Abstract
This thesis focuses on the pricing of the Chinese bond market. Specifically, we explore the return predictability of treasury bonds with macroeconomic factors utilizing machine learning methods (Chapter 2), the evaluation of corporate bonds with bond characteristics and aggregate predictors (Chapter 3), and the pricing power of equity jump and diffusion risks for corporate bond credit spreads (Chapter 4). All three chapters provide novel empirical evidence and draw conclusions that have implications for market participants, policymakers, and regulators.In Chapter 2, Bond return predictability: Macro factors and machine learning methods, we investigate the impact of macroeconomic variables on bond risk premia prediction via machine learning techniques. Using Chinese treasury bond yields from March 2006 to December 2022, we show that adding macroeconomic factors improves bond return forecasts and generates higher economic benefits to investors. This is achieved when the nonlinear relationship between macroeconomic variables and bond returns is modeled via machine learning methods. Furthermore, the importance of macroeconomic determinants changes along the yield curve. This chapter sheds new light on the information contained in macroeconomic variables for treasury bond valuation and highlights the importance of utilizing appropriate machine learning methods in this important treasury bond market.
In Chapter 3, Corporate bond returns prediction with machine learning methods, we explore the return predictability of corporate bonds using bond characteristics and aggregate predictors with machine learning techniques. Based on the Chinese corporate bond data from January 2008 to December 2023, we show that two sets of variables provide unique information for corporate bond return forecasts when utilizing machine learning methods, with bond characteristics playing the dominant role. Investment strategies based on machine learning forecasts, such as random forest, yield significant and sizeable economic benefits to investors. Moreover, the importance of determinants differs between linear and nonlinear models, with nonlinear models placing greater emphasis on skewness and kurtosis. Finally, we observe a substantial improvement in bond return predictability following the implementation of the no-bailout reform in the Chinese bond market, with default-risk-related indicators becoming more critical after the reform. These findings advance our understanding of the sources driving corporate bond return dynamics in China.
In Chapter 4, Jump and diffusion risks in corporate bond pricing, we examine whether equity jump and diffusion risks are priced in corporate bond credit spreads in China. Based on the stock and corporate bond data from January 2009 to June 2023, our results show that jumps have significant explanatory power for bond prices, with bonds issued by firms with higher jump variation having larger credit spreads. This effect is statistically significant and economically substantial. Furthermore, the impact of jump variation differs across issuer ownership and submarkets: jump risk has a stronger impact on the credit spreads of non-state-owned-enterprises (non-SOEs) and on bonds traded in the exchange market. Finally, both positive and negative jump variations explain bond credit spreads, with negative jumps affecting credit spreads primarily through their influence on default risk. This study highlights the link between equity and bond markets and sheds light on corporate bond valuation in China.
| Date of Award | 13 Jul 2025 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Xiaoquan Liu (Supervisor), Ying Jiang (Supervisor) & Fumin Zhu (Supervisor) |