ARTICLE IN PRESS

Technovation 29 (2009) 704713

Contents lists available at ScienceDirect

Technovation
journal homepage: www.elsevier.com/locate/technovation

Post-merger performance in the software industry: The impact of characteristics of the software product portfolio$
 Pierre-Majorique Leger , Louis Quach 1
   HEC Montreal, Department of IT, 3000 chemin de la Cte-Ste-Catherine, Montreal, Quebec, Canada H3T 2A7 o

a r t i c l e in f o

a b s t r a c t
This article studies the impact of the characteristics of software product portfolios on the performance of rms involved in a merger of software companies. The short-term nancial results reveal that markets generally seem to neglect the characteristics of software product portfolios when the merger is announced. Nevertheless, such portfolios appear to have a positive impact on the price/book value ratio of merged software rms. The empirical evidence presented in this paper suggests that, in the long term, the performance of business combinations in the software industry is related to certain factors that are attributable to virtual network effects. Crown Copyright & 2009 Published by Elsevier Ltd. All rights reserved.

Keywords: Software portfolio Merger Network effect

1. Introduction Many information technology (IT) specialists agree that the software industry has entered a phase of maturity in the last few years. Probably as a result, the merger-and-acquisition activity in the IT industry has intensied to unprecedented levels in recent years. Examples of high-prole deals concluded recently include the acquisition of Cognos by IBM ($5.0 billion), Business Object by SAP ($6.5 billion), Hyperion ($3.3 billion) and BEA Systems ($6.5 billion) by Oracle, MySQL by Sun Microsystems ($1.0 billion), and Mercury Interactive ($4.5 billion) and Opsware ($1.6 billion) by HP. And this does not even take into account the $7 billion acquisitions made by EMC and Computer Associates more than 150 acquisitions over the last decade. The merger-and-acquisition literature reveals that few business combinations achieve the performance level that was anticipated at the time the decision to merge was made (Datta et al., 1992). It is therefore legitimate to ask the following question: What factors determine the performance of a merger or acquisition in information technology? Few studies have specically covered the software sector. More specically, no study has yet considered the characteristics of the product portfolios held by the rms involved, which may well prove to be a determining factor in explaining the performance of mergers and acquisitions of software rms.

Based on network theory, this article hypothesizes that, beyond the traditional antecedents of the performance of business combinations, the performance of combinations of software companies should be positively impacted by the virtual network effects that result from the compatibility and complementarity of the new entitys software products. More specically, we examine the impact of these two factors on the short-term market performance of both entities stock at the time of the announcement, on the value of the transaction and on the long-term nancial performance of both entities, when measured alongside the impact of more traditional variables.

2. Review of the literature: the performance of business combinations In the nancial literature, mergers and acquisitions have always been a topic of great interest, and this continues to be true. A considerable proportion of these studies are interested in the creation of value for shareholders. For many authors, value creation is a good indicator of the performance of a business combination (Agrawal et al., 1992; Datta et al., 1992; Lubatkin, 1987; Singh and Montgomery, 1987). From a strictly nancial point of view, the impact of the business combination is often assessed based on share value. However, other specialists also measure performance in accounting and economic terms (e.g., return on investment) (Pautler, 2003) or by the level of synergy achieved (Larsson and Finkelstein, 1999). In all these cases, performance is associated with value creation. Consequently, it is important to investigate factors that are likely to result in value (Seth, 1990b). In this context, Brouthers et al. (1998) suggest that the motivations for mergers should be

$ We would like to thank Carl St-Pierre, Mohammed Jabir and Karine Marion for able research assistance. We would also like to recognize the nancial contributions of the FQRSC and NSERC. Corresponding author.  E-mail address: pierre-majorique.leger@hec.ca (P.-M. Leger). 1 Now works at SAP Canada.

0166-4972/$ - see front matter Crown Copyright & 2009 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.technovation.2009.05.016

ARTICLE IN PRESS
 P.-M. Leger, L. Quach / Technovation 29 (2009) 704713 705

perceived as key success factors. More than motivations, these key success factors can be termed performance drivers. In light of the existing literature, we propose to dene the performance of a business combination by the achievement of certain motivations and goals; the degree of achievement should be reected in the creation of value for shareholders. A number of authors have studied the factors inuencing the performance of business combinations (Brouthers et al., 1998; Lehto and Lehtoranta, 2004; Seth, 1990b). Taken together, these studies suggest that the main performance antecedents relate to four factors justifying the combinations economic potential: the potential for market growth, the potential for economies of scale, the potential for economies of scope, and the potential to acquire competencies. The following paragraphs dene each of these antecedents:

 Economies of scale constitute a classic motivation for mergers

and acquisitions (Brouthers et al., 1998). Combining into a new entity makes it possible to reduce average costs by consolidating production, as well as administrative, commercial, logistical and research and development services. In the context of a merger, economies of scope emerge when it costs less for the new entity to produce different varieties of a single product, primarily because of the consolidation of purchases, advertising and distribution, which are now done on a larger scale (Priest, 1994). Thus, economies of scope are one of the main reasons for marketing strategies such as the combined sale of products, the sale of related products or the sale of products under a single brand name (Cabral, 2001). Market growth suggests a gain in market share and an improvement in competitive positioning. This concept also integrates the notion of increased market power, that is, a rms ability to better control the prices, quantities or nature of the products it sells (Seth, 1990b). A business combination that is motivated by the acquisition of competencies has the goal of acquiring skills that are difcult or would take too long to develop internally, or that cannot be purchased in a transaction (Capron, 2005). The recent literature suggests that this factor may be crucial to the future success of the new entity (Gammelgaard, 2004).

