The Provincial Measurement and Contagion Of The Generalized Local Government Bond Risk In China Based On Complex Network

Abstract

This paper constructs the generalized local government bond index of 30 provinces and municipalities (except Hong Kong, Macao, Taiwan and Tibet) by python, extracts its volatility by GARCH model, constructs the inter provincial spillover network of local government bond risk based on binary Granger causality test, studies its spatial-temporal structure by complex network analysis method, the network density represents the risk concentration level, and investigates the influencing factors of risk spillover by QAP method. The results show that: <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{C}1$</tex> The risk spillover network of local government bonds in China has strong relevance, high transmission efficiency and great difficulty in risk prevention and control and disposal. <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{C}$</tex> Hubei, Gansu and Jilin are the amplifiers of risk, while Hubei, Guangdong and Fujian are more intermediaries and bridges of risk transmission. <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\subseteq 3$</tex> QAP regression analysis shows that the impact of geographical distance on Risk Spillover is weak in economic sense. Local government bond risk contagion is affected by industrial structure, local government financial resources, the development level and development structure of local financial market, which is the key factor driving the whole local bond risk spillover network. The above research based on the price perspective has important policy enlightenment for us to prevent the grey rhinoceros systemic financial risk superposition transmission of local bonds.