Data for: A multivariate analysis of forecast disagreement: Confronting models of disagreement with survey data
datasetposted on 18.07.2019 by Jonas Dovern
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Abstract of associated article: This paper documents multivariate forecast disagreement among professional forecasters and discusses implications for models of heterogeneous expectation formation. Disagreement varies over time and is positively correlated with general (economic) uncertainty. The degree to which individual forecasters disagree with the average forecast tends to persist over time. Models of heterogeneous expectation formation can be modified by introducing heterogeneous signal-to-noise ratios to match this feature. Furthermore, disagreement about correlations of different macroeconomic variables is high on average. In general, multivariate forecast data can be used more effectively than it has been to estimate models with heterogeneous expectations and to test the mechanisms used to generate disagreement in these models.