12-19-2012 09:02 PM
I have done some time series and panel data estimations in SAS by regressing stock returns on an index I constructed. I do not have any trouble carrying out the estimation, but I need some help with a particular interpretation.
I normally know how to interpret my coefficients, but I need to state how much the returns will change in percentages for one standard deviation increase in my index. Below is some background on the variables:
dependent variable: monthly stock returns on a stock market index in percentages (i.e. 6.23 is actually referring to 6.23%).
independent variable: an index on a particular activity.
My question sounds like standardized coefficients issue, but in these cases, the interpretation is something like the following: one standard deviation increase in this index leads to 0.21 standard deviation increase in the returns. This is not really what I need. What I am looking for something like the following: one standard deviation increase in the index leads to 0.47 increase in the returns.
Is there a way to tweak the regression procedure in SAS to obtain the interpretation I need? Or is there a particular way to tweak the coefficients?
I would really appreciate your help with this issue.
Thanks in advance for your time and help.