Detecting the interactions among firms in distinct links of the industry chain by motif

As a complicated system, the stock market includes listed companies from distinct industries. These industries can be connected as industry chains based on the relations of their product. The upstream, midstream and downstream of one industry chain are closely connected by value and information exchange. Hence, the fluctuation of a company's stock price in one link will affect those of the companies in other links. Exploring main intermediaries in the industrial chain and typical paths of stock price transmission can offer risk warnings to market participants, such as policy makers, company operators and investors. This paper combines multilayer networks and motifs to propose a new framework for studying the relationships among listed companies in different links of one industry chain. The results and implications are as follows: (1) There is an apparent feedback mechanism from the downstream (the finished automobile and charging) to the other links, especially the midstream (vehicle components). Policy-makers need to focus on monitoring the impacts of downstream changes on the midstream. (2) The important intermediary companies transmitting information between the links (layers) are identified. When one link has great fluctuation, the range of impacts on other links can be reduced by controlling the intermediary companies. (3) Ten typical spillover transmission paths linking the three links with the largest spillover amounts are identified. According to the path, the company managers can assess the possibility of the company being affected and adjust the risk prevention strategy in time. (paper: interdisciplinary statistical mechanics)