Dynamic features of China's photovoltaic listed companies in different periods: Based on partial Granger causality network

Abstract

As an important kind of renewable energy, China's photovoltaic (PV) industry has experienced many challenges in different historical stages. Many PV enterprises emerge in the stock market, correlate with each other, and form China's PV stock market network. However, the mutual influence between any two companies in the stock market is impacted by other enterprises and some external factors, which is not considered in the previous studies. In order to study the direct interdependence among China's PV stock markets, we propose the partial Granger causality network (PGCN) model. It is a first practice in the model to apply partial Granger causality to quantify the direct interactions between stock returns. The daily closing prices of 79 China's PV enterprises are selected, and the whole research period 2007.10.2–2016.10.3 is divided into four sub-periods according to three important time nodes. The PGCN in the overall period and four sub-networks are also constructed. Combined with the dynamic behaviors of networks' topological properties, the distribution of enterprise's influence, the conductive force of enterprises, and the stability of the stock market are analyzed. Meanwhile, the regional agglomeration development pattern is revealed, and top 10 influential enterprises are identified.