The literature indicates that the performance of a merger or acquisition is closely connected to each of the above-mentioned motivations. Thus, the studies by Seth (1990a) and Pautler (2003) show that value creation in acquisitions is associated with economic performance that emerges from economies of scale, economies of scope and market power. Moreover, Capron (1999) suggests that value creation in acquisitions results not only from a reduction in costs or the ability to dictate prices, but also from the opportunity to use a specialized resource that results from the merger and its potential to create synergies. However, the potential for synergies rarely guarantees the success of mergers or technological alliances. King et al. (2004) nd that, on average, acquisition activity does not contribute positively to the performance of acquiring rms. According to Bayona et al. (2006), share prices generally do not react to the announcement of technological alliances. Furthermore, Tuch and OSullivan (2007) observe that, in the short run, acquisitions have at best an insignicant impact on shareholder wealth, and that long-term performance analysis reveals overwhelmingly negative returns. Accordingly, the literature also offers a number of explanations for the surprising proportion of acquisitions that have failed in recent years, especially in technologically intensive industries, and much of the research points to integration as the decisive factor.

Prabhu et al. (2005) argue that, for acquisitions to promote innovation, rms must rst engage in internal knowledge development. Bannert and Tschirky (2004) also concentrate on the importance of internalizing external knowledge, and identify the lack of integrative decision-making, of systemic processes and of a holistic change of both companies during the integration as the main causes of failed acquisitions. Along similar lines, Yoo et al. (2007) develop a framework that reveals that mergers represent a discontinuity in knowledge sharing. Puranam and Srikanth (2007) offer a balanced point of view on integration by accounting for the qualitatively distinct ways in which acquirers leverage technology acquisitions. They contend that acquirers use the acquired rms existing knowledge as an input to their own innovation processes but do not rely on it as an independent source of ongoing innovation. To address this issue, Ku et al. (2007) propose a virtual collaborative framework to foster knowledge sharing within the new entity. From a managerial point of view, Graebner (2004) places particular emphasis on the role of managers from the acquired company in solving implementation dilemmas. Paruchuri et al. (2006) take these results one step further by hypothesizing that the productivity of corporate scientists at acquired companies is generally impaired by integration, due to the loss of social status and centrality in the process. This can be assumed to be particularly true of technological rms. Tsai and Hsieh (2006) claim that the application of two-stage grey decision-making can assist corporations in selecting technological assets to create wealth through mergers, whereas Haro-Dominguez et al. (2007) nd that the degree of absorptive capacity has a positive inuence on both internal and external acquisitions of technology. Shaver (2006) describes two mechanisms he calls the contagion effect and the capacity effect, which establish the paradox of synergy, whereby the excessive attention given to achieving synergy actually distracts the rm from its environment. Sorescu et al. (2007) point to a rm-driven, marketingdriven variableproduct capital, which they dene as product development and support assetsto explain why some rms make smarter acquisition decisions. Wang and Zajac (2007) posit that rms should examine factors such as resource similarity and complementarity, relational capabilities and partner-specic knowledge before considering an acquisition. The performance of a nancial combination is also inuenced by nancial factors. It is not uncommon for the often astronomical costs incurred in a merger to result in intense pressures on the new rms short-term market protability, and also on its longterm accounting protability; often the amounts invested generate little return. Thus, any study of the performance of business combinations must take the nancial context in which the merger occurs into account. The amount of the merger or acquisition transaction has a signicant impact on the new entitys future level of prots. However, in itself, the amount of a transaction alone does not take sufcient account of the reasons justifying this transaction. If an acquirer rm takes over another rm for a certain consideration, it also takes possession of the target rms assets. In addition, the ratio of the amount of the merger-and-acquisition transaction over the total value of the target rms assets would constitute a more appropriate performance antecedent than the amount of the transaction considered in isolation. In the nancial literature, this ratio is referred to as the price/book value ratio (P/B). If this ratio is greater than 1, it shows that the acquirer rm is ready to pay more than the value of the rm it is acquiring. An overly high P/B ratio tends to increase the new entitys debt burden, thereby hampering its future development; this situation is likely to be severely punished by the nancial markets (Leroy, 2003). More importantly, assessing the performance of a business combination

ARTICLE IN PRESS
706  P.-M. Leger, L. Quach / Technovation 29 (2009) 704713

also requires one to take into consideration the targets protability before the merger. Indeed, the ex ante protability of the target rm should be reected not only in the P/B ratio but also in the new entitys performance (Hollender, 1967). It should also be noted that the mergers economic potential appears to be discounted in the P/B ratio (Varaiya, 1987). To sum up, we can summarize our current knowledge of the performance of business combinations in the model presented in Fig. 1. Thus, the empirical literature claims that there is (i) a negative correlation between the P/B ratio and the new entitys nancial performance; (ii) a positive correlation between the target rms ex ante performance and the new entitys nancial performance; and (iii) a positive correlation between the factors inuencing the mergers economic potential and the new entitys performance; nally, (iv) the P/B ratio is positively inuenced by the target rms ex ante protability and by the mergers economic potential.

