Zaslat SMS: On not evaluating economic models by forecast outcomes