Event study methodologies in information systems research

https://doi.org/10.1016/j.accinf.2011.01.002Get rights and content

Abstract

Event studies are based on the theoretical framework of efficient capital markets and the notion that security prices include all information available to the market. As a result, announcements made by firms provide to market participants information that can be impounded into the market price. This paper investigates the use of event studies in information systems and accounting information systems research using a three-pronged approach. First, this paper provides a comprehensive survey of research that uses event study methodologies, where the events are announcements made by firms about issues related to information systems, e.g., announcements of the adoption of enterprise resource planning systems and of the effect of security breaches in firms' information systems. Second, this paper summarizes event study methodologies used in prior research, along with some of the key parameters and concerns associated with their implementation. Third, this paper provides remarks on key event study modeling issues, and it offers recommendations to researchers.

Research highlights

► We examine the use of event studies in the analysis of research questions related to information systems, especially accounting information systems. ► We examine differences in event studies in information systems as compared to other disciplines. ► We provide a comprehensive survey of information systems studies that use event study methodologies. ► We discuss the importance of issues such as confounding events, time, firm size, and future performance.

Introduction

Event studies have been a major focus of prior research because they provide a powerful setting to examine the informativeness of an event as assessed by market participants. An event study first requires identifying the event of interest, e.g., disclosure of the purchase of a particular type of software. After the event is defined, the period of time over which the stock price of the firm experiencing the event is determined. Then, the stock price changes beyond the “normal,” or expected changes, in response to the event announcement, are examined to determine the extent to which the event changes the market participants' evaluation of the firm.

The notion of efficient capital markets (Fama, 1970) provides a strong theoretical foundation for this basic event study methodology. Fama (1991, p. 383) notes that “security prices fully reflect all available information”. As new information is made available to the market, e.g., in the form of announcements about a firm's use of information technology, investors are expected to impound this information into the firm's stock price to capture the expected effect of the new information on the firm's value. As a result, the incremental effect of the information announcement on the value of the firm can be observed.

Event studies have been widely used in virtually all business and economics disciplines. Perhaps the first event study was published by Dolley (1933), who investigated the effect of stock splits on stock prices. The modern methodology of event studies was initiated by Ball and Brown, 1968, Fama et al., 1969, but the methodology has continued to evolve over time (MacKinlay, 1997).

MacKinlay, 1997, Binder, 1998, Kothari and Warner, 2006, and others provide analyses of event studies in finance. In addition, Dehning et al., 2003a, Dehning et al., 2003b, Roztochi and Weistroffer, 2008, Roztochi and Weistroffer, 2009a, Roztochi and Weistroffer, 2009b, and others provide reviews of different aspects of the use of event studies in information systems. However, this paper focuses on methodological issues as they relate to the use of event studies in information systems. In addition, this paper updates the literature that uses event studies in the information systems research area. Further, this paper evaluates the research questions examined in prior studies, and analyzes the comparative limitations of alternative methodological approaches.

At their most general level, event studies do not necessarily include or require stock market information. Instead, there could be a relationship between an event and a dependent variable. For example, Felcher et al. (2010) study the relationship between the event of “changing teachers” in a school and students' standardized test results. However, in this paper we assume that the event relates to an enterprise technology and the effect of the event is measured in a stock market response. Relating the effect of an event to a stock market response allows researchers to determine whether the event provides new, incremental information to stock market participants and the extent of the economic impact of the event on firm value.

The purpose of this paper is to survey the literature on event study methodologies related to information systems, and investigate some of the key concerns with using event studies in information systems. In so doing, we analyze parameters associated with the methodology of over 50 information system event studies. Because the paper is primarily concerned with the methodology used in the event studies, we do not focus on the actual results or conclusions of the specific studies. A comprehensive set of references is provided at the end of the paper for the reader who wishes to further examine these research studies.

This paper proceeds as follows. Section 2 describes how event studies differ in information systems, in contrast to accounting and finance studies. Section 3 investigates the basic event study methodology, laying out six different steps. Section 4 examines the possibility that confounding events can occur during the event window. Section 5 discusses the importance of time-related issues, among these are stationarity and meta events occurring over time. Section 6 investigates the impact of firm size on event studies. Section 7 analyzes the question, “after the event, then what?,” focusing on the analysis of future performance to determine whether the stock market is correct in its anticipation of the effect of the event. Section 8 analyzes limitations of an alternative market measure, and it describes why event studies are superior to an alternative approach to investigating the market response of technology adoption. Section 9 provides a summary of our key recommendations regarding event studies. Section 10 investigates the overall impact of information technology (IT) event studies, aggregating the results. Finally, Section 11 summarizes the paper, discusses its contributions, and provides potential extensions.

