Empirical Studies On Presence Of Peer Effects In Capital Structure Decision Making And On The Formation Mechanisms

Since Modigliani and Miller firstly propose M&M capital structure model in 1958,scholars in the field of corporate finance,by relaxing prior strict assumptions,further study how factors such as tax shield benefit,bankruptcy cost,agency cost,residual control right and free cash flow would affect firms’ capital structure decision,which lays the foundation for the modern capital structure theory.According to capital structure theory based on perspective of corporate finance,firms’ optimal capital structure is the function of their own characteristics.Neither other firms,actions nor characteristics play a role in the determination of focal firms’capital structure.In sharp comparison to the theory,in reality,due to the existence of some possible intertwined relationships among firms(such as operating in the same product market or doing business with upstream or downstream partners),firms’ own decisions are in essence learning behavior,feedback or even counter measures in response to other peer firms.In other words,given that theoretical mechanism of peer effects is at play,decisions of firms within a reference group are interrelated.This peer effects phenomenon applies not only to strategic and product competition decisions but also to corporate finance behaviors,for instance investment and financing,etc.Up till now,some foreign researchers have put some efforts in empirically detecting peer effects of corporate finance behaviors and made some significant progress.Although Chinese scholars have realized the importance of research on peer effects,much work needs to be done.Hence,this study focuses on peer effects of capital structure decision.By combining theoretical modeling and empirical examination methods and applying them onto financial data of Chinese non-financial A-share listed companies,this dissertation answers two important questions,i.e.,whether peer effects are present in capital structure decision and what are the underlying mechanisms behind the peer effects.Hopefully,our work will induce more dedication to empirical investigation on peer effects of corporate finance decisions.This dissertation has seven chapters.The first chapter is the introduction which brings up the research topic of our interest and its research implications.The definitions of key concepts are also provided.Then follow the research idea,specific research contents,methods and innovation of this dissertation.Our second chapter summarizes and reviews current literatures on peer effects in the field of economics and corporate finance.Third chapter introduces basic theory of capital structure and peer effects,both of which lay the theoretical foundation for analyzing peer effects in capital structure decision making.In addtion,some commonly used empirical identification models,as well as some indirect methods,are presented.Using financial data from Chinese non-financial listed firms,the fourth chapter empirically examines the presence of peer effects in capital structure decision by controlling for correlated effects.Basing on economically important and statistically significant relationship among debt decision of intra-industry firms,the fifth and the sixth chapter both theoretically analyze and empirically test whether two specific formation mechanisms(learning and competition)are able to lead to peer effects in capital structure decisions.Lastly comes the seventh chapter which outlines research conclusions,policy implications,possible research limitations and future research topics.Main research conclusions of this study can be summarized as follows:(1)Capital structure of non-financial Chinese listed companies exhibit peer effects.Using linear-in-means regression model,employing peer firms’ stock return shock as instrumental variable to control for reflection problem and applying firm’s fixed effects and differencing to partially diminish correlated effects,this paper affirms economically important and statistically significant relationship between focal firms’ capital structure and that of peers.In order to further distinguish correlated effects from above-mentioned significant relationship so that our conclusion is reliable,this study tests the asymmetric response of focal firms to peers’ debt change of different direction and different scale,as well as the existence of social multiplier effects.In addition,we construct another instrumental variable that is less influenced by correlated effects.All these supplementary evidence corroborates the presence of peer effects in capital structure decision.(2)Learning among intra-industry firms is proven to be a valid formation mechanism of peer effects in capital structure decisions.According to learning theory and capital structure decision model relying on priori information,private information and peer information from observing peers’ capital structure decisions,this dissertation theoretically derives and empirically testifies how heterogenous factors of learning(namely focal firms’ senior managers’ capability,external environmental uncertainties,peer firms’ informational disclosure quality and focal firms’ TMT informational exchange level)influence intensity of peer effects.In this way,this paper offers suggestive hints that one vital mechanism of peer effects in capital structure decision is learning.(3)Dynamic competition theory is another valid mechanism of peer effects in capital structure.This paper firstly derives optimal strategic debt decision under strategic pre-commitment theory and further deduces research hypotheses on how product competition strategy(substitutes or complements competition)would alter the direction of peer effects and how product competition interaction level,firms’previous debt level and industry competition level would moderate the absolute value of the slope of firm i’s optimal strategic debt against peer’s debt choice,hence the intensity of peer effects(specifically,positive/negative peer effects corresponding to complements and substitutes competition).Instrumental regressions performed on appropriate sub-samples support previous hypotheses.This study,thereby,implies that peer effects in capital structure decision represent,in nature,"attack and response" of debt decision among intra-industry competing firms.This research innovatively introduces peer effects,one popular research topic in education and social studies,into corporate finance decisions.Firstly,we perform empirical analysis to test the presence of peer effects in capital structure decision among Chinese non-financial listed firms,which is followed by investigation of formation mechanisms,i.e,learning and competition.To sum up,this dissertation not only extends current research angle of capital structure theory from inside the firm to outside the firm,but also provides ideas indicating how to i)control for reflection problem,a problem commonly seen when using linear-in-means regression model,ii)to distinguish endogenous peer effects and exogenous correlated effects and iii)to empirically test explanatory power of potential theoretical formation mechanisms.This research project is of importance in the sense that it,on the one hand,extends and deepens capital structure theory and,on the other hand,provides constructive insight into how firms shall optimize their capital structure decision in response to peers.