Cross-industry asset allocation with the spatial interaction on multiple risk transmission channels

Taking into consideration the real link and information risk transmission channels, we used a spatial econometric approach to construct an economic distance-based spatial weight matrix, which can capture the spatial interaction across industries, and built a return estimation model with spatial interaction using the matrix. On this basis, we derived the covariance matrix and constructed the cross-industry asset allocation model. The empirical results showed that 1) the spatial interaction has a strong explanatory power to return and integrating the spatial interaction on multiple risk transmission channels can improve the effectiveness of the return estimation model; 2) the covariance matrix includes unsystematic risk (idiosyncratic risk) and systematic risk (market risk and cross-industry spillover risk); 3) the asset allocation model with spatial inter-action can improve the performance of the portfolio and provide a valuable reference for in-vestors' risk management and investment decision.

Automated summary

By using a spatial econometric approach, this study constructs an economic distance-based spatial weight matrix to capture the spatial interaction across industries. A return estimation model is then built based on the spatial interaction, and the covariance matrix and cross-industry asset allocation model are derived. The empirical results demonstrate that spatial interaction has a strong explanatory power for returns, integrating spatial interaction on multiple risk transmission channels can enhance the effectiveness of the return estimation model, and the asset allocation model with spatial interaction can improve portfolio performance and provide valuable guidance for investors' risk management and investment decisions.