A carregar...
Publicação
Nowcasting earnings: a comparison with traditional forecasts
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.23 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
This thesis employs dynamic factor models to nowcast YoY percentage change of quarterly firm level earnings using time series observed at different frequencies. To this aim, we use
macroeconomic, financial, and textual data. We focus on evaluating the predictions of the dynamic
factor model against conventional econometric models, such as ARIMA and VAR, and analysts’
estimates. Our findings show that the nowcasting model performs comparably to these traditional
methods, with its predictive accuracy improving as the quarter progresses and releases are
incorporated. Therefore, we believe that future research could improve nowcasting accuracy and
extend the applicability of these techniques.
Descrição
Palavras-chave
Macroeconomic data Financial data Dynamic factor models Forecasting models Textual data