Work

Application of the Dynamic Factor Model on Nowcasting Sectoral Economic Growth With High-frequency Data

Year: 2024

Type: article

Abstract: Economic growth is crucial for planning, yet delayed data releases challenge timely decision-making. Nowcasting offers near-real-time insights using high-frequency indicators (released monthly, weekly, or even daily) to predict low-frequency variables (quarterly or yearly). This study uses high-frequency indicators (monthly), such as stock price changes, air quality, transportation data, financial conditions, and Google Trends, to nowcast quarterly GDP through the Dynamic Factor Model (DFM). The data used span from January 2010 until March 2023, which is split into two: January 2010 until March 2022 for training data and the rest as testing data. Compared to the benchmark Autoregressive Moving Average with Exogenous Variables (ARMAX) model, DFM demonstrates superior accuracy with lower symmetric Mean Absolute Percentage Error (sMAPE). In addition, to evaluate the model performance in nowcasting the GDP across the sector using DFM, the additional metrics, i.e., Root Mean Square Error (RMSE), Mean Absolute Deviation (MAD), and Adjusted R-squared, concluded that in the industrial and transportation sectors results in sufficient nowcasting of GDP, Meanwhile, In the financial sector, the results of the nowcasting GDP give poor estimation results that need improvement.

Source: MEDIA STATISTIKA

DOI: 10.14710/medstat.17.2.128-139

Authors: Putu Krishnanda Supriyatna, Dedy Dwi Prastyo, Muhammad Sjahid Akbar

Institutions: Institut Teknologi Sepuluh Nopember, Badan Pusat Statistik

Language: English

Topic: Forecasting Techniques and Applications

Subfield: Management Science and Operations Research

Field: Decision Sciences

Domain: Social Sciences

Open Access status: gold

Estimated APC paid: $32

Funders: Institut Teknologi Sepuluh Nopember, Kementerian Pendidikan, Kebudayaan, Riset, dan Teknologi