The Economic Consequences Of Network Centrality In Social And Economic Network

As globalization and financial integration deepens,the interconnectedness between countries,industries,companies,and investors makes major changes to the process of shock propagation and information transmission.As a result,the behavior of market participant and the formation of asset prices also changes.The network analysis tools emphasize the interconnectedness while traditional economic methodology and asset pricing theory neglect it.From the perspective of the social and economic network,the heterogeneity of market participants(nodes)does not only depend on their economic attributes but also depend on their network centrality – the number and strength of the connections and how they are connected.Given the variety and complexity of economic networks,the network position of different nodes varies greatly,and their roles in shock propagation and information transmission are also significantly different.China,as a country with increasingly important international roles,significant differences between domestic market participants,and closely related market participants,cannot afford to ignore the economic consequences of network positions.However,due to the complexity and endogeneity of the network structure,identifying the heterogeneity of the network position still faces many challenges.Therefore,it is of great practical value and theoretical significance to study the economic consequences of the position of nodes in the network.From the perspective of asset pricing,this paper examines the role of the network position in the process of shock propagation and information transmission and thereby the impact on the asset price,under the settings of spillover effect,information asymmetry,and herding,using industry input-output network,institutional investors network and fund colleague network correspondingly.This paper finds: First,central nodes which have more connections,are closer to other nodes,or connected to more central nodes can obtain “power” through two channels: First,central nodes are better connected and more reachable than other nodes,therefore having better access to information and greater exposure to risks.Second,central nodes have fewer constraints and more choices,thus having larger bargaining power and lower switching cost,which provides the possibility of collusion and risk resistance capacity.Since information and risk are the critical issues of the asset price,the “power” can bring agents advantages on investment returns.Second,social and economic networks contribute significantly to the process of shock propagation and information transmission and therefore affect the asset price significantly.Under each specific setting,the findings are as follows.First,based on 733 manually identified sectoral shocks from 1995 to 2011,this paper explores the spillover effects of sectoral shocks through input-output interactions.This paper finds that sectoral shocks can significantly affect downstream sectors even three connections away from the origin of a sectoral shock.Based on spillover effects,this paper investigates the role of the network position in the propagation of shocks and finds that central sectors are better at mitigating negative impacts of disruptions from other sectors.This paper then establishes a multi-period multi-sector model to investigate the relationship between input-output linkages and the origin of systematic risk.The general equilibrium demonstrates that the spillover effects through the inputoutput network can aggregate into undiversified systematic risk.Furthermore,the theory model finds that network centrality is positively related to the stochastic discount factor(SDF)and should be a pricing factor.Using the industry input-output data from China,this paper constructs the CMP(core-minus-peripheral)factor based on the industry network centrality and empirically demonstrated that the CMP factor is a new pricing factor.This paper finds that central sectors have greater exposure to systematic risk.Although central sectors can absorb spillover effects,they have a higher probability of being reached by shocks from other sectors,and the latter effect overrides the former.Therefore,central sectors have higher shock returns.This paper then investigates the impact of the network position under the setting of information asymmetry.Based on a manual collected account-level bidding data from 2010 to 2012,this paper examines how relationships between institutional investors affect their bidding behaviors in Chinese auctioned IPOs.This paper finds that central investors have better access to information,and network connections serve as a possible channel for collusion.Central investors are more likely to participate in IPOs,which reduces the participation uncertainty problem in auctioned IPOs.Second,information transmission through the network linkages can help mitigate the free-riding problem.Third,a denser relationship network will formulate a more homogenous belief about the value of IPO shares.After the IPO bidding process begins,investors in the denser relationship network bid less dispersedly and closer to the air price.Furthermore,connected investors can conspire to depress the bidding price.Consequently,IPOs with denser networks experience smaller dispersion in bidding prices and earn higher initial returns.Finally,this paper explores the information transmission effect in the mutual fund market from the perspective of herding.Using the mutual fund data from 2006 to 2018,I construct the economic network using colleague ties between mutual fund managers.This paper finds that central funds tend to buy and sell similar stocks with other funds in the same fund family and hold shares that are more heavily held by other funds,which indicate the herding through network connections.Central funds also have better investment ability(stock selection and market timing).The economic consequence is that central funds significantly outperform peripheral funds.Furthermore,the findings indicate that connections with other managers within the same fund family have significant impacts on the performance of mutual funds.In contrast,connections across different fund families have no significant effect.Overall,this paper documents the economic consequences of an agent’s position in the network.The findings of this paper can shed some light on the study of network heterogeneity,interconnectedness,and asset price in other settings.This paper also contributes to the identification of systematically significant industry or institution,the optimization of IPO design,and the efficiency of capital allocation.