Finally, we will propose a research model incorporating these variables. 3.1. Virtual network effects and the performance of combinations of specialized software rms Network effects have been dened as a gain in the value that a participant derives from a good in an industry when the number of participants consuming the same type of good increases (Economides, 1996; Katz and Shapiro, 1985). The higher the critical mass of users of a network of goods, the greater the value of the good is for each user. The fax and the telephone are classic examples of this phenomenon. Thus, a telephone has no value in the absence of a telecommunications network. In such bidirectional networks, the value of the network is equal to n*(n1), where n is the number of items. When a new item is inserted into the network, its value will increase by 2n. Economides (2001) introduced the concept of virtual networks, in which network effects also exist but are expressed in a slightly different way. He denes a virtual network as a collection of complementary and compatible technologies based on the same technological platform. A network effect is manifested in a virtual network by the mutual creation of value among the complementary and compatible technological components. In view of its nature, the software industry has properties that make it a particularly propitious business environment for the manifestation of virtual network effects. Thus, two compatible and complementary programs such as the Microsoft operating system and the Microsoft Ofce suite of applications can easily create value mutually and thereby promote the distribution of both. In this context, Ende and Wijnberg (2003) demonstrate through case studies that network effects offer an explanation for the increasing returns in the software sector. In other words, network effects contribute to the marginal prots of a product, which increase with the total quantity consumed or produced. Pehrsson (2006) identies four business relatedness classes, and nds that technology relatedness has the strongest positive performance effect. In order to measure the potential impact of virtual network effects on the performance of a combination of software rms, we propose to examine two main notions underlying the emergence of this phenomenon: software compatibility and software complementarity. As illustrated above, these two notions appear to be a sine qua non for the emergence of a virtual network effect. Compatibility and complementarity should therefore constitute indicators of the potential of network externalities within a product portfolio.

3. Research question and conceptual framework Although an extensive body of literature has examined the performance of business combinations, few studies have attempted to understand the antecedents of the performance of mergers of software rms. The software industry is characterized by a number of individual criteria that cumulatively justify our interest in it. The most noteworthy criterion is inherent in the intangible nature of software products. Essentially based on knowledge, the combination of software rms is associated with certain economic phenomena that are specic to the IT industry and that emerge from the characteristics of the product portfolio. More specically, this article will attempt to prove that, beyond the abovementioned antecedents to the performance of mergers and acquisitions, the potential for a network effect is a phenomenon that must be taken into consideration in evaluating the performance of a combination of businesses specializing in software. In other words, we are asking the following research question: Is the nancial performance of the rms involved in a software business combination inuenced by the potential network effect resulting from the characteristics of the new entitys portfolio of software products? We will rst present the concept of virtual network effects. Then we will operationalize this concept within two different notions: software compatibility and software complementarity.

Amount of transaction Value of target firms assets

+ + Ex ante profitability of target firms + Financial performance of business combination

Factors influencing the business combinations economic potential

Fig. 1. Factors cited in the literature as affecting the performance of business combinations.

ARTICLE IN PRESS
 P.-M. Leger, L. Quach / Technovation 29 (2009) 704713 707

3.2. The economic potential of software compatibility within the new entitys product portfolio Software compatibility is dened as the extent to which programs can work together and share data.2 Compatibility is assured by the issuance of standards by companies (e.g., compatible with Microsofts .NET standard) or consortiums of industry stakeholders (e.g., compatible with the XML standard). In this context, the results of the study by Kauffman and Li (2005) suggest that if a company wants to adopt a technology, it is more protable to wait until this technology becomes a standard or attains a critical mass threshold before investing. Moreover, it is important to think of compatibility as relative; a program will be more compatible in the absence of limitations on functionalities, in the presence of portability, and by assuring that it functions with older versions (downward compatibility) or future versions (upward compatibility) of other programs. Several other authors have studied compatibility and its advantages. According to Farrell and Saloner (1986), the three main benets of compatibility are the interchangeability of complementary products, ease of communication, and cost savings. In the view of Economides (1991), incompatibility is not a technical problem so much as a cost problem: it is possible to make any two products compatible provided that one accepts the costs of transitioning from incompatibility to compatibility. In the context of a business combination, if the products owned by the rms involved in the merger are compatible, this should reduce the investments the new entity needs to make to market a unied product portfolio. In addition, software compatibility can be perceived as a benet for customers in the sense that it allows the joint use of software and thus gives them access to new functionalities without making any additional investments. In other words, in addition to conferring technical advantages, compatibility is directly related to nancial investments: the more compatible the software products are, the lower the nancial investments required to make them work together. The compatibility of the products held by the new entity should therefore have a positive impact on the performance of the merged rm, but it should also increase the P/B ratio of the transaction, given the anticipated economic potential of this factor. However, it is important to keep in mind that, even though compatibility is expected to have a positive inuence on both these measures, the P/B ratio can be expected to have a negative impact on long-term nancial performance if the acquisition calls for a signicant increase in debt levels. We therefore formulated the following hypotheses:

