Working Paper: CEPR ID: DP18496
Authors: Kenza Benhima; Elio Bolliger
Abstract: Using individual inflation and GDP growth forecasts by professional forecasters for a panel of emerging and advanced economies, we provide direct evidence that foreign forecasters update their forecasts less frequently than local forecasters (about 10% less frequently) and make larger errors in absolute value (up to 9% larger). The foreign forecasters' less accurate forecasts are not due to a more irrational expectation, but to less precise information. The asymmetry is stronger at shorter horizons and when forecasting inflation. In general, the asymmetry is not stronger when forecasting is more uncertain. Taken together, our results provide a basis for disciplining international finance and trade models with heterogeneous information. On the methodological side, we provide tests that identify differences in information frictions across groups.
Keywords: information asymmetries; expectation formation
JEL Codes: E3; E7; D82
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Foreign forecasters (F37) | Lower frequency of updates (C41) |
| Foreign forecasters (F37) | Larger forecast errors (C53) |
| Information asymmetries (D82) | Larger forecast errors (C53) |
| Local forecasters (Q47) | More precise private information (C81) |
| Larger forecast errors (C53) | Asymmetry stronger for inflation (E31) |
| Larger forecast errors (C53) | Asymmetry stronger for shorter forecasting horizons (C22) |