Dynamic Factor Model to Nowcasting the Sectoral Economic Growth Using High-Frequency Data Putu Krishnanda Supriyatna, Dedy Dwi Prastyo, Muhammad Sjahid Akbar
Institut Teknologi Sepuluh Nopember
Abstract
Economic growth is an important indicator used for evaluation and planning by many users. Unfortunately, the release of economic growth data has a delay of more than a month. A delay in the availability of economic indicators will impact unpreparedness in dealing with an economic condition. On the other hand, the timely availability of economic indicators can optimize their use in evaluating the economic situation. The nowcasting method is a solution to get an overview of economic conditions more quickly and in near real-time. To provide information quickly, other indicators are needed as proxy components that can be obtained in real-time using high-frequency data. The predictive power of high-frequency data is better because it contains more information. Moreover, economic growth data are only available quarterly. Therefore, such information will be missing if the high-frequency variables used as proxy components are transformed into quarterly data. Therefore, a method is needed to overcome these issues. One method that can be utilized to solve this problem is the DFM (Dynamic Factor Model), which can perform nowcasting using high-frequency data. The predictors used in this nowcasting approach are indicators with a short time lag so that it can provide a quick estimation. Also, forecasting will be carried out in this study using the ARIMAX (Autoregressive Integrated Moving Average with Endogenous Variables) method to see the impact of high-frequency data, with the monthly variables will be transformed into quarterly in the ARIMAX method. The nowcasting results between the DFM and ARIMAX methods will be compared to conclude which method performs better. In addition, the same procedure will be applied to nowcasting the economic growth at the sectoral level.