3.3. The economic potential of software complementarity within the new entitys product portfolio Two products are described as being complementary when their joint use adds more value for the customer than the sum of the separate use of the same products (Carlaw and Lipsey, 2002). Product complementarity within a single portfolio is operationalized by the presence of functionalities within the different programs that create added value for the user when used jointly. This union leads to a gain in value for the user in the form of an addition to or a functional enrichment of the product portfolio. The merger of rms that have complementary product lines is also a frequent motivation for a combination. By offering an integrated product line, the new entity hopes to acquire an additional economic rent corresponding to the benet to users. Empirical evidence shows that resource complementarity increases the potential to create greater synergies in acquisitions, which leads to better performance in the long run. Thus, the results of studies by Harrison et al. (2001) clearly show that resource complementarity is a common characteristic of all acquisitions that perform well. Seth (1990b) also suggests that mergers and acquisitions in related industries may benet certain parties more but generally create as much value as mergers and acquisitions involving rms whose activities are not related. According to Harrison et al. (2001), it is important not to confuse complementarity and similarity: two activities or products may be complementary without necessarily being similar. This distinction is all the more important given that, before 1991, research suggested that acquisitions in related sectors should lead to better performance by the acquirer rm (Kusewitt, 1985; Singh and Montgomery, 1987). Harrison et al. (1991), on the other hand, show that synergistic gains from combinations of resources are much greater when the resources are complementary, rather than similar. Finally, by establishing a classication of three layers of complementary software activity, namely (i) systems software, (ii) middleware software, and (iii) applications software, Silva and Iyer (2006) emphasize the importance of software complementarity in a context of mergers and acquisitions. According to these authors, in a merger or acquisition between software rms, when the acquiring rms product is complementary to that of the target rm, in accordance with their classication of software products, better performance is observed in terms of abnormal returns than in mergers and acquisitions involving at least one non-software rm. In other words, complementarity is a source of value creation in software mergers and acquisitions. To sum up, a positive relationship should exist between product complementarity in a single software portfolio and the performance of the merged rm. Moreover, the economic potential associated with complementarity should also increase the transactions P/B ratio. Once again, the inverse relationship between the two dependent variables must be kept in mind. Based on these arguments, we made the following hypotheses: H2a. The greater the complementarity between the product portfolios in a business combination of software companies, the better the targets short-term market performance will be. H2b. The greater the complementarity between the product portfolios in a business combination of software companies, the higher the value of the transaction will be. H2c. The greater the complementarity between the product portfolios in a business combination of software companies, the better the new entitys nancial performance will be.

H1a. The greater the compatibility between the product portfolios in a business combination of software companies, the better the targets short-term market performance will be.

H1b. The greater the compatibility between the product portfolios in a business combination of software companies, the higher the value of the transaction will be.

H1c. The greater the compatibility between the product portfolios in a business combination of software companies, the better the new entitys nancial performance will be.

Microsoft Computer Dictionary, Fourth Edition. Redmond, WA: Microsoft Press, 1999, p. 100.

ARTICLE IN PRESS
708  P.-M. Leger, L. Quach / Technovation 29 (2009) 704713

3.4. A revisited model specic to the software industry Beyond the traditional antecedents of performance in mergers and acquisitions presented in Section 2, we propose that the specic elements in the software industry related to the potential for externalities in virtual networks should also be considered as factors explaining the performance of combinations of software companies. The following conceptual model presents an overview of the performance antecedents of software business combinations (Fig. 2).

4. Methodology To answer the research question, we tested the research model on the 60 largest (in terms of the acquisition value) combinations of public rms specializing in software (SIC 7373 and 7372) that took place in the United States between 1990 and 2003. The following sections present the operationalization of the performance variables and the independent nancial and qualitative variables. 4.1. Operationalization of performance variables The performance of each business combination was assessed over the short and long term. We used event studies to measure the short-term market performance of the acquirer rm and the target rm. An event study is a method of measuring market performance that is based on the theory of efcient and rational nancial markets (Fama, 1970). According to this theory, the effect of a signicant event is reected immediately in the value of a nancial security (MacKinlay, 1997). Event study constitutes a methodological approach that makes it possible to study how nancial markets assess the impact of an event on a rms value. This method is widely used to study the performance of business combinations as anticipated by the markets (Agrawal et al., 1992; Das et al., 1998; Lubatkin, 1987). In information technology, the approach has been used by Dos Santos et al. (1993), among others.

Event study makes it possible to calculate the abnormal return on a security at the time an event is announced. Abnormal return corresponds to the difference between the rms actual observed return and the normal return that should have been observed if the event had not occurred. Each rms normal return was estimated using the standard approach proposed by MacKinlay (1997), and the calculation was done using the Eventuss module in SASs software. The market information needed to measure abnormal return was obtained from the CRSP (Center for Research in Security Prices) database. The abnormal return was calculated for the following analysis windows: 1 day from the announcement date (1,0), 1 day to +1 day after the announcement (1,1), 3 days to +3 days (3,3), 5 days to +5 days (5,5) and 10 days to +10 days (10,10). The model for estimating the normal return was calculated over the days prior to the announcement. Long-term performance was measured by the change in various accounting indicators that are widely used to assess the protability of the new entity (Banker et al., 2004; Miller, 2000). The information that we used to measure the new entitys longterm performance included: (i) return on assets (ROA), (ii) return on equity (ROE), (iii) prot margin, and (iv) annual sales. The data needed to calculate these indicators were obtained from the COMPUSTATs (Standard & Poor) and OSIRISs (Bureau Van Dick) nancial databases. We measured the growth in these variables over periods of one year and two years after the announcement of the merger.

