Working Papers

[1] Patents as Collateral: How Disclosure Shapes Lending and Innovation (job market paper)

Abstract: This paper examines how expanded peer patent disclosure affects the use of patents as collateral and how, in turn, lenders' collateral preferences shape firms' patenting decisions. I use the American Inventor's Protection Act of 1999 (AIPA), which requires the United States Patent and Trademark Office (USPTO) to publish patent disclosures 18 months after filing, as a plausibly exogenous increase in patent disclosure. After the AIPA, firms that are more dependent on external finance and hold larger patent portfolios pledge more patents as collateral, both granted patents and pre-grant applications, than other patenting firms. Conditional on pledging, these firms pledge more patents in technology classes with many comparable patents, and this comparability matters most for less experienced lenders' collateral decisions. Finally, the same finance-dependent firms subsequently file patents that are more comparable and less novel, consistent with firms tilting their patenting decisions toward more pledgeable but lower-novelty patents in response to lender preferences. The results suggest that disclosure facilitates patent-backed lending and, through this channel, allows lenders' collateral preferences to influence the direction and risk profile of firm innovation.

• Solo-authored

• Dissertation committee: Sugata Roychowdhury (Chair), Beverly Walther, Regina Wittenberg Moerman, Efraim Benmelech

• Presentation: Northwestern University, 2026 Wharton Innovation Doctoral Symposium, 2026 WashU Dopuch Accounting Conference PhD Poster Session (scheduled)

[2] Talking to hire: How small firms pave their way to hire when large firms lay off

Abstract: We study how firms change their hiring and disclosure strategies when they face greater hiring opportunities. In the setting of the technology industry’s massive layoffs, we find evidence that small firms signal increased hiring intent and enhance their information environment when large firms in the same industry engage in layoffs. Specifically, small firms post more jobs following large firms’ downsizing, with the increase primarily concentrated in roles related to information technology. Small firms with greater hiring intent further increase the length of their job postings to provide more information to their prospective employees. Using state-of-the-art machine learning technology, we analyze the content of job postings and find that small firms’ job postings following layoffs contain more information about the company and the job description, as well as more specific keywords related to corporate culture and information technology. Using LinkedIn data, we present preliminary evidence of small firms’ strategies being successful, as we observe a disproportionate flow of employees towards small firms. In additional analysis, we find evidence of small firms engaging in more innovative activities following the layoffs, consistent with human capital being an important driver of innovation. Overall, our paper suggests that firms traditionally at a disadvantage in attracting top talent respond to large-scale layoffs as an opportunity to grow their workforce. Further, firms seeking to hire during layoffs enhance the informational content of job postings to attract and inform prospective employees during periods of increased labor supply.

• With Jung Min Kim and Sugata Roychowdhury

• R&R at Journal of Accounting and Economics

• Developed from second-year summer paper

• Presentation: Northwestern University, Connecticut University, 2024 Yale Summer Accounting Conference, 2025 Hawai’i Accounting Research Conference, 2025 UIUC Young Scholars Research Symposium, 2025 Labor and Accounting Group Conference, 2025 University of Michigan Harvey E. Kapnick Accounting Conference

[3] Net zero coefficients in accounting research

Abstract: A surprising number of empirical accounting studies report regression results that include two coefficients that are economically large, statistically significant, and roughly equal in magnitude with opposite signs. Searching five leading journals from 1996 through 2025, we find the pattern in 9 percent of articles. Building on Kalnins (2018), we derive a set of conditions under which these opposite-signed coefficients, which we call net-zero coefficients, may be an econometric artifact rather than a true finding. In the typical case we study, each regressor is positively correlated with the dependent variable on its own, but the sign of one of the coefficients flips and the economic magnitude of both coefficients increases when the two variables are jointly included in the regression. We show that this occurs when the two regressors are highly correlated with each other but are differentially correlated with the dependent variable. We extend the framework to interacted regression designs and show that the net-zero coefficients between the interacted term and the main effect can also be spuriously generated by data features that do not genuinely reflect the underlying structural relationship. We demonstrate our findings with both real and simulated data. Finally, we propose a five-step diagnostic that researchers can use to help identify potential Type I errors and inflated magnitudes when such patterns emerge in their research.

• With Andrew Leone and Shibao Liu

• Presentation: Northwestern University

Selected Work In Progress

[4] Shifting Liability and the Allocation of Accounting Expertise: Evidence from the Rejection of the Audit Interference Rule

Executive Summary: We examine how auditors and their clients respond to changes in the allocation of litigation liability between them. Our setting exploits state-level rejections of the Audit Interference Rule (AIR), which allow auditors to more broadly invoke client negligence in allocating fault and thereby reduce auditors’ liability exposure relative to that of their clients. Empirically, we rely on a stacked difference-in-differences research design around state-level AIR rejections, examining responses at both the client-firm and audit-office levels. Across both levels, we include unit-by-AIR-event and AIR-event-by-year fixed effects. We document complementary responses on both sides of the audit market. Following AIR rejection, affected firms pay higher audit fees and expand their internal accounting workforce, with firms facing greater litigation risk hiring more experienced accounting personnel. Firms initially audited by non-Big N auditors also become more likely to switch to Big N auditors. At the same time, Big N audit offices exposed to AIR rejection expand their client portfolios, accept more high-risk clients, and increase their overall exposure to high-risk clients in rejecting states. We find no comparable changes among non-Big N auditors, consistent with litigation exposure having constrained reputable auditors’ willingness to serve risky clients prior to AIR rejection. Collectively, the evidence suggests that shifting liability from auditors toward clients induces firms to invest in both internal accounting expertise and higher-reputation external auditors, while reduced auditor liability makes risky clients more acceptable to Big N auditors. Our findings demonstrate that the allocation of litigation liability shapes not only firms’ investments in financial reporting expertise but also the matching of clients with auditors and, more broadly, the allocation of accounting expertise across firms.

• With Xiumin Martin and Sugata Roychowdhury