关键词: 季度实际GDP增长率, 实时预报, 因子混频数据模型, 单调性检验

Abstract: The paper tests whether the nowcast error of quarterly real GDP growth rate in China obtained by the factor mixed-frequency data models will gradually decrease with the adoption of new data information, which shows the statistical characteristic of monotonic decrease. The empirical results show that the nowcast monotonicity test results obtained from the factor mixed-frequency data models generally confirm that the nowcast accuracy of quarterly real GDP growth rate would be improved by adding more monthly updated real-time information, regardless of the length of the out-of-sample nowcast window and the data update interval. Moreover, the nowcast monotonicity tests are robust when the sample’s time-varying characteristics and stage attributes change. The test is an effective supplement and extension to the existing forecasting and nowcasting evaluation methods, and can provide valuable information and consultation for governments, institutions, and investors to assess the macroeconomic conditions in real time.

Key words: Real GDP Growth Rate, Nowcasting, FA-MIDAS, Monotonicity Test