Some comments on the shape of the gain–stocking rate curve