Product network and origin of common equity factor risks

Abstract

This paper extends the multisector asset pricing model based on production networks, achieving direct pricing of a firm’s equity assets at a more microscopic level within the product network. An Equilibrium-Aligned PageRank (EA-PR) centrality algorithm is proposed, and it is theoretically proven that EA-PR centrality can reflect the relative value of products and further derive the value of firms. A new equity asset pricing factor is constructed based on the EA-PR centrality of firms. The key findings can be summarized as follows: Firstly, the EA-PR centrality of firms shows a significant positive linear predictability for equity returns, and the returns associated with the EA-PR centrality factor cannot be fully explained by the Fama–French 6 (FF6) factors. Secondly, the EA-PR centrality factor demonstrates significant explanatory power for the FF6 factors and, to some extent, exhibits predictive ability for future returns of the FF6 factors. Thirdly, the product network model can effectively reflect the market competitiveness and growth potential of firms, and firms with high EA-PR centrality have stronger value capture capabilities, improving gross profit margin, cash flow, and investment growth in the future. Finally, this study empirically demonstrates the consistency between the product network model and the pricing theory of the $q^5$ model, which incorporates the expected growth factor.