Which Bonds Are More Expensive? The Cost Differentials by Debt Issue Purpose and the Method of Sale: An Empirical Analysis

Abstract

In this paper, we survey the cost differentials by debt issue purpose and the method of underwriting. We find that cost differentials are a function of the purpose for which the debt was issued, with bond purposes traditionally considered riskier facing higher borrowing costs. However, this effect is not uniform and varies by credit quality portfolios. Moreover, the method of bond sale is an important factor for true interest costs. Overall, the competitive method of sale consistently performs better than the negotiated method of sale in all regression models, even after correcting for self-selection. Guzman and Moldogaziev/Cost Differentials by Debt Issue Purpose and Method of Sale 79sale method matter? Other debt issuance decisions may also be of high importance-debt instrument structure choices, special options, credit enhancements, and payment structures and sources-all will have direct implications for municipal borrowing costs. Previous empirical studies suggest that credit ratings (both enhanced and underlying), call and put provisions, underwriter's method of sale, and a maturity structure of bond series are significant factors in a broad array of determinants that have been found to have a significant effect on municipal borrowing costs (Denison, Yan, and Robbins and Simonsen 2007). Researchers have also found that revenue bond issuers, compared to general obligation (GO) bonds, pay higher interest on their debt (Hopwell and Kaufman 1977;Peng and Brucato 2004;Yusuf and Liu 2008). However, with a few exceptions, these studies tend to focus on a single industry or aggregate the issue purpose types to low-risk and high-risk categories (Peng and Brucato 2004). Consequently, studies that address diverse industries and cost differentials associated with such diversity are rare. We attempt to fill this gap by presenting empirical evidence on cost differentials by debt issue industry in California and by discussing management decision choices on how to keep municipal debt issue costs as low as possible. We distinguish between debt industries issued for the purposes of higher and secondary education, development, public facilities, housing, utilities, environmental facilities, health care, transportation, and other miscellaneous purposes. In addition to issue purpose, we pay attention to the sources of debt repayment to account for the differences in interest costs based on perceived default probabilities by market participants. We use controls for market-related and issue/issuer specific characteristics. 1 We begin our paper by discussing the background rationale and the relevant literature from the field of state and local government debt finance. We then describe characteristics of the data that we use in this empirical research and carefully examine outcome and explanatory variables. In the discussion of variables, embedded are the findings from current municipal finance literature that provide theoretical and empirical justifications for the variables in the regression models. After the discussion of vari...

The Impact of Dodd–Frank on True Interest Cost of Municipal Bonds: Evidence From California

“…The empirical findings indicate that TICs on limited tax obligation, public enterprise revenue and revenue bonds are significantly lower than those of GO bonds, which is an unusual result. Other studies of California municipal debt, however, provide similar findings (Guzman and Moldogaziev ). Perhaps, essential service revenue bonds are not viewed as significantly riskier in California.…”

Public Borrowing for Private Organizations: Costs and Structure of Tax‐Exempt Debt Through Conduit Issuers

“…The estimating equation takes the form of traditional issuance cost and interest cost models following recent scholarship (Ely and Calabrese ; Guzman and Moldogaziev ; Robbins and Simonsen, ), with costs explained by issue, issuer, and market characteristics as follows: where Issue is a vector of variables representing characteristics of the bond issue that influence costs including the issue size, time to final maturity, credit rating (natural rating when unenhanced or enhanced rating), refunding status, method of sale, use of credit enhancement, whether an issue is taxable or a tax credit bond (Build America Bond, Qualified School Construction Bond, or Qualified Zone Academy Bond), and multiple measures of issue structure, among others (see Table ). We choose not to include controls for the remaining credit‐backing types in the primary analyses to estimate a general difference in costs between conduit revenue bonds with private borrowers and the remaining municipal bond market.…”

“…Guzman and Moldogaziev () identify statistically significant, and practically meaningful, differences between issuers choosing to report TIC rather than NIC in their California sample. Our broader sample also has significant differences between the type of interest rate reported by issuers and characteristics of those issues, but we find far fewer differences among the private borrowers' conduit revenue bonds in our sample and the type of interest rate reported.…”

Impact of Bankruptcy Eligibility Requirements and Statutory Liens on Borrowing Costs

“…It is generally expected that a limited GO pledge has a narrower set of payment sources than the unlimited variety, and is therefore considered riskier (Butler, Fauver, and Mortal 2009;Peng and Brucato 2004). Similarly, issues with longer maturities, inferior credit quality, and callable bonds have been found to be positively associated with bond yields and issuer costs (Butler et al, 2009;Guzman and Moldogaziev 2012). Consistent with the existing literature, we include a dummy variable if interest on the bonds was subject to federal income tax and control for state tax regimes by including the marginal income tax rates in each state at the time of the bond issue.…”

“…Therefore, yields are expected to be higher on negotiated sales (see Butler et al, 2009;Robbins and Simonsen 2007). Special district bonds are issued for both riskier business-type activities of municipal governments and for narrow special type municipal jurisdictions, and as a result will be positively related with yields (Fairchild and Koch 1998;Guzman and Moldogaziev 2012;Peng and Brucato 2004).…”