Fast and Slow Arbitrage: The Predictive Power of (Persistent) Capital Flows for Factor Returns

• 成果类型: Article
• 署名作者: Dong, Xi; Kang, Namho; Peress, Joel
• 署名单位: City University of New York (CUNY) System; Baruch College (CUNY); Bentley University; INSEAD Business School
• 刊物名称: REVIEW OF FINANCIAL STUDIES
• ISSN/ISSBN: 0893-9454
• DOI: 10.1093/rfs/hhaf036
• 发表日期: 2025
• 页码: 2936
• 关键词: long-run risk; cross-section; stock returns; momentum; anomalies; investors; demand; tests

摘要

We document that persistent aggregate capital flows to hedge and mutual funds predict monthly factor returns with an out-of-sample R2 reaching 6.6%. Transient flows display no such power despite being more predictable. We show-both empirically and theoretically-that persistent flows' predictive power stems from active fund managers' capital constraints. As a result, managers invest persistent, but not transient, capital flows into factor trading strategies, leading to factor-return predictability and factor momentum, yet greater price efficiency. Our key insight is that capital-constrained managers account for both current and anticipated future flows in the arbitrage sector, thereby incorporating the dynamics of capital into their strategies.