Doubly Bound: The Cost of Credit Ratings

Introduction

Introduction 

WHEN LOCAL GOVERNMENTS OVERPAY for financial services, the extra costs must be borne elsewhere. Since cities cannot print money, they must either raise taxes or reduce services to offset extra money paid to bond investors and financial industry providers. One type of financial service is the assignment of credit ratings, which tell bond investors how much risk they incur by purchasing the local government’s obligations. Since US local government debt issues rarely default, the rating process should be straightforward and inexpensive. Instead, the credit rating process imposes substantial direct and indirect costs on issuers, resulting in higher taxes and degraded public services.

In this study, I measure the financial burden of the credit rating system on US municipal bond issuers and propose a less costly alternative rating approach. In addition to the fees rating agencies collect from municipal bond issuers, I also estimate costs municipal bond issuers incur because they are rated more harshly than corporate and structured finance issuers. These costs take the form of higher interest expenses and premiums on municipal bond insurance policies—which municipalities would not need if they were rated on par with corporate bond issuers. I also discuss the connection between low ratings and accelerated swap termination payments that some municipalities have or will incur—although I don’t include this affect in the total.

After measuring the cost of the current rating system, I outline and assess a model-based municipal rating system that generally assigns higher ratings. I find that, despite being more lenient, this system would still have been able to alert investors to the risks of bonds offered by defaulting cities such as Detroit, Michigan; Harrisburg, Pennsylvania; and Stockton, California

A model-based rating system like the one proposed here would reduce municipal financing costs. It would also lead to lower rating fees by allowing greater automation of the credit ratings process. 

This study begins by surveying evidence that municipal bond ratings are indeed more severe than those for other asset classes. With that basis, I then estimate the overall cost of the municipal rating system. Finally, I describe the model based alternative and provide some preliminary results from the proposed rating model.