Working papers
Existing literature assumes that the prices of off-balance sheet securities, such as derivatives, are unaffected by the balance sheet constraints of financial intermediaries. This paper examines the swaption market and demonstrates that the spread between underlying swap volatility and swaption-implied volatility increased significantly following the Global Financial Crisis. Moreover, this spread co-varies with violations of covered interest rate parity, which serves as a proxy for the shadow cost of banks’ balance sheet constraints. Our analysis reveals that swaption prices are influenced by demand from fixed-income hedge funds, which is, in turn, impacted by the balance sheet constraints of their prime brokers. Our results indicate that derivative prices are affected by intermediary balance sheet constraints when the derivative is used to hedge an on-balance sheet asset. Additionally, financial intermediaries can propagate their balance sheet constraints more widely through the economy due to their role as prime brokers to diverse financial institutions, including leveraged hedge funds.
Stock Return Predictability of Volatility Spreads and Abnormal Turnover
Despite the effectiveness of volatility risk premium (VRP) in predicting stock returns, measuring it is challenging because of issues with estimating expected realized volatility. This paper examines how imprecise measurement of expected realized volatility affects the predictive power of individual equity VRP, captured by realized-implied volatility spread, for cross-sectional stock returns. We show that abnormal stock turnover proxies for an increase in noise trading, which results in poor realized volatility estimates and contaminates the predictive power of volatility spreads. Applying a mean reversion correction on realized volatility rectifies this problem and improves the stock return predictability significantly, increasing the returns to trading strategies based on individual equity VRP by 42% on average.
Volatility Extrapolation and Corporate Investments
The Great Financial Crisis (GFC) has led to a slow recovery in corporate investment despite unusually easy financing conditions. In this paper, we propose a new explanation for this phenomenon. We show that large volatility shocks combined with extrapolative expectations impact managers' perception of risk, resulting in lower investment well into the future. We use a differences-in-differences strategy to compare the investment of firms that faced a large vs. small shock and show that large shocks change a firm's perception of risk and reduce investment for several years after the shock. Our findings are robust to controlling for alternative explanations, such as increased financial constraints.
Published papers
Zombie Lending due to the fear of Fire Sales
Journal of Corporate Finance, 2025
This paper provides evidence of a new cost of fire sales: zombie lending by banks. Banks with high market share are more likely to internalize the negative spillovers of falling collateral prices during a fire sale. To prevent prices from falling further during a fire sale, these banks do not liquidate defaulted firms and instead give zombie loans to keep them alive. Using structural breaks in real estate prices to identify periods of fire sales in different MSAs, we provide evidence that banks with high market share give zombie loans to firms with relatively higher real estate assets during a fire sale. Further, congestion due to zombie firms in an industry reduces the investment and profitability of healthier firms. Overall, we highlight a new mechanism for zombie lending resulting from reduced collateral liquidation in markets prone to fire sales.
Journal of Finance, 2024
In statistics, samples are drawn from a population in a data-generating process (DGP). Standard errors measure the uncertainty in estimates of population parameters. In science, evidence is generated to test hypotheses in an evidence-generating process (EGP). We claim that EGP variation across researchers adds uncertainty—nonstandard errors (NSEs). We study NSEs by letting 164 teams test the same hypotheses on the same data. NSEs turn out to be sizable, but smaller for more reproducible or higher rated research. Adding peer-review stages reduces NSEs. We further find that this type of uncertainty is underestimated by participants.