Marketability and default influences on the yield premia of speculative-grade debt - includes appendices - Financial Distress Special Issue

Financial Management (Financial Management Association), Autumn, 1993 by Joel Shulman, Mark Bayless, Kelly Price

Following several years of severe price volatility, market uncertainty and several well-publicized controversies and improprieties, the number of new issue high-yield bonds virtually disappeared by 1990. However, declining interest rates during 1991-1993 coupled with a recovering U.S. economy have enabled the high-yield bond market to report total annual returns approaching 40%, and with it, record new issuances. Not surprisingly, interest in high-yield bonds has been rekindled, and investors and analysts alike have revisited unresolved questions regarding risk and associated yield spreads.

Much academic interest in the high-yield bond market has centered on determining whether these bonds collectively provide adequate compensation for their inherent default risk.(1) Other studies have focused on risks other than default. For example, Fisher |12~ was the first to explain yields on fixed-income securities in terms of three major characteristics: (i) the risk of default; (ii) sensitivity to interest rate changes, i.e., interest rate risk; and (iii) the risk that the holder will be unable to sell the instrument (marketability risk).

In this paper, we examine the determinants of yields, relative to Treasuries, for a sample of individual, seasoned high-yield bonds. We compare the impact of default on yield spreads for individual high-yield bonds by using a default model which produces an expected probability of default for each bond. We then model default risk, along with marketability risk and other characteristics on yield spreads using a factor analytic technique (LISREL).

Our empirical results indicate that both marketability risk and default risk are important in explaining yield spreads for high-yield bonds. Additional refinements in the criteria discussed in this paper should benefit practitioners and academics attempting to price high-yield bonds, and might lead to a more rigorous analysis of marketability, which has received little attention in the past.

Our paper has four sections. Section I provides motivation for the variables included in the yield spread model, and Section II describes the data. Section III describes the factor analytic technique (LISREL model) used to assess default and yield spreads. The final section presents our results and conclusions.

I. An Empirical Model of Yield Spreads

We employ a two-step procedure in exploring the determinants of yield spreads for high-yield bonds. The first step estimates default risk in the traditional spirit of default risk methodology. The default model summarizes publicly available information that would influence investors' expectations that a particular firm will default. The second step explains yield spreads as a function of default, calculated by the default model in step one, marketability, and other characteristics including convertibility.(2) We minimize differences related to interest rate risk by calculating yield spreads for risky bonds and Treasury bonds that are of equal Macaulay duration.

A. The Default Model

Previous research suggests a large number of factors which potentially affect the probability of default. These include profitability, liquidity, leverage and market assessment ratios. Reflecting the bankruptcy literature, we selected ratios which represent each of these factors. We selected only those variables which had demonstrated a significant relationship in prior empirical research. For example, we selected cash from operations as our profitability measure, net liquid balance as our corporate liquidity measure, market value of stock to book debt (market to book) as our leverage measure, and standard deviation of stock returns as our market assessment measure. We also added several new measures to the model to isolate other differences among companies issuing high-yield bonds. This variables include amount outstanding, frequency of trades, volatility of bond prices, convertibility, and underwriter prestige. A brief discussion of each variable follows.

1. Corporate Liquidity and Cash Flows

Studies by Beaver |7~, Altman |2~, |5~ and others document an inverse relationship between corporate liquidity, corporate cash flows and the probability of default. We assess liquidity using net liquid balance (Dambolena and Shulman |11~) and assess corporate cash flows using cash from operations (Gentry, Newbold, and Whitford |15~). Net liquid balance is the difference between all liquid financial assets and all callable liabilities. As the firm's level of liquid financial assets increases relative to its callable liabilities, corporate liquidity improves. We measure liquid financial assets as cash and marketable securities, and we measure callable liabilities as short-term notes payable and current maturity due on long-term debt. To facilitate interfirm comparisons, the difference between liquid financial assets and callable liabilities is scaled by total assets. A positive net-liquid-balance-to-total-assets ratio, NLB, indicates a surplus of cash over callable debt, while a negative NLB indicates a shortfall of cash relative to callable debt. Thus, NLB is directly related to liquidity and, therefore, inversely related to the probability of default.