I have a problem with dataset, that is the fact that in some time series data start to appear in a certain year but I need the value of those time series for the precedent year. For example, in the TFP time series for Euro Area datas starts to appear only in 1995, but since I have to calculate the TFP growth, I need data for 1994 for certain countries. So, my question is: how trying to estimate those missing values? I'm thinking to use a four-period backward looking moving average (i.e.: the average of the first four values that I have). Do you think it is correct? Are there other methods to do that? Thank you in advance.

  • $\begingroup$ There is no 'correct' way. Every method has its limitations. Why must you infer the 1994-1995 growth rate btw? Usually if you difference your (log) data you make peace with the fact that it will cost you a degree of freedom/datapoint. $\endgroup$ Sep 3 at 12:08
  • $\begingroup$ Thank you. I have to infer the 1994-1995 growth rate because my analysis starts from 1995, and so to have the inflation rate for 1995 I have to calculate the variation between the 1994 price level and the 1995 price level. $\endgroup$
    – Giord
    Sep 4 at 11:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.