Section snippets

The difference between event studies in information systems and event studies in other settings

Event studies have been used in a wide range of settings, including accounting and finance (e.g., MacKinlay, 1997). As an example, in finance, researchers have used event studies to examine the market effect of mergers and acquisitions. Additional examples in accounting include whether accounting disclosures contain information, based on whether the stock market reacts to the disclosure of information events. In general, in virtually any discipline, the basic methodology remains the same: there

Event methodology

MacKinlay (1997) outlined an event study methodology involving the following steps: (A) identification of the event of interest; (B) definition of the event window; (C) selection of the sample set of firms to be included in the analysis; (D) prediction of a “normal” return during the event window in the absence of the event; (E) estimation of the “abnormal” return within the event window, where the abnormal return is defined as the difference between the actual and predicted returns, without

Confounding events during the event window

Event studies are designed to capture the impact of a specific event. If another event occurs at roughly the same time of the event of interest, there would be a question as to what was the true cause of a change in market price. Accordingly, it is important to eliminate those announcements that may be tainted by another event or a set of events. As a result, a careful and thorough research design should investigate other announcements during the window of interest to determine whether there

The importance of time

Not surprisingly, “time” plays an important role in event studies, through issues such as stationarity, importance of the particular time period, and meta events.

The importance of firm size

As noted by Kothari and Warner (2006), firms experiencing an event can have a non-random size and be from a non-random industry. For example, in the mid 1990s a typical implementer of an ERP system was both a manufacturing firm and generally a very large firm (e.g., O'Leary, 2000). Thus, announcements of ERP systems were likely from similar industries and of similar size. As a result, event studies can potentially capture substantial industry and firm size effects, which suggests that some

After the event, then what? (relating stock market returns to accounting profits)

The stock market responds to event information because it anticipates a change in a firm's value. Accordingly, an important issue is whether the performance of a firm changes after the event. Thus, a complementary analysis to an event study could examine future performance to determine if the stock market's reaction to the event of interest is appropriate in terms of the event influencing future performance. Researchers (e.g., DeFond et al., 2010, Masli et al., 2010) have taken multiple

An alternative market measure to an event study

Hitt and Brynjolfsson (1996) and others have proposed a different market-based measure and approach to analyze the market impact of technology adoption. In particular, researchers have related information system developments to Tobin's q, which is defined as the ratio of market value to replacement value of assets. As an example, Bhardadwaj et al. (1999, p. 1017) found that “… the IT expenditure variable in the model increased the variance explained in q significantly. This indicates that the

Remarks on modeling and recommendations to researchers

Events studies, which ultimately test for market efficiency, are joint tests of market efficiency and the model of expected returns used to estimate abnormal returns. There are several important issues about which the information systems researcher should decide when designing the research and setting up the model of expected returns. First, when appropriate, we recommend using daily, rather than monthly security returns data because such data allow more informative examination of the event of

Does IT matter? Do IT event studies matter?

Although the purpose of this paper is to investigate the use of the event methodology in information systems, by bringing together a large number of these studies in a single setting it allows us to investigate another question. In particular, Carr (2003) investigates the question “Does IT Matter?” Based on the 50 or so IT event studies summarized in Table 1, it is clear that stock prices act as if IT does matter — at least the specific technologies investigated in those research papers. When

Summary, contributions, and extensions

This paper reviews and analyzes methodological features of event studies related to information systems announcements. In so doing, this paper provides a comparison of methodological parameters across a number of event studies in information systems that could be used to guide researchers. Further, this paper provides an updated literature survey on event studies in information systems, as well as it evaluates alternative research design choices and makes recommendations for researchers,

Acknowledgement

We thank Andreas Nicolaou, an anonymous associate editor, and two anonymous referees for their comments and suggestions on earlier versions of this paper. We also thank the participants at the Second Annual Pre ICIS Workshop on Accounting Information Systems, in St. Louis, Missouri in December 2010.

References (92)

  • J. Min et al.

    Tobin's q, agency conflicts and differential wealth effects

    Glob Finance J

    (2001)
  • R. Morck et al.

    Internalization: an event study test

    J Int Econ

    (1992)
  • D.E. O'Leary

    Gartner's hype cycle and information systems research issues

    Int J Acc Inf Syst

    (2008)
  • A. Acquisti et al.

    Is there a cost to privacy breeches? An event study

  • N. Aggarwal et al.

    Do markets prefer open or proprietary standards for XML standardization?

    Int J Electron Commer

    (2006)
  • M. Argawal et al.

    Market reactions to E-Business outsourcing announcements

    Inf Manage

    (2006)
  • R. Ball et al.

