@Rick_SAS has the answer!
https://blogs.sas.com/content/iml/2014/07/23/prediction-ellipses-from-covariance.html
Why do you need this prediction formula, what would you do with it if you had it? If you want to determine if additional points (not the ones used to create the ellipse) are within the ellipse, you can just add these additional points to the plot. Example:
data additional;
height1=60; weight1=80; output;
height1=70; weight1=95; output;
height1=80; weight1=50; output;
run;
data class;
set sashelp.class additional;
run;
proc sgplot data=class;
scatter x=height y=weight / markerattrs=(symbol=circlefilled);
ellipse x=height y=weight / alpha=0.05 lineattrs=(thickness=2 color=red) type=predicted;
scatter x=height1 y=weight1 / markerattrs=(symbol=plus);
run;
... View more