Chinese economic behavior in times of covid-19. A new leading economic indicator based on Google trends.

Abstract

Since December 2019 we have been living with a virus called SARS-CoV-2 which has led to health policies being given prevalence over economic ones, causing serious consequences with regard to China's economic growth. For this purpose, we have built a Real Time Leading Economic Indicator based on Google Trends that improves the performance of Composite Leading Indicators (CLIs) to anticipate GDP trends and turning points for the Chinese economy. First, we assess the effectiveness of this new leading indicator relative to China's GDP by analyzing its statistical properties. We use fractional integration techniques to show the high degree of persistence of the new Real Time Leading Economic Indicator (RT-LEI) for China. Second, we observe the same relationship between GDP and RT-LEI in the long term using a Fractional Cointegration VAR (FCVAR) model. Third, we use a multivariate Continuous Wavelet Transform analysis to show which leading indicator best fits GDP and to identify when a structural change occurs. Finally, we forecast, using Artificial Neural Networks and a KNN model based on Machine Learning, our RT-LEI predicting the conclusion of a bearish scenario, after which the recovery begins in mid-2022.