Research

[1] Retail Habitats Network (Job Market Paper) [07/2024]

Abstract: Retail investors choose stocks based on their personal preferences, which define their individual habitats. This paper proposes a novel network that unifies the habitats of various retail investors in a parsimonious way. In this network, two stocks are connected if they are frequently searched by the same Internet users, indicating that they likely share the same habitat. Consistent with this interpretation, I demonstrate that firms with similar habitat-forming characteristics -- such as headquarters location or stock price level -- are more likely to be co-searched. The habitats from existing studies explain only a small part of the co-searches, suggesting that the network provides unique information. Using the network, I argue that a retail investor's attention is limited to stocks within their habitat. Consequently, when one firm distracts attention by announcing earnings, other firms in the same habitat receive less retail attention and experience lower abnormal trading, volatility, and excess returns. The results are consistent with retail investors reallocating their limited attention across individual stocks within their habitats.

[2] Risk and Risk-Free Rates (with Mete Kilic and Zhao Zhang) [04/2024] [SSRN]

Abstract: We document that monetary policy shocks sharply change the correlation between financial market risk and the risk-free rate. While interest rates and the VIX index on average comove negatively, consistent with precautionary savings, this relationship turns positive on FOMC days. This shift is observed across various types of risk-free interest rates, including nominal, real, swap, short-term, and long-term rates, and is driven by unexpected interest rate shocks realized within the 30-minute FOMC announcement window. We develop a model featuring precautionary savings and levered investors to explain these results. In response to a positive transitory monetary policy shock, equilibrium interest rates and borrowing costs for levered investors rise persistently, leading to reduced risk appetite and heightened financial market risks.

[3] Risky Value Meets Behavioral Momentum [04/2022] [SSRN]

Abstract: This paper demonstrates that risk-based and behavioral cross-sectional asset pricing anomalies can plausibly coexist. To this end, I build a model in which risk-based value premium exists along with behavioral momentum. The value premium stems from differential exposures of stocks to rare disaster risk. Momentum is the result of behavioral underreaction to firm-specific news. The model reproduces several key patterns about value and momentum factor returns, including average excess returns, factors' correlation and the failure of CAPM in explaining both anomalies. The model also generates the reversal of momentum returns precisely due to the existence of a persistent value premium. The behavior of value and momentum during market crashes lends support to the model mechanism.

[4] Do Analysts React to FOMC Announcements?

Abstract: This paper shows that sell-side financial analysts respond to Fed's monetary policy shocks when revising their earnings expectations. The sign of the reaction is time-varying: before the end of 2012, an unexpected monetary policy tightening leads analysts to downgrade their earnings expectations, consistent with the standard New Keynesian view. The relationship inverts in December 2012, after the Fed explicitly promises to condition the future policy on economic outlook. The positive relationship suggests that analysts perceive monetary policy shocks as the news about economic fundamentals, in line with Fed's information channel. Finally, I find that the earning forecast revisions underreact to monetary policy shocks in both time subsamples. Overall, I show that Fed's monetary policy affects firm cash flow expectations both through the New Keynesian and the information channels.