Graduation Year

2026

Date of Submission

12-2025

Document Type

Open Access Senior Thesis

Degree Name

Bachelor of Arts

Department

Economics-Accounting

Reader 1

George Batta

Abstract

This study investigates whether perceived auditor independence, proxied through proportions of non-audit service fees and client dependence of the auditor, is associated with the outcome of debt contracts. While prior research examines how these proxies relate to auditor independence itself, and other strands of literature assess how audit quality influences debt contracting, few studies evaluate whether lenders interpret auditor independence signals in ways consistent with auditors. This study aims to contribute to the broader literature by examining the auditor independence proxies and its direct relationship to debt contract structure holistically. Two OLS models were employed to investigate the relationship between independence proxies and the cost of debt as well as the number of financial covenants employed. Additionally, both models were corrected for heteroskedasticity and industry fixed effects. Lastly, a logit model is run on a binary variable for the presence of financial covenants with the same independent variables employed previously in order to account for the lack of variance in the financial covenants variable.

The results dictate no statistically significant relationship between either auditor independence proxies and the dependent outcomes. Although the Client_Dependence variable was marginally significant at the 10% level of significance when regressed against the cost of debt, the variable lost significance after employing industry-fixed effects. The coefficients on independent variables lacked precision, and although some directions deviate from those reported in prior literature, their associated t-statistics indicate effects statistically indistinguishable from zero. These null findings should be interpreted cautiously due to several limitations, including the small sample size, potential omitted variable bias, and measurement limitations in variables such as Restatement, which assumes timely disclosure of restated financials. Additionally, the use of syndicated loan data presents its own constraints as they are usually issued by large or public companies, and its standardized cov-lite characteristics do not generalize to other sources of debt financing. Overall, the evidence suggests no consistent pattern in how auditor independence signals translate into debt pricing or covenant structure, highlighting the need for broader samples and refined proxies in future research.

Download

Share

COinS