PDF Ebook Treasury yields and corporate bond yield spreads: An empirical analysis
This paper empirically examines the relation between the Treasury term structure and spreads of corporate bond yields over Treasuries. This relation is of interest in its own right because it is essential to the calibration and testing of models that price credit sensitive instruments. More broadly, this investigation allows us to address the nature of the biases in commonly-used indexes of corporate bond yields; biases induced by the way these indexes are constructed.
Indexes of corporate bond yields are typically averages of yields on a set of bonds that are chosen based on the characteristics of the bond at the time the average is computed. For example, today's Moody's Industrials Aaa-rated Bond Yield is today's mean yield on a set of Aaa-rated bonds issued by industrial rms that have current prices not too far away from par and have relatively long remaining maturities. Much academic work interprets changes in yield spreads constructed with such indexes as indicative of changes in default risk. This interpretation is subject to a number of potential problems, two of which are likely more important than others.
The rst major problem, as discussed by Du e and Singleton (1995) in the context of indexes of new-issue swap spreads, is that these are \refreshed" indexes. The change in the yield from one period to the next does not measure the change in the mean yield on a xed set of bonds, but rather the change in the mean yield on two sequential sets of bonds that share the same features; in particular, that share the same credit rating. Hence using these indexes to investigate intertemporal changes in bonds' default risk is problematic because the indexes hold constant a measure of credit quality.
The second major problem is that corporate bond yield indexes are often constructed using both callable and noncallable bonds. Given the objectives of those constructing the indexes, such as Moody's, this is sensible. Until relatively recently, few corporations issued noncallable bonds, hence an index designed to measure the yield on a typical corporate bond would have to be constructed primarily with callable bonds. However, yields on callable bonds will be a ected by the value of the option to call. Variations over time in yield spreads on callable bonds will re ect, in part, variations in the option value.
Other, less important problems that a ect most indexes of corporate bond yields include coupon-induced changes in duration and state taxes. Given all of these biases, how appropriate is the assumption that changes in yield spreads for such indexes are driven by changes in default risk?
I use month-end data on individual investment-grade bonds included in Lehman Brothers Bond Indexes from January 1985 through March 1995 to examine how yield spreads vary with changes in the level and slope of the Treasury term structure. To illustrate the results discussed in this paper, consider the average month-t yield spread (over the appropriate Treasury instrument) for a group of noncallable coupon bonds with identical credit ratings and similar maturities. If the Treasury yield curve shifts down by 10 basis points between months t and t+1, the average yield spread on this group of bonds rises by between 0.6 and 3.6 basis points. In other words, the price increase in a corporate bond that accompanies a given decrease in Treasury yields is between 6 and 36 percent less than it would be if bond yields and Treasury yields moved one-for-one. The responsiveness of the spread is weak for high-rated bonds (e.g., it is statistically insigni cant for Aaa-rated bonds) and strong for low-rated bonds. Although these results are for coupon bonds, any coupon induced bias in these results is quite small.
I also conclude that, after adjusting for coupon-induced biases, changes in the slope of the term structure of Treasury yields are very weakly negatively related to changes in yield spreads. Surprisingly, although both yields on Treasury bonds and corporate bond yield spreads are linked to the business cycle (well-known facts that I con rm here), I nd it di cult to explain the relation between yield spreads and Treasury yields in terms of variations in default risk driven by the business cycle.
I nd that, compared with the results discussed above, changes in yield spreads constructed with refreshed indexes of noncallable bonds have much weaker links to changes in Treasury yields. This evidence indicates that refreshed indexes underreact to variations in credit quality because changes in credit ratings capture part of the variability of yield spreads over time.
By contrast, changes in yield spreads constructed with indexes of callable bonds have much stronger links to changes in Treasury yields because the value of the call option negatively varies with Treasury yields. For high-priced callable bonds, most of the variation in yield spreads is caused by variations in the value of the call option.
The next section discusses why the relation between yield spreads and Treasury yields is important and reviews some previous research. The third section describes the database I use. The fourth section analyzes the relation between yield spreads on noncallable bonds and Treasury yields. It also discusses possible biases induced by coupons and state taxes. The fth section attempts to interpret this relation in terms of default risk and the business cycle. The sixth section explores the behavior of yield spreads constructed with both refreshed indexes of noncallable bond yields and indexes of callable bond yields. Concluding comments are contained in the nal section.
Download
PDF Ebook Treasury yields and corporate bond yield spreads: An empirical analysis
Posted in :
