Institutional Trading around Corporate News: Evidence from Textual Analysis
结合新闻发布和机构交易的高频数据，研究发现机构投资者在新闻集群中最早发布后立即根据新闻基调交易，这种交易能预测后续几周的回报，表明机构通过快速解读公共信息促进价格效率。
Abstract We examine institutional trading surrounding corporate news by combining a comprehensive database of newswire releases on U.S. firms with a high-frequency database of institutional trades. To identify the ability of institutions to predict or quickly interpret news, we form “news clusters” of related news about a particular firm that occurs in rapid succession. We find that institutions chiefly trade on the tone of news directly after the earliest news release in a cluster, and such news-motivated trading predicts returns over the following weeks. Our results suggest that institutional investors contribute to price efficiency through the speedy interpretation of public information.