4.2. Operationalization of independent nancial variables The P/B ratio and the targets ex ante protability were obtained from the SDC Platinum database. More specically, the P/B ratio corresponds to the transaction value divided by the value of the target rms assets. As for ex ante performance, we used the target rms earnings per share (EPS) at the time the merger was announced, as suggested by Burger and Webster (1978).

Amount of transaction Value of target firms assets

+ + HIB and H2B : + Ex ante profitability of target firms + Financial performance of combination of software companies

Factors influencing the business combinations economic potential

Factors associated with the potential for network effects

H1A and H2A : +

Fig. 2. Proposed research model.

ARTICLE IN PRESS
 P.-M. Leger, L. Quach / Technovation 29 (2009) 704713 709

4.3. Operationalization of independent qualitative variables There is no structured data source documenting the composition of software rms product portfolios. In order to assess the characteristics of the new entities software portfolios, we did historical research in order to determine their composition at the time of the decision to merge. This historical research was based on a survey of articles published in the press at the time each merger was announced that described the makeup of the software portfolios of the parties involved. Using Proquest, we chose to collect announcements of mergers and acquisitions in specialized publications such as The Wall Street Journal, Dow Jones & Company, the Financial Times, and any other relevant newspaper articles that captured information concerning software portfolios at the time of the announcement. An average of three articles was found for each announcement. Using the coding approach suggested by Larsson (1993) and Bullock and Tubbs (1987), a coding scheme for the newspaper articles was developed in order to measure the six variables (acquisition of competencies, economies of scale, economies of scope, market growth, software compatibility and software complementarity) representing antecedents to the performance of business combinations. Appendix 1 presents the operational denition of each variable and an example given to the coder to make it easier to understand. With regard to the operationalization of the measurement of these factors, we decided to assess each variable at the time of coding on a scale of three levels of intensity: from no intensity (0) to high intensity (2). Therefore, based on the evidence gathered from the articles, the characteristics of the software portfolio resulting from each merger were evaluated with regard to the six research variables using this three-level intensity scale. The classication scheme was pretested on ten business combinations, which enabled us to rene the measurement tool. A coding guide was also created based on the recommendations of Bullock and Tubbs (1987). This document included the coding scheme and presented the necessary information to code the case studies and the process that the coder had to follow in identifying the research variables. The coding was done by two coders; one third of the coding was evaluated by both coders, with an interrater reliability rate of 80%. Any variances were discussed and recoded by consensus, as recommended by Larsson (1993). 4.4. Descriptive statistics and correlations among research variables No correlation among the different independent variables was greater than 0.5, which indicates a satisfactory level of discriminating validity among the variables (Tabachnick and Fidell, 2001). The P/B ratio, EPS, software compatibility and economies of scale were transformed to ensure that their distribution was normal.

nancial values, i.e., single indicators); (ii) it allows these indicators to be treated as dependent and independent variables in the model (i.e., P/B ratio is considered in our model as both a dependent and an independent variable); and (iii) it considers the intercorrelations within each set of criterion variables. In other words, path analysis allows one to calculate parameter estimates and t estimates for simultaneous equations. LISREL software (Submodel 2) was used to do this analysis (Joreskog and Sorbom, 1995). 5.1. Short-term model When we analyzed performance in terms of cumulative abnormal return (CAR), in the case of acquirer rms, only software compatibility seems to have a signicant positive inuence in the (3,3) window. These results support hypothesis H1a. Target rms CAR is negatively affected by the P/B ratio in all analysis windows. As previously mentioned, this is largely attributable to the perception of high P/B ratios as involving unsustainable amounts of debt. With regard to complementarity, our results revealed an unexpected impact. In the short term, complementarity has a negative impact on market performance in certain analysis windows. Thus, the results of our research do not allow us to accept hypothesis H2a. Finally, in the (3,3) and (5,5) windows, CAR is also positively impacted by economies of scope (Table 1). The results obtained are in agreement with those of previous studies; within short windows of analysis, the CAR of target rms is clearly superior to that of acquiring rms. From a nancial point of view, this is also aligned with the convergence of the stock price of the acquirer and that of the target in anticipation of the merger at a previously specied price, which normally lies between that of the acquirer and that of the target. In addition, the difference in performance in favor of target rms seems to be much more pronounced in the software industry than in other industries. Our results show that software compatibility is taken into consideration by acquirers in determining the acquisition value of the target rm. Indeed, Table 2 reveals that in addition to the target rms EPS, the acquisition of competencies, and market growth, software compatibility has a positive impact on the P/B ratio. In other words, acquiring rms are willing to pay a substantial premium for a target whose product is readily compatible with their own. This ultimately conrms hypothesis H1b. However, the results are inconclusive in terms of complementarity, so we cannot accept hypothesis H2b. These ndings indicate that the characteristics of product portfolios have an impact not on short-term market performance but on the price/book value ratio. More specically, they suggest that nancial markets tend to neglect the characteristics of product portfolios in evaluating the impact of a combination on the value of the acquirer rms shares. Thus, rms themselves seem better able to evaluate product compatibility and complementarity than nancial agents, who appear to be less sophisticated in this regard. 5.2. Long-term model In the long term (Table 3), the results show that software compatibility, economies of scale and market growth have a positive impact on sales growth. As expected, performance in terms of return on assets (ROA) and prot margin is negatively affected by P/B ratio, again due to increased debt ratios, and positively affected by the target rms EPS, the acquisition of competencies and economies of scope. Performance in terms of return on equity (ROE) is inuenced by the P/B ratio, the target

