In recent years, the term “nowcasting” has entered the economic and financial jargon to describe the real-time development of macroeconomic and financial conditions. These nowcasting models can leverage information from potentially noisy, higher-frequency data releases to estimate current values of infrequently published indicators.
In this whitepaper, we introduce a macroeconomic nowcasting model based on RavenPack EDGE to predict real-time economic activity of four important emerging market economies: Brazil, Russia, India and South Africa (BRIS).
● Our study finds that RavenPack news analytics help improve timeliness and accuracy of traditional macroeconomic nowcasting models. Specifically, we find that:
● Out-of-sample forecasting error improves by 5% to 25% depending on the country and the prediction period. Prediction accuracy is significantly enhanced during periods with a sparsity of core macroeconomic releases.
● Aggregating news analytics over relatively short time horizons (30 and 90 days) provides an optimal assessment of the prevailing macroeconomic situation in a timely fashion.
Nowcasting models are powerful tools for mitigating uncertainty when formulating monetary policy or investment decisions.
Get the whitepaper and find out how RavenPack EDGE can help to improve your nowcasting models.
Request White Paper