Does liquidity spill over? A study on liquidity patterns between Corporate Bond and Equity markets across issuers

Karoliina Hienonen

Student thesis: Master thesis


This paper seeks to establish common patterns between liquidity in publicly traded equities and corporate debt (bonds). The underlying logic follows a straightforward path: by focusing on purely issuer level factors it is argued that if the operational and informational content regarding cash flows and asset valuation is ultimately embodied at the level of the issuer, and if both securities essentially represent claims to these same assets, it follows that trading patterns and therefore liquidity should reflect similar dynamics to some extent, while acknowledging the fundamental structural differences in trading the two. This hypothesis is put to test with various panel regression methodologies that incorporate multiple issuer variables in the hopes of explaining the relationship in a more nuanced manner. The studied sample is comprised of some of the largest global issuers trading in the North American market– as measured by market capitalization, and the variables modelled include some of the most commonly applied financial metrics, embodying angles of both, risk and return. For liquidity, this paper uses the popular Amihud ILLIQ as a proxy. The econometric approach has its point of departure in simply treating liquidity in each of the asset class as the dependent and independent variables in turn, while introducing additional issuer metrics as control variables in three more stages. These stages are based upon variable nature, where those relating to firm fundamentals such as size and leverage are introduced first, then complemented by variables more influenced by market speculative factors, such as Beta and a valuation ratio Price-to-Book. Finally a set of lagged variables are included. After it has been confirmed that the panel sample at hand demonstrates a fixed effect at the cross-sectional level, the relationship is studied through a series of panel regression estimations with one way cross-sectional fixed effects. The robustness of the established results is finally evaluated via timewise subsamples. It was found that indeed a robust issuer level linkage seems to exist, however it is more described by the cross-sectional fixed effects rather than liquidity in the other market or any of the introduced firm factors. In other words, the commonality seems to be better described as flowing from the top (issuer) rather than spilling over from one market to the other. Results regarding the firm factors do not give clear enough indication on linkages for clear conclusions to be drawn in a robust manner, apart from firm size which economically seems obvious as information content is seen to increase with size. Nevertheless, as the fundamental goal of this paper was simply to find evidence that relative liquidity in each of the markets is driven by the common denominator, the issuer, this result in its simplicity is considered meaningful with multiple insights for further studying the relationship with more sophisticated approaches.

EducationsMSc in Applied Economics and Finance, (Graduate Programme) Final Thesis
Publication date2016
Number of pages95