A New Event Study Approach for the Announcement of Information Technology Investments – Joint Events, Multiple Announcements and Window Identification

• Main Authors: Chen-yu Lee, 李琛瑜
• Other Authors: Jao-hong Cheng
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/83678573101779711012

博士 === 國立雲林科技大學 === 管理研究所博士班 === 97 === It has been considered that information technology (IT) enables enterprises to raise their competitive advantage and organization performance, and enterprises tend to increase IT investments continuously, hence the benefit of IT investments has elicited great concerns. Financial theory suggests that the market value of firms will be reflected on the stock price in response to IT investment announcements; therefore, the event study method is employed in this research. IT investments are classified into two categories including IT application and IT infrastructure to be discussed respectively by previous researchers; nevertheless, according to relative literatures and real cases, it indicates that both categories mentioned above are often included in IT investment announcements as a joint event as well as multiple announcements. It is probable to neglect the information content of subsequent duplicate announcements if just taking the first announcement day under imperfect market efficiency. Thus, the joint events of information technology investments as well as multiple announcements are taken into account in this study. Previous researchers usually set fixed-time event windows subjectively by try-and-error (Dos Santos et al., 1993; Chatterjee et al., 2002). However, too-long windows would probably include the noises of irrelevancies, whereas too-short windows would probably exclude the effects of target events. In order to obtain objective identification, the event window in this study is measured by using the iterated cumulative sums of squares (ICSS) algorithm for detecting change points. The sensitive analysis of cumulative average abnormal returns (CAAR) under various window settings that proposed by this study and previous researchers will be implemented to manifest the effect of window identifications in this study. The results indicate that positive CAARs are generated for the joint events of IT investments with multiple announcements, and a significance level of CAAR estimated by subsequent duplicate announcements is reached; moreover, the later duplicate announcements the less amount of CAARs; in other words, a later duplicate announcement results in a less amount of CAAR. The joint events with multiple announcements have information contents under imperfect market efficiency.