Working Paper: CEPR ID: DP15451
Authors: Gabriel Prezquirs; Elena Diaz
Abstract: This paper develops a novel indicator of global economic activity, the GEA Tracker,which is based on commodity prices selected recursively through a genetic algorithm. TheGEA Tracker allows for daily real-time knowledge of international business conditionsusing a minimum amount of information. We find that the GEA Tracker outperformsits competitors in forecasting stock returns, especially in emerging markets, and inpredicting standard indicators of international business conditions. We show that aninvestor would have inexorably profited from using the forecasts provided by the GEATracker to weight his/her portfolio. Finally, the GEA Tracker allows us to present thedaily evolution of global economic activity during the COVID-19 pandemic.
Keywords: Global economic activity; Commodity prices; Factor models; Genetic algorithm
JEL Codes: F44; G17; Q02
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| GEA tracker (D58) | improved forecasting accuracy (C53) |
| selected commodity prices (Q02) | GEA tracker (D58) |
| GEA tracker (D58) | investor profitability (G11) |
| selected commodity prices (Q02) | global economic conditions (F69) |
| GEA tracker (D58) | stock market predictability (G17) |