QUESTION ON THE DATASET OF THE "RISK SHOCKS" (aer2014) paper

Dear All,

I am trying to obtain the smoothed shocks from the AER paper by Christiano, Motto and Rostagno, Risk Shocks. In their paper, it is claimed that the dataset is from 1985q1 to 2010q2, which means there are 102 observations. However, there are 118 observations in their attached files on the AER’s webpage. It is puzzling for me. I wonder if anyone knows the starting and ending time of those 118 observations? Thank you in advance!

They use presample = 16, i.e. the first 16 observations are used to initialize the Kalman filter, but they do not enter the likelihood computation.

Thanks, Johannes! Just to confirm: the presample they used was the 16 quarters before 1985q1? Is that right?

Yes, that is correct

Failed to say thank you on time. But it is better than nothing. Thank you very much!