Municipal bonds are used by cities in the United States to raise funds for public projects. Cities are concerned with how they can improve their credit ratings so that they can borrow funds at lower cost. In this study, we examine the determinants of municipal bond ratings for three rating agencies, which are Moody's, Standard and Poor's, and Fitch Ratings. We construct predictive models that associate cities' characteristics with bond ratings. Our goal is to understand what cities need to improve so that they can achieve better credit ratings. As a basis for this research, we provide an understanding of ordered multinomial logistic regression theory. This regression method also leads us to other topics in mathematics such as maximum likelihood estimation and the Newton-Raphson iteration method. This independent study project focuses on learning about the determinants of municipal bonds' credit ratings and how to measure the effects of these determinants on the ratings. We learn about maximum likelihood estimation and the Newton Raphson method to understand how ordered multinomial logistic regression works, and apply this knowledge to build models that can answer our questions.


Hartman, James

Second Advisor

Burnell, James


Business Economics; Mathematics


Municipal Bonds, Credit Ratings, Logistic Regression, Ordered Multinomial Logistic Regression

Publication Date


Degree Granted

Bachelor of Arts

Document Type

Senior Independent Study Thesis



© Copyright 2018 Thao Nguyen