Vieses comportamentais em projeções macroeconômicas^*

Abstract

This paper documents the optimism in the forecast of GDP and IPCA. That is, it shows that GDP forecasts are systematically higher than what actually happens, and IPCA projections are systematically lower. Then it tests whether there is an anti-herding bias related to a publicity hypothesis regarding institutional incentives for the behavior of professionals who make such forecasts. Finally, it verifies whether the revision of the forecast still carries the previously identified bias. The results confirm the existence of an optimistic bias in the GDP and IPCA projections and this bias remains in the following revisions for GDP. However, the publicity hypothesis seems to be not confirmed for the aggregated forecasts by types of institutions.

Keywords:
Biases; Forecast; GDP; Inflation; Boletim Focus