Explaining the Use of Graphs In

Explaining the use of graphs in

corporate takeovers

Keith Houghton

ANUCollege of Business and Economics, The AustralianNationalUniversity,

Canberra, ACT 0200,Australia.

McCombsSchool of Business. University of Texas,

Austin, Texas78712, USA.

and

Stephen Smith

Department of Information Systems, The University of Melbourne,

Melbourne, VIC 3010, Australia.

June 2006

Acknowledgement

The authors gratefully acknowledge the assistance of Dr Liz Roberts and Professor Christine Jubb for comments on an earlier draft of the paper. Remaining errors are the responsibility of the authors. Comments on the present draft are welcome to either author. Please do no quote present version of the paper without the authors’ permission.

Explaining the use of graphs in corporate takeovers

Abstract

The inclusion of graphs in financial reports is widespread yet little is known of what may motivate or explain their presence. This study analyses thepresence and extent of the use of graphs in formal documents which, for regulatory reasons, are required to be issued duringcorporate control(takeover) contests in Australia.

The results suggest that in a takeover contest, the presence andfrequency of the use of graphs is nota “neutral” or random phenomenon. Several factors explain the extent (and,to a lesser extent,the presence) of the use of graphs in takeover documents. These factors include: the nature of the takeover contest (hostile or friendly), the financial performance of the target, and takeover bid value. Inconsistent with expectations, the ownership dispersion(a proxy for the degree of sophistication of users of the financial reports)did not have a significant effect.

The results are consistent with the position that graphs are used by corporate management as a “tool” in corporate takeovers to positively manage the information signals provided to users of financial reports.

1

Explaining the use of graphs in corporate takeovers

1Introduction

For modern corporations, the appearance and content of financial reports is closely managed by those responsible for their production. This is true both for disclosures which areproduced as a consequence of regulatory requirements, as well as those which are voluntary and produced at the discretion of management. The presence and extent of graphical disclosures contained in financial reports are almost always discretionary in the hands of corporate management; although if included the nature of the disclosures is subject to certain constraints.While we knowa great deal about the information content and relevance of conventional financialmeasurements and related disclosures;by contrast relatively little is known about matters relatingto non-conventional information cues contained in financial reports including graphical representations. This study develops and tests several propositions about the circumstances under which graphs[1] are used in a certain class of financial reports. It seeks to contribute to a fuller understanding of the role of graphs in the communication of financial information.

The documents examined in this study are those issued during corporate control contests commonly referred to as takeovers. This setting was chosen because: (1) decisions made on the basis of these documents are often economically significant, and (2) the sometimes adversarial nature of these contests means that various participants have observable (or at least inferable) and differing economic incentives.

2Background

2.1Use and regulation of graphical information in accounting

Graphs are said to be a common way of illustrating corporate performance including financial performance (Beattie Jones, 1996). Some of the impetus behind graphically presenting financial data may be driven by the perceived needs of stockholders. This may be the motivation behind Section 409 of the Sarbanes-Oxley legislation[2], which encourages the use of graphs as a way to make reports easier to understand. Indeed as observed in Kosslyn (1994), graphs may assist in attracting and holding the reader’s attention. Editors and graphic designers frequently use information graphics to entice the reader into a story and also assist in the understanding of its content[3](Schriver, 1996). Similarly, it can be argued that corporate managers may find graphs a useful way of drawing attention to selected performance indicators in information releases such as annual reports (Beattie Jones, 1992).

While the inclusion of a graph in a financial report is largely, if not exclusively in some circumstances[4] at the discretion of corporate management, once included there are constraints in what the graph (or other non-conventional information) may depict. Specifically, there are constraints that require the absence of contradictions between the graphical representations and the underlying financial data. In many countries auditing standards constrain the way in which graphs can be constructed. Auditing standards in Australia (AUS 212), the United Kingdom (SAS 160), Hong Kong (AG 3.255), New Zealand (AS 518), Singapore (SSA 14 (1996), South Africa (AU 322), the United States (SAS 8), and the International Standards on Auditing (ISA 720) require that “other information” does not contradict the audited numerical financial statements[5].

Legislative requirements also influence graphical representations. For example, in Australia the relevant legislation, theCorporations Act(2001), specifically prohibits all types of misleading or deceptive conduct in relation to disclosure documents including prospectuses used in takeover contests. Moreover, investors have demonstrated their willingness to litigate where omissions or misstatements occur in certain related financial reports (prospectuses)(see Baxt, 1995; Foster, 1996). The Corporations Actcovers formal takeovers and, in Section 670A, prohibits the issuing of a document if there is “a misleading or deceptive statement in the document”. Importantly, the Act (Section 9) defines the term “statement” very broadly to include any information that “… conveys a message”. While there is no case law on the matter known to the authors it seems reasonable to conclude that this broad definition would include non-conventional information representations including graphs.

