Corporate Governance and Corporate Bond Liquidity

Abstract

Using a unique dataset of corporate bond trading information and corporate governance evaluation scores, this study examines the determinants of corporate bond market liquidity in Korea. In particular, this study explores whether corporate governance performance of a company influences liquidity of bonds issued by the company. The paper reports three important findings. First, the issue size and age of bond are important determinants of bond liquidity. Second, liquidity of corporate bonds is influenced by changes in macroeconomic conditions. Third, and most importantly, better corporate governance increases liquidity of corporate bonds. This result suggests that corporate governance is an important determinant of bond liquidity, as it lowers transaction costs by improving transparency and reducing asymmetry of information. This paper contributes to the literature by providing new evidence that corporate governance performance is an important determinant of liquidity in corporate bond markets.

Key Words:

• Corporate bond
• trading volume
• corporate governance

Acknowledgement

This paper is the extension of one chapter of Hyun Jin Lee's Ph.D. dissertation, “A Study on Bond Trading and CDS Spread,” and title of the chapter is “The determinants of trading volume of corporate bonds: focusing on the corporate governance.”

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 O'Hara (1995) defines liquidity as the ability to trade swiftly at low costs. Although liquidity is easy to define theoretically, it is not easy to measure accurately and credibly in practice, except in frequently traded market.

2 The inventory paradigm was first introduced by Demsetz (1968) and was further developed by Stoll (1989) and Ho & Stoll (1981). According to the inventory paradigm, a bond with a lower trading volume requires a higher inventory holding cost for the dealers. As this inventory holding cost is eventually transferred to investors as a part of the transaction cost, a lower trading volume decreases the liquidity of the bond by raising the transaction cost. For example, it is easier for dealers to adjust the size of the inventory for bonds with a higher trading volume, which implies a lower inventory holding cost for bonds with a higher trading volume than for those with a lower trading volume. Thus, bond trading volume can be directly linked to liquidity since it affects the dealer's inventory holding cost. (Bamber, 1986; Krinsky & Lee, 1996; Bamber et al., 1999; Harris & Raviv, 1993)

3 For the homogeneity of the sample bonds, the dataset only includes straight bonds. Non-straight bonds such as asset-backed securities, convertible bonds, guaranteed bonds and option-embedded bonds are excluded from the sample. Moreover, for better reliance, only the bonds issued by KRX-listed companies are included. The trading volume of bonds issued by Korea Securities Dealers Automated Quotation (KOSDAQ)-listed companies are more likely to be affected by speculative components. The total number of corporate bonds issued by these 200 sample companies over the period 2003-2007 is 1141. However, 181 corporate bonds were never traded over this 5-year period. In order to focus more on the corporate bonds that had been in demand in the market, this study only uses the 959 corporate bonds that had been traded at least once over the given period. A statistical description for the sample of 1141 corporate bonds is reported in Appendix Table 2, and a dataset including all of the 1141 corporate bonds will later be used for Tobit analysis.

4 For over-the-counter trading, the KOFIA keeps all transaction data from both sellers’ and buyers’ sides. In order to avoid possible double-counting problems, this study only uses selling side data for over-the-counter trading.

5 This study uses dummy variables that indicate the months when the central bank changes the direction of key interest rates, instead of the key interest rate variable. The use of interest rate variables incurs serious imperfect multi-collinearity problem with other control variables such as interest rate risks and price volatility variables.

6 The first stage estimation result suggests that these four financial performance indicators are good instruments for the corporate governance performance. The F-statistics from the first stage regression is 310.2, which confirms that there is no weak instrument variable problem.