A student stated that the first-order autoregressive error models ( 12.1 ) and (12.2) are too simple' for business time series data because the error term in period t in such data is also influenced by random effects that occurred more than one period in the part. Comment.