Nowcasting Indonesian Economic Growth at Risk with MIDAS-Quantile Regression
Turfah Latifah (a), Muhammad Sjahid Akbar (b*), Dedy Dwi Prastyo (c)

Department of Statistics, Faculty of Science and Data Analytics, Institut Teknologi Sepuluh Nopember (ITS),
Jl. Teknik Mesin 175 Sukolilo, Surabaya, 60111, Indonesia

* m_syahid_a[at]statistika.its.ac.id

Abstract

This study aims to nowcast Indonesian economic growth using the Growth at Risk (GaR) model based on Mixed Data Sampling-Quantile Regression (MIDAS-QR). The MIDAS-QR approach enables the prediction of economic growth risks at various quantile levels, thus providing an overview of the best and worst-case scenarios. This approach integrates multi-frequency data grouped into the Financial Conditions Index (FCI), External Financial Environment Index (EFEI), and Macroeconomic Prosperity Leading Index (MPLI) using Principal Component Analysis (PCA). The model is evaluated using Quantile Mean Absolute Error (QMAE) and Quantile Root Mean Squared Error (QRMSE) to measure prediction accuracy. The findings reveal that the MIDAS-QR model significantly enhances the accuracy of predicting GaR for Indonesian GDP compared to conventional models based on the QMAE and QRMSE metrics. The model effectively captures the risk dynamics at various quantile levels, particularly upside and downside risks, demonstrating a reduction in prediction errors compared to traditional methods. The integration of high-frequency data through FCI, EFEI, and MPLI allows for timely detection of potential downturns, providing critical insights for policymakers to take preemptive actions. This study underscores the importance of utilizing high-frequency data in economic forecasting to improve policymaking effectiveness in Indonesia. The MIDAS-QR model not only enhances the accuracy of GDP growth predictions but also serves as a valuable tool for assessing economic risks, real-time monitoring, and decision-making, especially in times of economic uncertainty.

Keywords: Indonesian Economic Growth at Risk (GaR), MIDAS-Quantile Regression, Principal Component Analysis (PCA), Quantile Mean Absolute Error (QMAE), Quantile Root Mean Squared Error (QRMSE)

Topic: Development economics