5. Analysis and discussion A path analysis using the observed variables was done to test the research model based on the data gathered on the 60 software company combinations. Path analysis is a type of multiple regression analysis belonging to the class of structural equation modeling (SEM) techniques (Joreskog and Sorbom, 1995). Path analysis is commonly used in many disciplines (e.g., psychology and marketing), and is considered especially appropriate for theory testing (Bagozzi, 1980). SEM goes beyond ordinary regression models to incorporate multiple independent and dependent variables. Path analysis is appropriate for this paper because (i) it directly uses observed variables (all our variables are

ARTICLE IN PRESS
710  P.-M. Leger, L. Quach / Technovation 29 (2009) 704713

Table 1 Inuence of performance factors on short-term market performance. CUMULATIVE ABNORMAL RETURN CAR 1,0 Beta Acquirer rms Price/book value ratio Target rm EPS Acquisition of competencies Compatibility Complementarity Economies of scale Economies of scope Market growth R2 (%) Target rms Price/book value ratio Target rm EPS Acquisition of competencies Compatibility Complementarity Economies of scale Economies of scope Market growth R2 (%) Sig. CAR 1,1 Beta Sig. CAR 3,3 Beta Sig. CAR 5,5 Beta Sig. CAR 10,10 Beta Sig.

0.23 0.07 0.03 0.17 0.19 0.11 0.1 0.15 11.69 0.32 0.16 0.08 0.04 0.17 0.13 0.09 0.07 13.73

* NS NS NS NS NS NS NS

0.18 0.05 0.19 0.15 0.26 0.07 0.07 0.08 7.87 0.37 0.1 0.002 0.03 0.23 0.26 0.17 0.08 20.36

NS NS NS NS * NS NS NS

0.11 0.0005 0.22 0.04 0.15 0.02 0.08 0.02 5.28 0.33 0.11 0.0002 0.19 0.22 0.25 0.26 0.19 26.16

NS NS * NS NS NS NS NS

0.02 0.06 0.1 0.05 0.04 0.004 0.13 0.004 3.51 0.24 0.11 0.09 0.22 0.12 0.25 0.27 0.26 22.35

NS NS NS NS NS NS NS NS

0.04 0.16 0.25 0.03 0.13 0.11 0.02 0.005 12.12 0.18 0.02 0.07 0.04 0.2 0.22 0.11 0.09 8.92

NS NS * NS NS NS NS NS

** NS NS NS NS NS NS NS

*** NS NS NS * ** NS NS

*** NS NS NS * ** ** *

** NS NS * NS * ** **

NS NS NS NS NS * NS NS

****po.001; ***po.01; **po.05; *po.10; NS: Not signicant.

Table 2 Inuence of performance factors on the value of the transaction. Price/book value ratio Gamma Price/book value ratio Target rm EPS Acquisition of competencies Compatibility Complementarity Economies of scale Economies of scope Market growth R2 (%) Sig.

Table 3 Inuence of performance factors on the new entitys performance. Sales growth 1 year Beta 2 years Sig. Beta NS NS NS ** NS ** NS *** 0.07 0.04 0.14 0.3 0.06 0.26 0.05 0.17 20.11 Return on assets 1 year Sig. Beta NS NS NS * NS * NS NS 0.62 0.28 0.32 0.01 0.25 0.01 0.26 0.08 57.82 2 years Sig. Beta Sig.

0.22 0.31 0.3 0.06 0.01 0.24 0.26 30.59

** ** ** NS NS ** **

Price/book value ratio Target rm EPS Acquisition of competencies Compatibility Complementarity Economies of scale Economies of scope Market growth R2 (%)

0.17 0.02 0.18 0.41 0.12 0.34 0.12 0.4 34.34

**** 0.52 **** *** 0.21 ** *** 0.16 NS NS 0.11 NS * 0.13 NS NS 0.01 NS ** 0.32 ** NS 0.06 NS 48.45

Return on equity 1 year 2 years Sig. Beta **** ** NS NS * NS ** NS 0.002 0.13 0.13 0.22 0.27 0.09 0.13 0.24 12.83

