[an error occurred while processing this directive]

Macro loops make things possible, they don't make them fast

Instead, use data steps. One thing that macros seem good for is running rolling regressions, or computing rolling means or standard deviations. A modification of the method used in section 7.3 should be used instead.