Academic journal article Quarterly Journal of Business and Economics

Bond Ratings and New Issue Municipal Bond Pricing: Path Analysis Results

Academic journal article Quarterly Journal of Business and Economics

Bond Ratings and New Issue Municipal Bond Pricing: Path Analysis Results

Article excerpt

Bond Ratings and New Issue Municipal Bond Pricing: Path Analysis Results

Abstract

Although research on municipal debt pricing has included both the bond

rating and economic information as exogenous variables, other

information is often of little value to investors due to lack of both

timeliness and reliability. This study employs path analysis to resolve

the role of the bond rating in the context of such economic data. The

results suggest that the bond rating assumes considerable importance in

the pricing of municipal debt.

Introduction

Bond pricing research and the comments of practitioners attest to the importance of the bond rating as a means of measuring credit quality both for corporate and municipal debt issuers. Ingram, Brooks, and Copeland (1983) suggest that the rating takes on particular importance in the pricing of municipal debt due to the lack of timely, reliable public information in some standardized format for municipalities. Robbins, Apostolou, and Strawser (1985) conclude that the financial statements of municipalities are inadequate when considered in the context of investor needs. Thus, investors may rely heavily on the bond ratings in lieu of economic data if the cost of securing and interpreting those data is excessive.

Although Liu and Thakor (1984) provide a model for examining this relationship between economic information and municipal bond pricing, their methodology is limited in its ability to measure the marginal effects. It is unclear from the Liu and Thakor analysis specifically how the economic variables affect new issue yields, given the existence of bond ratings. As Cranford and Stover (1988) observe, the Liu and Thakor model examines merely the total effects of economic information rather than deciphering the direct effect and the indirect effect via rating. Therefore, an empirical model is required to decompose the total relationship between economic information and new issue debt pricing into direct and indirect effects. The indirect effects are those influences of the exogenous variables on new issue debt pricing through the formers' respective effects on the bond ratings given to the new issues.

Path analysis (Goldberg and Duncan, 1973) is employed to quantify this distinction between the total effect of economic variables on debt pricing and their indirect effects via the bond rating. Based on two separate samples of initial municipal bond offerings, the path analysis results provide a more complete measure of the relationship between economic information and new issue bond pricing. The goal of path analysis is to provide plausible explanations of observed correlations by constructing a model of cause-and-effect relations among the variables. The observed relationships then are decomposed into a sum of path coefficient terms representing the direct and indirect paths. This explicit measure of indirect effect is missing from the Liu and Thakor's (1984) analysis.

In particular, the model quantifies the relative roles of economic information and bond rating. Based on the results, it is suggested that state administrators and the investment community could rely extensively on the bond rating in examining new issue municipal debt pricing. Thus, administrators could place more emphasis on ratings as a means of avoiding average quality pricing due to an adverse selection problem. The investment community could reduce its search costs. The direct effects of economic variables specifically on new issue pricing are relatively inconsequential in the pricing of municipal debt new issues.

Empirical Methodology

General Path Model

Based on m economic variables and n bonds, the Liu and Thakor model is stated as:

(1) [Mathematical Expression Omitted] (2) [Mathematical Expression Omitted] where:

[Mathematical Expression Omitted] = an n dimensional column vector of ones; R = an n dimensional vector of ratings; Z = an n x m matrix of economic variables; Y = an n dimensional vector of yields; [[Lambda], [[Alpha], and [[Sigma] = scalar coefficients; [Pi] and [Beta] = m dimensional column vectors of regressio n coefficients; and [Xi] and [Epsilon] = n dimensional column vectors of stochastic error terms. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.