一、主题：Information Glitters on China’s Connected Market

二、主讲人：祝小全，清华大学五道口金融学院在读博士研究生，清华大学国家金融研究院民生财富管理研究中心研究员，中国财富50人论坛青年研究员。她的主要研究方向是资产定价、机构投资者交易和资本流动。目前，她已在学术期刊上发表多篇论文，包括《经济研究》《世界经济》、Accounting and Finance等。

三、时间：2019年12月25日，12:30-13:30

四、地点：学院南路主教913会议室

五、主持人：吴锴，中央财经大学金融学院讲师

Abstract: This paper explores a persistent mechanism through which investorsprofit from the potential information contained in capital flows from China’sconnected market. Using a complete history of daily filings about stock-levelholdings of all Hong Kong and international (northbound) investors, we documentthat weekly changes in northbound shareholdings (HPC) have a positivecross-sectional predictability for future stock returns. A long-short portfoliothat exploits this differentiating preference earns abnormal returns of up to61 basis points per week, which cannot be explained by common factors. Threesets of evidence suggest the information edge and trading impact of northboundinvestors: (1) HPC has stronger predictability around earnings announcements infirms with more international exposure and global accessibility (fundamentalinformation edge hypothesis); (2) Combined with the similar tradingbehavior with opportunistic insiders, the sign of decay in returnpredictability after the “penetrating” supervision on investor identity alsosuggests that some northbound investors are probably mainland insiderswith private information advantage that tend to make round-trip investment on Ashares to hide identity (round-trip trading hypothesis); (3) Theimitative herding of mainland investors following informed northbound flowswould help boost the future price of specific stocks suffering excessiveattention in the short run (copycat herding hypothesis). Besides, wealso observe similar cross-sectional predictability for H-share returns insouthbound institutional investors’ trading behavior.