On aggregation bias in fixed-event forecast efficiency tests
This note shows that problems due to aggregation in fixed-event forecast efficiency tests are not as severe as they are in unbiasedness tests. We also show that first lags of consensus revisions should be avoided in the tests.