This study investigates whether media management exists in private placements and public offerings. In order to precisely analyze media management on private placements and public offerings, we use Support Vector Machine (SVM) as measure to train news term so that we can establish news feature classification model. We use news, which from 1996 to March 2016 in TEJ to analyze and collect study data. We end up collect 3083 total news in “before the announcement day” period, including 586 private placement news and 2497 public offering news. In “after announcement day” period, we have 3031 total news, including 586 private placements news and 2445 public offering news. According to our regression results, we find that media management exists in private placements and public offerings. In addition, private market prefers to control news coverage by increasing positive news and decreasing negative news.