Martin Randall comments on Evaluating the historical value misspecification argument