TBC
Nowcasting refers to the practice of using recently published data to update key economic indicators that are published with a significant lag, such as real GDP. The aim of this course is to familiarise participants with cutting-edge nowcasting tools that facilitate the use of mixed-frequency data in regression models. The course begins by establishing the importance of nowcasting for more timely and appropriate policy formulation during crisis periods such as the Great Financial Crisis or COVID-19. It then reviews the standard nowcasting regression-based procedures available, including the Bridge, MIDAS, and U-MIDAS estimators, both with and without dynamic factors. The course also reviews the more general state-space/Kalman filter approach to formulating and estimating a nowcasting model with mixed-frequency data. Procedures for combining nowcasts from the distinct models are considered, along with statistical procedures for evaluating the accuracy of a sequence of nowcasts. Each topic is complemented by a hands-on workshops and assignment using country-specific data using the EViews econometric package. The workshops and assignments are an integral component of the course designed to illuminate the actual steps required to generate a nowcast.
The course will be delivered by faculty from the IMF’s Singapore Training Institute (IMF-STI).
At the end of the course, participants should be able to: (1) understand and be proficient in the steps required to manage time-series data in EViews; (2) estimate an OLS regression and calculate its associated forecasts in EViews; (3) formulate several useful statistical procedures in EViews, including consolidation of time series from higher to lower frequencies, interpolation techniques, seasonal adjustment and use of leading indicators; (4) identify appropriate high-frequency indicators useful for the nowcasting macroeconomic variables and prepare them for use in a nowcasting exercise; (5) formulate and estimate a nowcasting regression using several approaches (including Bridge, MIDAS, and U-MIDAS estimators); (6) generate a nowcast from the base regression and consolidate competing forecasts using combination forecasts; (7) evaluate the accuracy of the nowcast using several forecasting performance indicators; and (8) apply the nowcasting tools to their own country data and interpret the nowcast appropriately in policy making settings.
Participants are expected to have an advanced degree in economics or equivalent experience, a basic understanding of time-series econometrics and be comfortable using the EViews econometric software package. It is strongly recommended that applicants have completed a few general macroeconomic courses.
Some of the EViews exercises might require more recent versions of EViews, as earlier versions might not be able to handle certain functionalities. Ideally you should have EViews version 12 or later.
