on the volatility of commodity prices (using an index).
These macro announcements are published quarterly/monthly. Furthermore, I have calculated the daily volatility of the index price.

Vol Price index: (12 announcements)
(dependent variable: independent variable)

There are gaps in the data series (days where there are no announcements). I have removed 'no activity' days and still get the error 'insufficient number of observations'.The number of observations per announcement is on average 80 (largest 99, smallest 33).

Is there an explanation for this error? How do gaps in data series get solved without estimating values?
I have attached a screenshot of my excel values to demonstrate it!

Thank you for your help in advance! :)

Greetings,
Jake

Re: Gaps in time series with N/a values-Macro Announcements/

Posted: Wed May 23, 2018 9:07 am
by EViews Gareth
When running a regression you need data for every variable in the regression for an observation to be included. i.e. only rows with zero NAs will be used.