A12: Fragmented Intermediation, Bond Risk Premia, and Market Stability

Intermediation in the corporate bond market has undergone a tremendous change over the last two decades. First, due to stricter regulation, bank-affiliated dealers that are typically marginal in many asset classes withdraw from providing dealer services. Second, mutual funds and ETFs that are often restricted through their investment policies to a rather limited asset universe hold an ever increasing proportion of the total outstanding volume in their portfolios. We aim to investigate this change in intermediation from institutions that are marginal in many asset classes to fragmented intermediation with intermediaries that are only marginal for particular bonds or bond market segments. Our project starts with a descriptive analysis of how intermediation has shifted from dealers to asset managers. To this end, we assess differences in the mixture of intermediation in the cross-section of bonds and analyze its relation to bond characteristics. A comparison of intermediation during different stress events will reveal how intermediation works during such market phases and how conditions changed over time. We then proceed to a theoretical and empirical analysis of how a bond’s holding structure and the way the bond is intermediated impact its price and risk premium. We also analyze the differences in the roles of ETFs and mutual funds to shed light on the impact of different liquidity management concepts on bond prices. In the last part of the project, we seek to understand the role of stress events and how different kinds of intermediation mitigate or reinforce market shocks. The mismatch between the liquidity of shares of mutual funds, which can often be sold at daily net asset values without costs, and the illiquid bonds in their portfolio could incentivize investors to sell their shares as quickly as possible and, for that reason, might lead to fragility during stress events. To capture a bond’s fragility, we develop an indicator for the likelihood that a particular bond suffers from fragility in case of a possible future stress event. It will be interesting to see whether such a fragility indicator is priced in the cross-section of bond returns.

Project Leader: Prof. Dr. Philipp Schuster and Prof. Dr. Marliese Uhrig-Homburg