Research on the Correlation between China’s Juglar Cycle and Stock Price Fluctuation Based on Singular Spectrum Analysis

Yin Zhujia; Hu Hui; Tang Tanling

doi:doi:10.16538/j.cnki.jsufe.2019.06.002

Summary

The judgment of the new round of the global Juglar cycle has been recognized internationally, and there are indications that China is in the early stage of the new round of the Juglar cycle. Based on monthly data from January 2006 to December 2018, this paper studies the correlation between the Juglar cycle and the stock price fluctuation cycle of equipment manufacturing industry in China. Firstly, according to the characteristics of the Juglar cycle and the stock price fluctuation cycle of equipment manufacturing industry, 9 indicators are respectively selected from macro level, industry level and micro level to construct the comprehensive index of the Juglar cycle, and 9 representative secondary industry stock price indexes in the Shenyin Wanguo industry classification are compiled into the comprehensive index of the equipment manufacturing stock price fluctuation cycle. Secondly, the entropy method is used to calculate the scores of the two periods. Then, periodic characteristics and periodic fluctuation relations are studied by using the singular spectrum analysis method from the perspective of the combination of time and frequency domains. It is found that the first and second decomposition sequences of the Juglar cycle are periodic items, and the first, second, third, fourth, and eleventh decomposition sequences of the equipment manufacturing stock price fluctuation cycle are periodic items, and by means of the diagonal averaging, the periodic sequence of the Juglar cycle and the equipment manufacturing stock price fluctuation cycle are obtained; then, according to the periodic sequence of the Juglar cycle and the equipment manufacturing stock price fluctuation cycle, the overall correlation coefficient of these two periods is 0.6352, which indicates that there is a strong correlation between the two cycles; moreover, after combining the two periodic fluctuation graphs, it is found that the stock price fluctuation cycle of equipment manufacturing industry is ahead of the Juglar cycle for about 3-6 months, which indicates that the stock price fluctuation of China’s equipment manufacturing industry can reflect the future trend of the Juglar cycle to a certain extent, showing that the stock market is a barometer of macro economy. Finally, the singular spectrum analysis and autoregressive analysis methods are combined to construct the prediction model. It is found that the prediction error of the model is small and the effect is good in the next 12 months. The research findings in this paper confirm the close relationship between China’s economic cycle and stock market volatility, which can provide important reference for investors to optimize investment strategies and government departments to formulate macro-control policies.