Green Moral Hazard: Estimating the Financial and Non-financial Impacts of CEO Incentives
Latest Version
Dissertation Committee: Ivan Marinovic (Chair), Maureen McNichols, and Kevin Smith

I develop a novel structural model for analyzing the financial and non-financial implications of CEO compensation contracts that include incentives tied to non-financial performance. By applying this model to green incentives, I find that they motivate CEOs to reduce carbon emission intensity by 1.8% per year but at a financial cost of 1.3% of firm value annually. As green performance is an imperfect signal of CEOs’ actions toward green outcomes, a “green moral hazard” arises: the principal should offer CEOs a premium for the risk added by green incentives. I estimate that this green moral hazard is substantial, accounting for $1.72 million of the total moral hazard cost of $2.05 million. These results suggest that green incentives pose an important economic trade-off: while green incentives can lead to meaningful environmental improvements, they impose substantial costs on the firm.

Presented at: AES Junior Accounting Theory Conference, AAA Doctoral Consortium

Moral Hazard and the Value of Information: A Structural Approach
(with Jeremy Bertomeu and Ivan Marinovic)
R&R at Journal of Accounting Research (JAR)
Executive compensation contracts use information from markets and accounting to elicit efficient incentives. We structurally estimate the contribution of each performance to quantify the relative importance of price versus accounting. For plausible risk-aversion coefficients consistent with the sample, the cost of moral hazard, defined as the risk premium to be paid to elicit incentives, amounts to $4.17 million; further, in counterfactuals, we show that absent reliable accounting performance measures, average compensation increases by approximately one eighth versus more than doubling if price information is unavailable (e.g., a non-public firm). At high risk-aversion levels, dropping accounting or price information would make it infeasible to elicit high effort, with a maximal potential loss in firm value of $6.24 million without accounting and $16.13 million without price information. These results offer a first attempt to quantitatively assess the value of accounting signals in executive contracts.

Learning or Catering? Effect of Managerial Myopia on Disclosure and Investment Feedback
Financial markets can have real effects through managers learning from prices or catering to investors to boost prices. I study how the two incentives shape a myopic manager’s disclosure and investment decisions, in a “feedback effects” model with investors who can acquire private information and a manager who learns from prices when making an investment decision. I provide a novel mechanism through which disclosures can incentivize information acquisition, even though the disclosed signal and the investor’s private signal are substitutes. Moreover, I show that myopic managers disclose high signals, even if doing so reduces information in price. On the contrary, value-maximizing managers resort to discretionary disclosure only when their information quality is low and the investment opportunity is risky.

Deviations from the Law of One Price across Economies
(with Hyeyoon Jung)
In a model with agents facing constraints heterogeneous across economies, we provide a novel explanation for an understudied yet economically significant deviation from the Law of One Price across FX forward markets. Specifically, we document a substantial divergence between the exchange rate for locally traded forward contracts and contracts with the same maturity traded outside the jurisdiction of countries during the global financial crisis, and that the magnitudes varied across currencies. The model predicts that (1) the basis increases with the shadow costs of constraints across time and increases with the country-specific FX position limit across countries; (2) the shadow cost of each constraint non-linearly increases as the intermediary sector’s relative performance declines below a threshold; and (3) higher shadow cost of the position limit predicts lower future excess return on local-currency denominated assets, as buying local assets relaxes the FX position limit constraint imposed on the intermediaries. We test the model predictions and find consistent evidence in the countries with tight position limits.

Disclosure and Stock Market Participation
(with Kevin Smith)
Standard setters often cite investors’ confidence in the stock market and its impact on their desire to participate in this market as key motivations for disclosure regulations. This paper models how the informativeness and complexity of public disclosures influence market participation among unsophisticated investors who face ambiguity about the distribution of firms’ cash flows. In contrast to the common perception, we demonstrate that disclosure has countervailing effects on participation. While it reduces ambiguity, which encourages participation, it can also diminish the equity risk premium, especially during times of market stress, which discourages participation. Consequently, while simpler disclosures generally increase participation, more informative disclosures may have the opposite effect. We explore the implications for investor welfare and the design of optimal disclosure regulation.