Working Papers
Working Papers
Present: Maryland Brownbag
Abstract: This paper documents a mechanism by which institutional investors support asset prices through the asymmetric execution of routine trades. Institutional buy orders generate a 50–100% larger price impact than equivalent sell orders, thereby creating price support for existing holdings even without net buying. The asymmetry intensifies near reporting period ends when valuation incentives peak, and for larger positions where price increases yield greater portfolio-wide gains. Quasi-exogenous flow-induced trades rule out alternative explanations such as information asymmetry. The mechanism has three asset pricing implications: (1) 1% institutional turnover without net buying raises contemporaneous returns by 0.1%; (2) high institutional ownership stocks earn 5.7% cumulative four-factor alpha over 18 months before fully reversing, explained by predictable turnover; and (3) a long-short strategy exploiting within-quarter seasonality yields 13.2% annualized alpha. As institutional ownership has risen from 32% to 73% since 1980, understanding this execution-based price support is increasingly important.
Present: CICF 2025, Transatlantic Doctoral Conference LBS 2025, EasternFA 2025, SWFA 2025, AFA 2025 PhD Poster, SFA 2024, University of Maryland
Invited Industry Presentation: TwinBeech Capital, Wolfe Research 8th Annual Global Quantitative and Macro Investment Conference
Media Coverage: Money Stuff / Bloomberg
Abstract: Stocks that close just above a round number (e.g., $6.10) tend to rise and outperform those that close just below (e.g., $5.90) by 24.6 basis points the next day and 46.1 basis points over the following week. I attribute this predictability to limit order clustering: concentrated buy orders at round prices support stocks above the threshold, whereas concentrated sell orders restrain those below. The effect is robust to various firm characteristics, appears throughout intraday half‑hour intervals, and replicates in 18 international equity markets. Exogenous price shifts around stock splits and dividend payments provide causal evidence. Consistent with a retail‑trading channel, the spread is larger in stocks with heavy retail activity, and trades executed exactly at round numbers are smaller. Order clustering also slows the incorporation of earnings‑announcement news and contributes to short‑term return reversal. Overall, these findings reveal how a high‑frequency microstructure friction can shape return dynamics over much longer horizons.
Present: AFA 2026 (scheduled), CICF 2025, SFS Cavalcade Asia-Pacific 2024, EUROFIDAI-ESSEC Paris December 2024, University of Maryland, CUHK-Shenzhen*, George Washington University*, HKUST*, University of Florida*, University of Macau*, University of Notre Dame* (*by coauthor)
Abstract: A disproportionately large fraction (70%) of the stock momentum reflects return continuation on the same weekday (e.g., Mondays to Mondays), or the same-weekday momentum. Even after accounting for partial reversals on other weekdays, the same-weekday momentum still contributes to a significant fraction (20% to 60%) of the momentum effect. The same-weekday momentum is hard to square with traditional momentum theories based on investor misreaction. Instead, we provide direct and novel evidence linking it to within-week seasonality and persistence in institutional trading. We also offer the first piece of direct evidence that seasonal fund flow drives seasonal institutional trading.
4. "Perpetual" Futures (with Albert "Pete" Kyle)
Faculty-Student Research Award, University of Maryland
Abstract: This study investigates the unique funding rate mechanism of perpetual futures contracts and its potential for price alignment with spot assets in traditional financial markets.
5. Informative Equity Price and Corporate Lending (with Brandon Yueyang Han)
Abstract: This study demonstrates how lenders learn from informative equity value to adjust their debt supply.
6. Attention and Memory (with Zhi Da)
Abstract: This study examines how investor attention in the recent or distant past transforms into active or passive memory, triggering future trading.