Nowcasting and Short-term Forecasting of Chinese Macroeconomic Aggregates: Based on the Empirical Study of MIDAS Model

Liu Han,Liu Jinquan
2011-01-01
Abstract:The trend and fluctuations in quarterly GDP have great influence on the revenue and expenditure of governments, the profits and financial situation of corporate,and even have effects on the income and expenditure of families and individuals. They are the key variables in macroeconomic forecasting,prediction and analysis. Traditional macroeconomic forecasting models are based on the same frequency data,high frequency and ultra high frequency data must be change into low-frequency data,which ignored the information of high-frequency data,and timeliness of forecasts and the accuracy of prediction are decreased. Mixed data sampling ( MIDAS) models are used for forecasting and nowcasting Chinese quarter GDP,and empirical results show that the export is the major factor caused the depreciate of Chinese economic growth during the period of financial crisis. The results verify the comparative advantage of MIDAS models in the accuracy of macroeconomic short-term forecasting,and have significant feasibility and timeliness in nowcasting.
What problem does this paper attempt to address?