2.2Impression management in accounting

The provision of graphs in reports can be analysed from the perspective of a wider body of literature dealing with the management of perceptions generally,sometimes referred to as “impression management”. This wider set of literature tells us, for example, that subjects make differing decisions when presented with tabular information (somewhat like conventional accounting reports) as opposed to graphically presented data (see Dilla Steinbart, 2005). Also we know that improperly designed graphs can affect a reader’s choices (Arunchalam, Pei Syeinbart, 2002). The importance of research into the various methods of altering perceptions of company performance has beenacknowledged; “… while corporate annual reports have been transformed from minimal and legalistic exercises in financial statistics into a creative and flamboyant mix of text, images, and accounting data, such changes have gone largely unexamined in the research literature” (Hopwood, 1996, p55). Some would argue that little has changed since Hopwood made these observations. Hopwood further asserted that accounting research into corporate reports have, to date, been narrowly focussed on the accounts. This work he argues, typically examines accounting methods while ignoring the wider influences of the documents, and the type and method by which information is conveyed. These comments seem equally applicable to a wider range of financial information sources including prospectuses, takeover documents and other public corporate documents.

Although a number of distinguished scholars have published research into the use of non-conventional disclosures (including graphs) in financial reports, the absence of a significant body of literature in this area is somewhat surprising. Our surprise stems from: (1) graphs, photographs and other graphic design elements have been widely used by many organisations as “rhetorical” devices in reports to shareholders since the early 1960s (Graves, Flesher, et al & Jordan. 1996; Preston, Wright, et al.& Young 1996); and (2) a large body of research in diverse disciplines such as cognitive psychology and marketing has established that the format used to present information significantlyinfluences the message conveyed.

3The Research Setting of the Present Study

Corporate takeovers were chosen as the context for this study because they are a clearly defined event in which certain groups of information providers are required by legislation to inform specified stakeholders (stockholders) so that these stakeholders might make a more informed choice as to a particular course of action (generally retain or sell shares). Formally, a corporate takeover is an economic event that involves one company (“bidder”) attempting to acquire the shares of another (“target”) in order to gain a material equity interest. This may either concur with the wishes of the target’s board of directors (and management), that isa “friendly” takeover, or without the agreement of the target’s board;a “hostile” takeover. For the purposes of this study, takeovers are defined as those events that involve applications to Australia’s corporate regulator, the Australian Securities and Investments Commission (ASIC), and lodgement of documents pursuant to the relevant legislation relating to corporate control. This definition means that for any one contest there are two parties only, a bidder and a target. It is possible there many be multiple bidders in the same time period with a “shared” target. This study relates only to entities where the target company is a publicly tradedcompany. The study is also limited to a narrow period when access to data,in the exact form it was provided to stockholders and not reworked to conform to a regulatory template, was available.

Given data availability, a study of takeovers avoids several of the problems associated with other settings where information is proprietary or where a self-selection or survivor bias may be present.From an accounting perspective, studying information presentation behaviour in takeovers is also worthwhile because of the absence of similar research in this context.

The remainder of this paper is organized as follows. In the next section we provide some theoretical support to suggest that, certain company characteristics, the nature of the takeover (friendly or hostile), and the company’s role in the takeover (bidder or target)can help explain the presence and extent of graphs in takeover documents. The hypotheses developed are then testedin the next section. We then provide the results of the two models developed (one for presence of graphs and the other for extent of use) and a discussion of those results. Conclusions and implications are then explained.

4 THE LITERATURE AND THE DEVELOPMENT OF HYPOTHESEs

4.1Signalling and agency theory

As noted above in most instances, there is no obligation to include graphs in accounting documents. One might conclude that the theory behind the use of graphs is similar to other voluntary disclosures. This may not be entirely correct as other voluntary disclosures generally providenew information, whereas a graph generally reintroduces and/or highlights information already contained elsewhere in the financial report.

In part, the voluntary disclosure literature uses either signalling theory(e.g. Penman 1980; Dye 1985, 1986; Hughes 1986; Clarkson, Dontoh, et al.Richardson &Sefcik 1992; Dye 1998) and/or agency theory(e.g. Ruland, Tung et al& George 1990; Mak 1996) as the theoretical basis for research into factors influencing the volume, nature and type of information disclosure. The basic proposition of signalling theory is that firms with good news (such as positive earnings forecasts) have incentives to voluntarily disclose that information in order to distinguish themselves from less desirable firms[6](Verrecchia 1983; Dye 1985; Verrecchia 1986; Dye 1998). Therefore, by using signalling theory, graph production can be explained as a means of graphically highlighting selected aspects of performance so that firms can differentiate themselves from less desirable alternatives.

As noted below, we argue that theoretically, certain company characteristics, their position in the takeover (bidder or target) and the nature of the takeover (friendly or hostile) can explain the presence and extent of the usage of graphs.

4.2Company characteristics

Each company involved in a takeover contest has certain characteristics that place them in differing circumstances. These include profitability, nature of the ownership of the company and the economic scale of the takeover contest (that is, the size of the target company).

