03-11-2015 01:51 PM
What procedure are you using to calculate partial autocorrelation? PROC TIMESERIES offers the CORR option which in turn offers the PACFSTD that will give partial autocorrelation standard errors. This seems like the only way I could find to get close on this. If you post in the Forecasting and Econometrics forum, there are folks there that are much more knowledgeable about capabilities.