Publication:
Re-doing the Audit, with Isabel J. Cho and Clive Lennox. Journal of Accounting and Economics, 2026, 82 (2)
Abstract: This study investigates the practice of re-audits—where an incoming audit firm re-audits the prior year’s financial statements previously examined by a predecessor audit firm. We examine the costs and benefits of such re-audits. We find that re-audits are more likely when the risks of misstatement are high and when incoming auditors can expect to win more clients by identifying material misstatements overlooked by predecessor auditors. Our results show that incoming auditors who perform re-audits are more successful in winning new clients from predecessor auditors. Moreover, re-audits result in significantly more restatements. However, consistent with the resource-intensive nature of re-audits, we also find that they are associated with longer audit delays and higher audit fees. Overall, we conclude that re-audits have important consequences for financial statement users, incoming auditors, and predecessor auditors.
Presented at the AAA Western Region Doctoral & Faculty Interchange, Egyptian Online Seminars in Business, Accounting and Economics*, Fox and Haskayne Accounting Conference*, Hong Kong University of Science and Technology*, International Symposium on Audit Research, University of Alberta Accounting Research Conference*, University of Alabama*, and University of Southern California
Received the Best Paper Award at the Fox and Haskayne Conference
Working Papers:
To Hide or Highlight? How Borrowing Firms Summarize Their Loan Contracts (Dissertation)
Abstract: Loan announcements in Form 8-Ks typically include the full contract as an exhibit and a summary in the main file. This study examines how firms construct these summaries, which contain no incremental information beyond the underlying contracts and therefore should be irrelevant absent information processing costs. I find that firms are more likely to omit covenant details from summaries when the likelihood of covenant violation is high, consistent with incentives to downplay credit risks and operational restrictions. This tendency is stronger when investors face higher costs of processing the underlying contracts. Less detailed summaries also predict a higher likelihood of future covenant violations and lower profitability, suggesting that summarization choices themselves convey information. Using SEC log files, I further show that investors respond to less detailed summaries by consulting the full contracts, but only when omissions are readily detectable. Overall, the findings demonstrate the strategic use of loan summaries and highlight the importance of information processing costs, as disclosures that provide no new information can nonetheless reflect deliberate reporting choices and shape investor behavior.
Committee: Clive Lennox (co-chair), Regina Wittenberg-Moerman (co-chair), Maria Ogneva, Lorien Stice-Lawrence, and Odilon Câmara
Presented at AAA/Deloitte Foundation/J. Michael Cook Doctoral Consortium, AAA Global Connect, CAPANA Annual Research Conference, and University of Southern California
Received the Best Paper Award at the CAPANA Annual Research Conference
Lender-borrower Interactions and Loan Contracting, with Allen Huang and Xinlei Li
Abstract: In commercial lending, lenders interact with borrowers to source new loans and assess credit risk. Using proprietary data from a bank, we study what drives lender–borrower interactions and how they shape loan contracting. Interactions increase with borrowers’ business prospects and risks, consistent with prospecting and credit assessment motives. Although interaction-derived information is not incorporated into the bank’s internal risk ratings, loans initiated after interactions are priced more efficiently, with interest rates that better predict future defaults, especially when interactions are more likely to generate new information. Loan officers use interaction-derived information to approve loans that would otherwise be rejected, but not to reject loans that would otherwise be approved; this asymmetry is more pronounced when their prospecting incentives are stronger. These findings highlight how loan officers collect and use interaction-derived information with their prospecting and screening incentives and have implications for research on loan officer incentives and the role of information in credit decisions.
Presented at the AAA Annual Meeting, Chinese University of Hong Kong*, City University of Hong Kong*, EAA Annual Congress*, Fudan University*, Harbin Institute of Technology Shenzhen*, Hawai'i Accounting Research Conference*, National Taiwan University*, Peking University*, Southern University of Science and Technology*, Tsinghua University*, University of Illinois Chicago*, UC Davis*, UC Irvine*, and the Eleventh International Conference of the Journal of International Accounting Research
Work in Progress
Auditor Responses to Financial Crises, with Isabel J. Cho and Clive Lennox
Presented at University of Southern California
* Presented by co-author.