Marketability and Default Influences on the Yield Premia of Speculative-Grade Debt

• Published in: Financial management Vol. 22; no. 3; pp. 132 - 141
• Main Authors: Shulman, Joel, Bayless, Mark, Price, Kelly
• Format: Journal Article
• Language: English
• Published: Albany, N.Y Financial Management Association 01-10-1993; Financial Management Association International; Blackwell Publishing Ltd
• Subjects: Bond markets; Bonds; Cash flow forecasting; Cash flow statements; Coefficients; Convertible bonds; Corporate bonds; Credit risk; Debt restructuring; Default; Default (Finance); Equity; Financial Distress Special Issue; High yield investments; Impacts; Influence; Interest rates; Junk bonds; Loan defaults; Marketing; Price volatility; Prices and rates; Standard deviation; Statistical analysis; Studies; Variables; Volatility; Yield; Yield curves; United States > US
• ISSN: 0046-3892; 1755-053X
• DOI: 10.2307/3665933

Following several years of severe price volatility, market uncertainty and a number of well-publicized controversies, the issuance of high-yield bonds virtually ceased by 1990, suggesting that interest in these instruments might disappear as well. During 1991 and 1992, however, total annual returns on high-yield bonds approached 40%, and with it record new issuances. Analysts and investors alike have since revisited the unresolved questions regarding risk of high-yield bonds and their associated yield spreads. Prior academic interest in the high-yield bond market focused on determining whether these bonds collectively provided adequate compensation for their inherently higher default risk. Such studies were ordinarily conducted using portfolios of bonds rather than individual bonds and typically centered on new issues, not seasoned issues. In this paper, we explore determinants of yield spreads between high-yield bonds and U.S. Treasuries for a sample of individual, seasoned high-yield bonds. We examine the influence of default risk, marketability risk, convertibility and underwriter prestige on yield spreads and offer hypotheses on the importance of these components. Our modeling procedure explores both the impact of default risk and the impact of marketability on yield spreads. We model default risk using a combination of traditional firm-based liquidity and leverage variables computed from publicly available financial statements and stock price information. Variables include: net liquid balance, standard deviation of stock returns, cash from operations and market-to-book ratios. We also use bond information such as issue size and convertibility in assessing default risk, and use frequency of bond trades and volatility of bond prices to assess marketability risk. We compiled our sample using the set of all 838 industrial firms identified in Standard & Poor's COMPUSTAT PC Plus which had bonds rated below BBB- in any quarter during the study period 1980 through 1992. We then collected all available bond price, stock price and financial statement data on these firms from a variety of sources. We eliminated any firm and bond from the final dataset which had incomplete data since our factor analytic modeling procedure (LISREL) requires complete data on every bond we use in the study. As a consequence of using this methodology (which mitigates several econometric difficulties inherent in other econometric techniques), we reduced our final dataset to 107 bonds representing 78 firms. Our results support prior academic findings which associate greater default risk with larger yield spreads. In addition, after controlling for the impact of default, we find that our marketability measures contribute to the explanation of yield spreads. For example, yield spreads narrow when bond trading frequency increases, and yield spreads tend to increase when bond price volatility increases. We interpret these results as evidence that investors demand compensation for reductions in marketability. Since speculative-grade debt probably exhibits more pronounced differences in default and marketability risk than other debt, relationships uncovered in this paper may contribute to a greater understanding of yield spreads for all debt. In particular, we anticipate that marketability risk, which has received little attention in prior empirical research, will become more important in future work regarding yield spreads. Moreover, we believe additional refinements of our model criteria might potentially benefit practitioners and academics attempting to price risky debt. For example, aspects of this paper could assist bond analysts in deciding which variables to collect, monitor and model. More importantly, we believe that extensions of our model might lead to a better understanding of relative bond price efficiency and mispricing.