    An empirical evaluation of accounting numbers

    J Acc Res

    (1968)
  • M.E. Barth et al.

    Cost of capital and earnings transparency

  • E Bartov et al.

    Capital market consequences of filing late 10-Qs and 10-Ks.

  • R. Benbunan-Fich et al.

    Effects of web traffic announcements on firm value

    Int J Electron Commer

    (2004)
  • R. Benbunan-Fich et al.

    Measuring the value of refining a web presence

    J Electron Commer Organ

    (2005)
  • A.S. Bhardadwaj et al.

    IT effects on firm performance

    Manage Sci

    (1999)
  • J. Binder

    The event study methodology since 1969

    Rev Quant Finance Acc

    (1998)
  • K. Campbell et al.

    The economic cost of publicly announced information security breaches: empirical evidence from the stock market

    J Comput Secur

    (2003)
  • A. Cannella et al.

    Effects of executive departures on the performance of acquired firms

    Strateg Manage J

    (1993)
  • M.M. Carhart

    On persistence in mutual fund performance

    J Finance

    (1997)
  • N. Carr

    IT doesn't matter

    Harv Bus Rev

    (2003)
  • H. Cavusoglu et al.

    The effect of internet security breach announcements on market value: capital market reactions for breached firms and internet security developers

    Int J Electron Commer

    (2004)
  • D. Chatterjee et al.

    Examining the shareholder wealth effects of announcements of newly created CIO positions

    MIS Q

    (2001)
  • D. Chatterjee et al.

    The shareholder-wealth and trading-volume effects of information-technology infrastructure investments

    J Manage Inf Syst

    (2002)
  • E. Daniel et al.

    The impact of ICT enabled off shoring announcements on share prices

    J Enterprise Inf Manage

    (2009)
  • M. Dardan et al.

    The valuation of ecommerce announcements during fluctuating financial markets.

    J Electron Commer Res

    (2005)
  • S. Dardan et al.

    The impact of customer-related IT investments on customer satisfaction and shareholder returns

    J Comput Inf Syst

    (2006)
  • M.L. DeFond et al.

    Does superior knowledge management increase shareholder value?

  • B. Dehning et al.

    Reviewing event studies in MIS: an application of the firm value framework

  • B. Dehning et al.

    The value relevance of announcements of transformational information technology investments

    Manage Inf Syst Q

    (2003)
  • B. Dehning et al.

    Reexamining the value relevance of ecommerce initiatives

    J Manage Inf Syst

    (2004)
  • J.C. Dolley

    Characteristics and procedure of common stock split-ups

    Harv Bus Rev

    (1933)
  • B. Dos Santos et al.

    The impact of information technology investment announcements on the market value of the firm

    Inf Syst Res

    (1993)
  • P.H. Dybvig et al.

    Tobin's Q does not measure performance: theory, empirics, and alternative measures

    (2010)
  • T. Erickson et al.

    On the accuracy of different measures of q

    Financ Manage

    (2006)
  • M.L. Ettredge et al.

    Information transfer among internet firms: the case of hacker attacks

    J Inf Syst

    (2003)
  • E. Fama

    Efficient capital markets: a review of theory and empirical work

    J Finance

    (1970)
  • E. Fama

    Efficient capital markets II

    J Finance

    (1991)
  • E. Fama et al.

    The adjustment of stock prices to new information

    Int Econ Rev

    (1969)
  • Felcher, J. Song, J. and Poindexter, S., “School Reforms often overlook the instructors,” Los Angeles Times, Wednesday,...
  • Cited by (113)

    • Does privacy breach affect firm performance? An analysis incorporating event-induced changes and event clustering

      2022, Information and Management
      Citation Excerpt :

      However, the stock market data is not likely to be stationary over time. Therefore, it suggested breaking the period into different groups [59]. A lack of stationarity in data may find the results stronger for one time period and weaker for another period.

    • Responses to SEC comment letters on cybersecurity disclosures: An exploratory study

      2022, International Journal of Accounting Information Systems
    • Apologize or justify? Examining the impact of data breach response actions on stock value of affected companies?

      2022, Computers and Security
      Citation Excerpt :

      Two event windows were analyzed. These consist of one day before the announcement and one and two days after the announcement; hence, we could minimize the impact of unrelated events and still get a first impression of whether the effect may be longer than one day (Konchitchki and O'Leary, 2011; MacKinlay, 1997; Pettengill and Clark, 2001). Each company's closing price for each day in the estimation period is required to conduct the relevant market-adjusted event.

    View all citing articles on Scopus
    1

    Tel.: + 1 213 740 9399.

    View full text