Prot margin 1 year Sig. Beta NS NS NS NS NS NS NS * 0.56 0.22 0.39 0.12 0.43 0.1 0.4 0.13 60.12 2 years Sig. Beta **** ** *** NS *** NS *** NS 0.34 0.19 0.04 0.31 0.07 0.19 0.3 0.29 39.18 Sig. ** * NS ** NS NS ** **

rms EPS and economies of scope. In the long term, complementarity has a negative effect on the prot margin one year after the announcement of the merger, so hypothesis H2c is not supported. Our results indicate that, in the long run, nancial factors (especially the price/book value ratio) tend to inuence the combinations performance in terms of ROA, ROE and prot margin. Nevertheless, contrary to the results of the short-term model, our analysis suggests that characteristics of the software product portfolio play a signicant role in the performance of merged software rms. More specically, software compatibility appears to be an antecedent to the new entitys long-term performance as measured by increased sales one and two years after the announcement of the merger. This supports hypothesis H1c. As for the main antecedents of the nancial performance of the business combination mentioned in the literature, the results of our study conrm that they have a positive impact, depending on the type of performance considered (ROA, ROE, increased sales or prot margin).

Beta Price/book value ratio Target rm EPS Acquisition of competencies Compatibility Complementarity Economies of scale Economies of scope Market growth R2 (%) 0.63 0.24 0.17 0.01 0.23 0.01 0.25 0.01 57.56

****po.001; ***po.01; **po.05; *po.10; NS: Not signicant.

6. Conclusion The goal of this study was to better understand the performance of mergers and acquisitions in the software sector in the short and long term. The short-term results reveal that

ARTICLE IN PRESS
 P.-M. Leger, L. Quach / Technovation 29 (2009) 704713 711

nancial markets generally seem to neglect the characteristics of software product portfolios. Nevertheless, such portfolios appear to have a positive impact on the price/book value ratio of combined software rms. In the long term, the empirical evidence presented in this paper suggests that the performance of business combinations in the software industry is connected to certain elements that are attributable to virtual network effects. The importance of software compatibility, which is associated with virtual network effects, in explaining the performance of combinations of software rms, constitutes the most important nding of this paper. Compatibility constitutes a factor with a positive inuence on the performance of the new entity in terms of increased sales in the short run after the announcement of the merger. This conrms that acquirers could potentially gain by better integrating technology issues into their M&A decisionmaking (James et al., 1998). Acquirers are fully aware of the benets associated with merging a software portfolio that is compatible, and our results show that this business potential is accounted for in the acquisition value of the target rm. However, it appears that nancial markets are more shortsighted and tend to neglect the long-term benets of software compatibility; they fail to take the potential synergy of the combined software portfolio into account when valuing the acquirer rms shares. Financial professionals focusing on long-term growth would probably benet from being better able evaluate factors that are attributable to virtual network effects. There are several limitations on this study. Only US rms were taken into consideration, because of the wealth of information available on these mergers in the documentary databases. It is not necessarily possible to generalize these results to the European and Asian markets, which may have different attitudes. The rms short-term market performance represents the nancial markets perception of the combinations potential and not its actual shortterm performance. Finally, it should be mentioned that the instruments used to measure performance factors in this study have certain inherent limitations. First of all, the measurement is based on public information or, in other words, secondary data. In addition, the qualitative variables used in this study were measured on a scale with only three levels of intensity. Although the use of a more precise scale would have allowed for greater discrimination, it would also have increased the subjectivity applied in assessing fairly general information. Furthermore, this study did not take into account organizational factors affecting post-acquisition performance. According to Chakrabarti (1990), the post-acquisition success of rms depends on the strategic t between the merging companies as well as on their organizational integration. Our results are specic to the software industry and we have no evidence that they are generalizable outside the context of this study. However, compatibility and complementarity are likely to be important drivers in other technological industries where network externalities are important, such as industries where software is embedded in equipment, as in the telecommunications sector. Investigating recent mergers and alliances in software-intensive sectors would provide a better perspective on the generalizability of the current results. This research makes contributions that are relevant to both the nancial and IT industries. First of all, the study emphasizes that it is important for nancial specialists to take the contents of a product portfolio into consideration. In the short term, nancial markets often appear to ignore certain key characteristics of product portfolios, and especially compatibility, which does in fact have an impact on the success of software business combinations in the long term. It is important for managers to better position these characteristics in the rationale they use to justify the synergy between the partners involved in a combination of software rms.

Appendix 1. Operationalization of variables (extract of coding scheme)

Variable

Operational denition By acquisition of competencies, we mean the acquisition of technical know-how or specic technologies, which are difcult to imitate or copy and which require a corresponding nancial investment.

Example

Acquisition of competencies

Software compatibility

Compatible programs are based on the same standards and require few or no investments to make them work together.

Software Two products are said complementarity to be complementary when their joint use adds more value for the customer than the sum of the values of the separate use of these products.

Economies of scale

Economies of scale refer to the reduction in average per-unit costs because of increased volume or rationalization of operations.

