Essays on Bonds: A Comparative Approach

This thesis comprises three comparative empirical chapters that explore distinct areas in bond market literature. The first chapter investigates whether Islamic bonds (sukuk) should be rated the same way as conventional bonds. Using a sample of Malaysian bonds, we address this question by examining the credit rating determinants of each type (conventional and Islamic) and testing the significance of Islamic bond features. Our results based on ordered probit and support vector machines show new evidence of the distinction between the two types of bonds, suggesting that their rating methodologies should differ. Sukuk and conventional bond ratings seem to share some common determinants, but their sensitivity to these determinants vary. Moreover, our findings suggest that conventional bond ratings are driven by a smaller set of financial variables, whilst Islamic bond ratings are triggered by a wider set of variables including Islamic structure variables. The most accurate bond rating predictions are achieved using tailor-made individual Islamic and conventional bond rating models. The support vector machines outperform the ordered probit model across all our samples and increase the bond rating prediction accuracy by more than 20%. Hence, to get the best results we suggest using support vector machines to forecast sukuk and conventional bond ratings, separately.

The second chapter investigates and compares herding in the US corporate bond and equity markets between January 2008 and December 2018. Our initial unconditional tests detect significant herding in speculative grade (high yield) corporate bonds only. However, once we condition for market liquidity and volatility, we find significant asymmetric herding behaviour in both markets and their credit rating portfolios. The results suggest that investors are inclined to collectively herd towards the market consensus during high market liquidity and low volatility. Interestingly, the herding effects are more pronounced in corporate bonds than in equities. The findings are robust even after simultaneously conditioning for both liquidity and volatility market states. Our further tests also provide new empirical evidence of the existence of cross-asset herding spillovers from US corporate bonds to their respective equities. The direction of the cross-herding effect holds during the 2008 global financial crisis, but switches post-crisis from equity to corporate bonds.

The final chapter explores the impact of a wide range (151) of Brexit-related news announcements on UK and EU government bond and stock market volatilities during April of 2010 to February 2020. Using the Heterogeneous Autoregressive (HAR) model, we find strong evidence of the importance of Brexit news carrying valuable information for both markets. Specifically, the initial regressions combining all Brexit news show that the volatility of the Eurozone's government bonds and the UK stock market considerably increased on the announcement days of Brexit news. The detected volatility responses to the news are more substantial in government bonds relative to equities. We also find that the degree and direction of the Brexit news effects varies depending on the news category. Particularly, among the different types of news, the markets seem to be most sensitive to UK votes and UK political news. In addition, we document leading and lagging effects of the news, suggesting that the volatility responses often extend for more than a day. Interestingly, the findings reveal that the Eurozone markets are more suspectible than the UK to Brexit news. Lastly, the contribution analysis shows that during the 3-day event window, aside from UK votes, all Brexit news categories induce higher levels of volatility that temporarily destabilise the UK and Eurozone markets.