Research

Research Interests

My research examines how information frictions shape investor decisions and market discipline in financial intermediation. I study how disclosure design affects investors' responses to mutual fund fees and how analysts' incentives and capacity constraints shape the information they produce. These projects examine both the information investors receive and their ability to use it to evaluate financial intermediaries.

Job Market Paper

1. Fee Salience and Mutual Fund Market Discipline

   Abstract: This paper asks whether disclosure design, rather than disclosure quantity, is a key friction in mutual fund markets. I study the SEC's 2022 Tailored Shareholder Report (TSR) rule, which changed how fee information is presented but did not change fee schedules or add new fee content. Using staggered share-class adoption tied to predetermined reporting calendars, I find that flow–fee sensitivity becomes about 29% steeper after TSR adoption. The change is much stronger in retail share classes, where the combined post-adoption steepening is about 3.7 times the institutional response, and is concentrated where pre-reform reports were harder to process: longer, less readable, and more bundled. On the supply side, fee adjustments are modest and mixed, while high-fee share classes are more likely to exit through liquidation or merger. A decomposition of investor fee burden indicates that most of the post-reform decline in retail expense ratios reflects capital reallocation and exit rather than within-fund repricing. Overall, the evidence suggests that weak investor response to fees partly reflects presentation frictions, and that redesigning existing disclosures can strengthen fee competition.

Working Papers

1. IPO Underwriting and Peer-Firm Analyst Optimism (with Yuchou Peng)

   Presentations: AsianFA 2025, AFBC 2025, HKU Business School, Erasmus University (RSM)

   Abstract: Across 1,730 U.S. IPOs, lead-underwriter analysts become more favorable toward issuers' product-market peers during underwriting. Buy probabilities rise 5.5% and sell probabilities fall 21.8%, relative to their unconditional means. This Peer-firm IPO Optimism (PFIO) rises before listing, declines afterward, and strengthens with peer visibility and valuation complexity. Neither EPS nor long-term growth forecasts become more optimistic, and EPS forecast accuracy does not detectably change. These patterns favor strategic IPO support over sincere optimism. Affected buys and upgrades yield lower subsequent abnormal returns but receive no detectably smaller announcement responses, potentially preserving analysts' incentives to promote IPOs through optimistic peer research.

2. Discounting Chinese Industrial Policy (with Andrew Sinclair) [SSRN]
   Revise & Resubmit, European Economic Review

   Media coverage: VoxChina

   Presentations: USC (China Workshop), University of Toronto, HKU Business School, USC (Finance)

   Abstract: Each year, the Chinese central government issues dozens of policy documents directing its local officials to promote the development of targeted industries. Although central support is valuable in expectation, the duration and intensity of support are uncertain and only revealed over time. We find that small supported firms earn between 3.6% and 7.2% annualized excess returns relative to standard Chinese asset-pricing benchmarks. We demonstrate that the pattern of these returns is consistent with rational discounting of policy-support uncertainty, and is difficult to reconcile with alternative explanations based on systematic mispricing. Together, our results reveal an important market dynamic that is rooted in China's unique political economy.

3. Why (How) Do Analysts Issue Multiple Forecasts in a Day? (with Tse-Chun Lin)

   Presentations: AFA 2025 (poster), AsianFA 2025, HKU Business School

   Abstract: Analysts increasingly issue earnings forecasts for multiple firms on the same day (“forecast clustering”), accounting for 55.9% of one-year-ahead EPS forecasts in 2024. Because clustering is associated with lower forecast accuracy, we examine why clustering occurs and when analysts avoid it. Clustering is more prevalent when workload creates scheduling congestion, including concurrent earnings announcements, elevated news activity at other portfolio firms, and exact-second same-industry forecast batches. Clustering is less common for career-relevant firms and for forecasts accompanied by recommendation revisions, consistent with strategic effort allocation under capacity constraints. Firms persistently exposed to clustered forecasts exhibit wider bid–ask spreads and higher illiquidity.

Publications

1. Retail Hedge Funds (with Andrew Sinclair)
   Journal of Alternative Investments, 2025