Reference-Dependent And Relative Leverage Premium

Corporate finance and asset pricing, which are enduring topic both in the academic and the business field, are known to be the core of modern finance. However, the relationship between the capital structure and asset pricing seems to be a puzzle. China is the largest developing country in transition economies. With the SME and GEM set up, China's multi-level capital market basically formed. The market environment of China is unique compared to other developed western countries represented by the United States. The capital structure and asset pricing decisions of China may exhibit different characteristics. However, there are few existing studies for big firms in developing countries like China, let alone the study on small and mediums firms. What are the dynamic characteristics of the capital structure of SMEs? Are the firm-specific determinants of target leverage identified in developed countries also adapted to small firms? Does leverage help explain cross-sectional expected returns in SMEs? Under the background of economic globalization and the era of new economic, the above questions are what we would explore in this paper.Under the framework of dynamic capital structure, we the reference point of firms' capital structure and also the leverage premium effect using the data from SME and NASDAQ during 2004 to 2013. First, we employ a dynamical model to estimate the determinants of target leverage and adjustment speed of the two markets using two-step system GMM method. Then we estimate the target leverage based on the results and decompose leverage into target leverage and relative leverage, to examine whether relative leverage help explain cross-sectional expected returns in SME and NASDAQ.The results show that the adjustment speed of SME is larger than that of NASDAQ because of weaker institutional environment. In addition, the firm-specific determinants of target leverage identified for larger firms also hold for small firms. However, institutional differences could indirectly affect their roles. We find that relative leverage is significantly and positively related to expected returns both in SME and NASDAQ. In addition, the positive relation keeps significant even when the observed leverage is controlled. Instead of the observed leverage, relative leverage which considers reference-dependent is the relevant variable in explaining stock returns.