Hi,
I have to calculated predicted PD for year 2015, 2016 and 2017 for Bucket1, Bucket2, Bucket3 and Bucket4 using a Z_score and an Index value. This Index value is a random value. The index value is to be generated such that, when we calculate predicted PD for each year then the sum square error of Predicted Pd and Actual PD shoud be minimum.
I am able to do this in Excel using Run solver, and proc optmodel is an equivalent for this. But I am not able to use proc opt model in SAS EG.
Attached is the Input table and desired output table in excel file attached:
Table 1: Need to use Z Score from this
Table 2: Bucket-wise Actual PD for each year
Table 3: Highlighted yellow values are index values generated, and used in calculating predicted values
Table 4: Min sum of squares between actual and predicted.
Please help with this, as I am in urgent need to perform this calculation in SAS.