This dissertation studies the daily institutional investors trading patterns before and after public news announcements in the US equity market, such as Merger and Acquisition announcement and release of macroeconomic indicators. Do institutional investors have inside information, or do they have superior models before news announcement? Using a high frequency institutional trading dataset that combines intraday NYSE Trades and Quotes (TAQ) data with the quarterly institutional ownership report (13F) by a reduced-form model, this dissertation tests the hypothesis of institutional investors trading on inside information 1993 to 2004. I find that most institutional investors are informed traders who accumulate shares before good news or before takeover announcements as early as 30 days ahead. Institutional investors do not have superior models in that they only buy the actual future targets and sell the forecasted ?rumor? stocks from an acquisition probability model. By reversing their positions on and after the announcement day, they realize positive profits. Further, I confirm that the pre-event trading pattern of institutional investors is associated with stocks that have high probability of informed trading.