This dissertation focuses on high-frequency volatility-return relationship and dealer trading strategies in the corporate bond market, delivering a comprehensive analysis on corporate bond pricing through two chapters.

In Chapter 1, we study the relation between high-frequency return volatility and cross-section corporate bond returns. Utilizing intraday transactions, we employ the good minus bad volatility constructed from the weekly realized variance to predict future corporate bond returns. Both portfolio and cross-sectional regression results confirm the negative predictability of the relative signed jump variation in capturing additional information beyond bond characteristics and conventional risk factors. Enabled by the recent development of econometrics techniques and newly available intraday transaction records, we shed light on linking high-frequency bond return volatility to the cross-sectional determinants of corporate bond returns.

In Chapter 2, we examine dealer markups in rising and falling corporate bond market. We show that dealer markups increase faster when price rises than they decrease when price falls. The asymmetries are more pronounced for bonds with higher illiquidity and larger search costs, which can be explained by the search-based theory. Moreover, we find the dealers’ asymmetric behavior is more significant during the distress times, and we provide evidence that the implementation of the Volcker Rule weakens the asymmetries. Our results indicate that illiquidity and higher search costs are important factors impeding price discovery in the OTC corporate bond markets.