The issue seems to be due to holidays and other days when the market was closed. Data is in 5-day week format.
I tried to work around it by adding an IF condition to the sample statement, doing the following:
Code: Select all
smpl @all IF bond_30_otr_yld<>NA
series bond_30_otr_ch_yld=bond_30_otr_yld-bond_30_otr_yld(-1)
Unfortunately, it fills in NA values where there are NAs in the series (day of NA and following day), even when I use the IF condition in the sample statement. Is there an easy work-around? Different differencing command or a function for deleting dates from workfile based on NAs in a series?