Forecasting by factors, by variables, or both?
We consider forecasting with factors, variables and both, modeling in-sample using Autometrics so all principal components and variables can be included jointly, while tackling multiple breaks by impulse-indicator saturation. A forecast-error taxonomy for factor models highlights the impacts of loc...
| Main Authors: | , |
|---|---|
| Format: | Working paper |
| Published: |
University of Oxford
2012
|
| _version_ | 1797052941117423616 |
|---|---|
| author | Castle, J Hendry, D |
| author_facet | Castle, J Hendry, D |
| author_sort | Castle, J |
| collection | OXFORD |
| description | We consider forecasting with factors, variables and both, modeling in-sample using Autometrics so all principal components and variables can be included jointly, while tackling multiple breaks by impulse-indicator saturation. A forecast-error taxonomy for factor models highlights the impacts of location shifts on forecast-error biases. Forecasting US GDP over 1-, 4- and 8-step horizons using the dataset from Stock and Watson (2009) updated to 2011:2 shows factor models are more useful for nowcasting or short-term forecasting, but their relative performance declines as the forecast horizon increases. Forecasts for GDP levels highlight the need for robust strategies such as intercept corrections or differencing when location shifts occur, as in the recent financial crisis. |
| first_indexed | 2024-03-06T18:37:36Z |
| format | Working paper |
| id | oxford-uuid:0bc9e862-0b1c-432c-a8a5-c494c6bb96d0 |
| institution | University of Oxford |
| last_indexed | 2024-03-06T18:37:36Z |
| publishDate | 2012 |
| publisher | University of Oxford |
| record_format | dspace |
| spelling | oxford-uuid:0bc9e862-0b1c-432c-a8a5-c494c6bb96d02022-03-26T09:31:20ZForecasting by factors, by variables, or both?Working paperhttp://purl.org/coar/resource_type/c_8042uuid:0bc9e862-0b1c-432c-a8a5-c494c6bb96d0Symplectic ElementsBulk import via SwordUniversity of Oxford2012Castle, JHendry, DWe consider forecasting with factors, variables and both, modeling in-sample using Autometrics so all principal components and variables can be included jointly, while tackling multiple breaks by impulse-indicator saturation. A forecast-error taxonomy for factor models highlights the impacts of location shifts on forecast-error biases. Forecasting US GDP over 1-, 4- and 8-step horizons using the dataset from Stock and Watson (2009) updated to 2011:2 shows factor models are more useful for nowcasting or short-term forecasting, but their relative performance declines as the forecast horizon increases. Forecasts for GDP levels highlight the need for robust strategies such as intercept corrections or differencing when location shifts occur, as in the recent financial crisis. |
| spellingShingle | Castle, J Hendry, D Forecasting by factors, by variables, or both? |
| title | Forecasting by factors, by variables, or both? |
| title_full | Forecasting by factors, by variables, or both? |
| title_fullStr | Forecasting by factors, by variables, or both? |
| title_full_unstemmed | Forecasting by factors, by variables, or both? |
| title_short | Forecasting by factors, by variables, or both? |
| title_sort | forecasting by factors by variables or both |
| work_keys_str_mv | AT castlej forecastingbyfactorsbyvariablesorboth AT hendryd forecastingbyfactorsbyvariablesorboth |