Date:
Sun, 17/05/202016:00-17:45
Location:
online
Nathan Goldstein
Bar-Ilan University.
Expectations Formation and Forward Information: What Do Forecasters Really Know about the Future?
with Yuriy Gorodnichenko
Abstract: Subjective adjustment is a well-known practice in forecasting. Yet, it is ignored by models of expectations formation. We propose to model forecast adjustment as a practice driven by availability of noisy signals about the future. Introducing forward information cannot only explain apparent deviations from rational expectations documented in the literature, but additionally resolve puzzling empirical patterns, which we document in the Survey of Professional Forecasters. We provide further direct evidence that SPF forecasts are driven by noise in forward information, rather than noise in realized data, as suggested by standard models of noisy information. Our approach is also useful for studying the role of news shocks. We propose a simple method for extracting the forecast adjustment due to forward information from data of survey forecasts and find a highly predictive power of future movements in fundamentals by this component.
The online meeting will be held on Sunday, May 17, 2020, 16:15-17:45
Join Zoom Meeting
https://huji.zoom.us/j/96473227980?pwd=SEk5aG1wUE10TUxWcWwrM05mTkx3Zz09
Bar-Ilan University.
Expectations Formation and Forward Information: What Do Forecasters Really Know about the Future?
with Yuriy Gorodnichenko
Abstract: Subjective adjustment is a well-known practice in forecasting. Yet, it is ignored by models of expectations formation. We propose to model forecast adjustment as a practice driven by availability of noisy signals about the future. Introducing forward information cannot only explain apparent deviations from rational expectations documented in the literature, but additionally resolve puzzling empirical patterns, which we document in the Survey of Professional Forecasters. We provide further direct evidence that SPF forecasts are driven by noise in forward information, rather than noise in realized data, as suggested by standard models of noisy information. Our approach is also useful for studying the role of news shocks. We propose a simple method for extracting the forecast adjustment due to forward information from data of survey forecasts and find a highly predictive power of future movements in fundamentals by this component.
The online meeting will be held on Sunday, May 17, 2020, 16:15-17:45
Join Zoom Meeting
https://huji.zoom.us/j/96473227980?pwd=SEk5aG1wUE10TUxWcWwrM05mTkx3Zz09