Profitability

From a signalling theory perspective, companies with higher levels of profitability have incentives to highlight this performance to distinguish themselves from other firms. Therefore, one explanation for the usage of graphs is that managers incorporate them into reports to emphasize success (see Steinbart, 1989, Beattie Jones, 1992; 1994). The reverse is also true. When performance has been unfavourable, a graph would highlight poor performance perhaps leading to additional monitoring or other less desirable actions by principals.Therefore, managers may be hesitant to include a graph. In the context of this study it is the profitability of the target that is most at issue for it is the performance of the target that is often central to the arguments in favour or not in favour of the takeover contest.

Nature of ownership

The type of information released in annual reports and takeovers is likely to be influenced by a firm’s ownership structure (Jensen Meckling, 1976). If shareholders are financially sophisticated investors, then one could argue that detailed accounting and other technical data are likely to be of greatest benefit[7]. However, other types of investors may have different information preferences. While graphics may be ignored or used with lesser weight by somesophisticated report users, they may well be used in the buying, holding and selling decisions of less experienced, less wealthy, less educated and financially unsophisticated investors (Chang Most 1985; Epstein Pava 1993)[8]. Therefore, the greater the proportion of shares held by this type of (generally economically smaller) investor, the more likely that graphs will be included. One can argue that graphs are of more use to this type of investor because they match this group’s information and information processing circumstances. This study conjectures that the more disperse the shareholder base, the more likely this type ofunsophisticated investor will be represented in the share register. Therefore, it is argued that the more disperse the shareholdings of the target, the greater the likelihood that graphs will be provided in takeover documents as shown in this study[9].

Size and takeover value

Several studies (Buzby 1975; Chow Wong-Boren 1987; Cooke 1991) suggest that firm size is an indicator of the amount of resources available for preparing and disseminating information[10]. Applying this to the takeover context, the value of a takeover rather than the value of each firm per se may be more indicative of the amount of resources likely to be used by participants to produce and disseminate information.

It is argued, that a target company’s profitability and ownership structure and the value of the takeover each explain the presence and extent (number) of graphs disclosed in takeover documents by information suppliers. This leads to the following formal hypothesis:

H1The value of the takeover, the target firm’s profitability and ownership structure will be positively and significantly related to the frequency of the presence of graphs and extent of graph usage in financial reports issued by bidder and target companies in the context of a takeover context.

4.3Adversarial conflict and the presence and extent of graphs

In a takeover, the level of adversarial conflict essentially depends on whether a takeover is “hostile” or “friendly”. A hostile takeover is distinguished by the presence of mutually exclusive actions by the bidder and the target (accept or reject). In this circumstance, the bidder and target send different signals to shareholders to support these differing proposals. To gain support (accept or reject) from shareholders, bidder and target management also have incentives to interfere with each other’s signals. Adversarial conflict is likely to be high because the outcome often has a significant impact on an organisation’s future, but whether the takeover is successful or not is often dependent on the actions of many individuals and corporations not under the direct control of management of either the target or the bidder[11]. The association between adversarial conflict and information release suggests that graphs are more likely to be used in hostile rather than friendly takeovers.

In a friendly takeover, documents from the bidder and the target send complementary signals about the bid. So, rather than having an adversarial relationship, the firms are likely to jointly emphasise the benefits of synergy and the disadvantages of individual existence. Because the firms involved recommend the same course of action (complementary signals), less information needs to be produced and both produce similar price signals and so do not need to counter noise created by the other’s documents as happens in a hostile takeover. Therefore, graphs are less likely to be used in friendly takeovers because (1) less information will be produced in total (information can be jointly produced); and (2) the lower adversarial conflict reduces the likelihood that impression management devices such as graphs will be used.

The previous discussion on the effect of the nature of the takeover (friendly or hostile) will influence the presence and extent of graphs. This leads to the following hypothesis:

H2Adversarial conflict will be positively and significantly related to the frequency of the presence of graphs and the extent of graph usagein financial reports issued by bidder and target companies in the context of a takeover context.

4.4Role in the takeover

A number of studies indicate that management may manipulate the contents of financial reports for strategic purposes[12]. This can be observed in one form where competing parties choose to issue markedly different sets of accounts for the same entity in support of their respective arguments (McBarnet D &J,. 1992; McBarnet, Weston et a. 1993).

Agency theory provides an explanation for different stakeholders to behave differently in respect of information issuance. It proposes that one reason takeovers occur is to replace managers who are not maximising shareholder wealth (Jensen 1988)[13]. For this reason, differential information release strategies will occur between bidder and target companies as the management of the latter faces the possibility of replacement (Mørck, Shleifer et al& Vishny 1988). In these cases, the larger quantity of information produced and highlighted by managers of the target resisting the bid may be part of an entrenchment strategy (Shivdasani 1993)[14].

This leads to the following hypothesis:

H3The presence of graphs and the extent ofgraph usage will be significantly greater in financial reports issued by target companies compared with bidder companies.