"Why Firms Go Dark: Evidence from Leaked Covert Political Ties," with Omrane Guedhami, April Knill, and Baixiao Liu.

[Abstract]: Using three disclosures of donor lists from U.S. “dark money” groups, we compare firms whose covert political donations were exposed to similar non-exposed firms to shed light on why corporations establish covert political ties. Our findings reveal that, on average, exposed firms experience a 2.7% increase in Tobin’s Q following exposure. However, firms exposed as donors to dark money groups that advocate for a rival candidate’s defeat observe a 17.8% decrease in Tobin’s Q and lower future operating performance, in line with negative reputation effects. These negative effects are particularly pronounced when the candidate supported by the dark money group loses the election and when contributions to the group do not align with the firm’s publicly disclosed political contributions. Additional analysis shows that firms adjust their corporate political disclosure activities in response to exposures. Overall, our findings suggest that firms contribute politically through dark money groups to avoid reputational damage due to associations with “dirty politics,” retaliation by rival candidates, and inconsistency with established political ideologies.

"Now You See Me:  Broadband Internet, Social Media,  and Corporate Social Responsibility," with Hosein Maleki and Todd Gormley.

[Abstract]: Our study examines the impact of mobile broadband internet and social media activity on firms' corporate social responsibility (CSR) policies. We use the expansion of 3G networks in US counties as a natural experiment and find that the availability of 3G around a firm's headquarters is associated with a significant improvement in their CSR ratings. Reputation concerns, driven by the possibility of online activism and social media backlash of corporate policies, are a major driver of the results. The effect becomes stronger for firms that receive the most attention from the media. Further, firms' CSR ratings improve not only if 3G is available around the headquarters but also in other locations strongly linked to the firm's domicile county. Higher CSR ratings happen concurrently with real social and economic improvements, such as a significant reduction in toxic chemical release after 3G networks become available around manufacturing facilities.

"Institutional Ownership and Investment Activity of Firms: An Examination of Real Estate Investment Trusts," with Prashant Das, Frank Gyamfi-Yeboah, and Philip Seagraves. (under review).

[Abstract]: Prior research has examined the impact of institutional ownership on firms' decision-making, such as asset expansion. We explore whether institutional investors influence investment strategy and whether it has a material association with firm performance. Our study underscores a dynamic link between equity ownership structure and managerial focus on asset expansion strategies. We examine a large sample of real estate investment trusts (REITs) that adopt two strategies to acquire new assets: (1) developing new assets themselves or (2) acquiring existing assets.  Institutional investors have held a majority of shares in REITs. We find that ownership by long-term institutional investors is generally associated with significantly enhanced development activity. In recent years, investment companies have shown a preference for ground-up asset development, while insurance firms have shown a preference for acquiring existing assets. Compared to acquisition, development is associated with superior firm performance, and firms with higher acquisition levels tend to disclose less.

“Agency Cost of Choosing Green,” with Mariya Letdin, Chongyu Wang, and Tingyu Zhou.

[Abstract]: This study examines the financial implications of corporate green-certified headquarters, evaluating whether such investments align with enlightened value maximization or represent agency costs detrimental to shareholder value. Our findings reveal a statistically significant decline in firm value following green-certified occupancy, with the effect being more pronounced for firms occupying both LEED- and Energy Star-certified buildings. Temporal analysis confirms that valuation declines occur post-certification, ruling out reverse causality. We further show that institutional ownership and stronger governance amplify the negative impact, suggesting that green-certified headquarters are perceived as managerial excess rather than value-enhancing investments. Stock market analysis indicates that firms moving into green buildings underperform their counterparts, with a trading strategy of shorting such firms yielding positive abnormal returns.  Our results provide empirical evidence that green-certified headquarters function as indirect managerial compensation, imposing agency costs that diminish shareholder value. These findings challenge the assumption that corporate environmental investments necessarily contribute to long-term financial performance, emphasizing the importance of governance.

“Optimizing Real Estate Portfolios: The Role of AI in Geographic Diversification,” with Timothy Dombrowski.

[Abstract]: This study investigates the data analysis capabilities of GPT-4o in real estate portfolio selection by integrating predictive modeling, model evaluation, and investment decision-making into a fully autonomous AI-driven framework. Unlike the earliest large language models (LLMs) that primarily process textual data or recent LLMs such as OpenAI's o1 and DeepSeek's R1, which are designed for complex reasoning, GPT-4o actively executes code and conducts quantitative analysis using the Code Interpreter tool. Leveraging a dataset of Zillow home price data and several predictive factors, the AI-generated portfolios achieved three of the four top-performing Sharpe ratios in our out-of-sample backtest. Further, we find that data obfuscation -- removing city names, states, and dates -- reduces geographic diversification and produces lower Sharpe ratios than the unobfuscated portfolios. Overall, our findings highlight the potential of generative AI in advancing data-driven portfolio management.

“After the Storm: Reallocation in Commercial Real Estate,” with Meagan McCollum and Jason Walter.

[Abstract]: Exposure to natural disaster risk is an important consideration for commercial real estate market participants. In the past 30 years, in the state of Florida alone, the total estimated economic loss from hurricanes is over $250 billion, which doesn't directly consider insured losses that commercial properties may face. While Florida has adopted many new policies, such as revisions to building codes, changes to insurance regulations, and incentives to develop greater resiliency throughout the physical building stock, our knowledge is limited regarding how commercial properties, in aggregate, are impacted by disaster risk. This study examines three major dimensions of possible reallocation in the FL CRE market from 1995 to 2023. Our findings reveal that hurricanes significantly increase the likelihood of use code changes in commercial properties, with a more pronounced effect in higher-category storms. Additionally, the impact on market dynamics, such as sales price and volume, is significant, with a marked decrease in both aspects following hurricane events. This trend highlights the market's vulnerability to natural disasters. Finally, we plan to utilize these results to examine how disasters via commercial real estate can impact industry agglomeration in fourteen major Florida MSAs.

“Work-From-Home, Task-Switching, and Employee Performance,” with Meagan McCollum, Eric Olson, and Jason Walter.

[Abstract]: This study examines the impact of work-from-home and hybrid schooling on employee performance evaluations during the COVID-19 pandemic. Using a unique dataset from a major U.S. public company, combined with local school district operational data, we analyze how different schooling modes affect employee evaluations. Our theoretical model predicts that hybrid schooling increases task-switching, leading to reduced productivity. Empirical results confirm that employees in hybrid schooling districts experience significantly lower evaluation scores compared to those in in-person settings. Our results highlight that local school districts opting for virtual instruction days can create unintended negative effects on the local workforce, as working parents must balance their professional responsibilities with supervising their children's at-home learning.

“Are Investors Missing the Total Return Forest for the Excess Construction Trees?” with Ryan Chacon, Varsha Jain, and Hans Norby.

[Abstract]: We investigate whether excess construction or middle-class income growth is the more prominent driver of multifamily rent growth in major U.S. metropolitan statistical areas (MSAs) over the past 19 years. Using data from CoStar, we perform a series of horse race tests, using both traditional methods and machine learning methods, that compare the relative importance of net deliveries to the growth of median household income in explaining the growth of effective rents. Our results consistently show that middle-class income growth dominates. Using standardized variables to facilitate comparison, we consistently find that income growth has more than twice the impact on rent growth as net deliveries. These findings suggest that recent concerns about oversupply in the multifamily sector may be overblown and provide opportunities for long-term investors.