SilverStreams product portfolio, particularly its J2EE-based application server, will be a linchpin of Novells transformation into an e-business solutions and platform company. (Novell gets J2EE boost from SilverStream deal, Elizabeth Montalbano. CRN. Jericho: June 17, 2002, No. 1000; p. 8) Like most Web services companies, SilverStreams development tools are based on Suns Java computing system, which makes it easier to write code that runs on many different operating systems. (Transforming Novell SilverStream deal moves rm deeper into web software, Hiawatha Bray. Boston Globe, June 11, 2002; p. C1) The goal is to integrate BMCs Patrol reporting tools with BGSs Best/I analysis tools. The combined tools would let information systems managers detect and predict problems with business-critical servers and applications. (BMC buys BGS systems to help IS maintain service, Patrick Dryden. Computerworld. Framingham: Feb. 9, 1998. Vol. 32, No. 6; p. 24) Rather than trying to integrate the companies systems and technologies, which makes such mergers complex, Oracle will simply milk ongoing revenues from PeopleSofts customers, transfer the best

ARTICLE IN PRESS
712  P.-M. Leger, L. Quach / Technovation 29 (2009) 704713

Economies of scope

Market growth

salesmen, developers and technology to Oracle, and close much of the rest down (The soft sell, Financial Times. London: June 7, 2003. p. 16) Economies of scope Network Associates, are dened as an formerly called McAfee increase in efciency Associates, said it will resulting from the use TIS software in a expansion of the range suite of programs, of goods or services similar to the way produced by the Microsoft Corp. has company. bundled collections of personal computer software (Network associates to pay $300 million to buy trusted information systems, Don Clark. Wall Street Journal. New York: Feb. 24, 1998. p. 1) Growth refers to the The opportunity to concept of increased gain small and midsize market share and users was also a factor power. for Peregrine whose 46,000 customers are primarily large companies (Peregrine swoops in to buy Remedy, Ann Sullivan. Network World. Framingham: June 18, 2001. Vol. 18, No. 25; p. 16)

References
Agrawal, A., Jaffe, J.F., Mandelker, G.N., 1992. The post-merger performance of acquiring rms: a re-examination of an anomaly. The Journal of Finance 47, 16051621. Bagozzi, R.P., 1980. Causal Models in Marketing. Wiley, New York. Banker, R.D., Chang, H., Janakiraman, S.N., Konstans, C., 2004. A balanced scorecard analysis of performance metrics. European Journal of Operational Research 154, 423436. Bannert, V., Tschirky, H., 2004. Integration planning for technology intensive acquisitions. R & D Management 34, 481494. Bayona, C., Corredor, P., Santamaria, R., 2006. Technological alliances and the market valuation of new economy rms. Technovation 26, 369383. Brouthers, K.D., Van Hastenburg, P., Van Den Ven, J., 1998. If most mergers fail why are they so popular?. Long Range Planning 31, 347353. Bullock, R.J., Tubbs, M.E., 1987. The Case Meta-Analysis Method for OD. Research in Organizational Change and Development. JAI Press, Greenwich. Burger, A.D., Webster, S.K., 1978. The management accountant looks at EPS vs. ROI: conict in measuring performance. Management Accounting. (pre-1986) 60, 1924. Cabral, L.M.B., 2001. Optimal brand umbrella size. Working Papers 01-06, New York University, Leonard N. Stern School of Business, Department of Economics. Capron, L., 1999. The long-term performance of horizontal acquisitions. Strategic Management Journal 20, 9871018.   Capron, L., 2005. Les beneces et les risques des acquisitions horizontales. Les  echos 2005, lart de la strategie, /http://www.lesechos.fr/formations/strate gie/articles/article_7_7.htmS [Accessed December 4, 2005]. Carlaw, K.I., Lipsey, R.G., 2002. Externalities, technological complementarities and sustained economic growth. Research Policy 31, 13051315. Chakrabarti, A.K., 1990. Organizational factors in post-acquisition performance. IEEE Transactions on Engineering Management 37, 259268. Das, S., Sen, P.K., Sengupta, S., 1998. Impact of strategic alliances on rm valuation. Academy of Management Journal 41, 2741.

Datta, D.K., Pinches, G.E., Narayanan, V.K., 1992. Factors inuencing wealth creation from mergers and acquisitions: a meta-analysis. Strategic Management Journal 13, 6784. Dos Santos, B.L., Peffers, K., Mauer, D.C., 1993. The impact of information technology investment announcements on the market value of the rm. Information Systems Research 4, 123. Economides, N., 1991. Compatibility and Market Structure. New York University, Leonard N. Stern School of Business, Department of Economics. Economides, N., 1996. The economics of networks. International Journal of Industrial Organization 14, 673699. Economides, N., 2001. The Microsoft antitrust case. Journal of Industry, Competition and Trade 1, 739. Ende, J.V.D., Wijnberg, N., 2003. The organization of innovation and market dynamics: managing increasing returns in software rms. IEEE Transactions on Engineering Management 50, 374382. Fama, E.F., 1970. Efcient capital markets: a review of theory and empirical work. The Journal of Finance 25, 383417. Farrell, J., Saloner, G., 1986. Installed base and compatibility: innovation, product preannouncements, and predation. The American Economic Review 76, 940955. Gammelgaard, J., 2004. Access to competence: an emerging acquisition motive. European Business Forum 17, 4447. Graebner, M.E., 2004. Momentum and serendipity: how acquired leaders create value in the integration of technology rms. Strategic Management Journal 25, 751777. Haro-Dominguez, M.D.C., Arias-Aranda, D., Llorens-Montes, F.J., Moreno, A.R., 2007. The impact of absorptive capacity on technological acquisitions engineering consulting companies. Technovation 27, 417425. Harrison, J.S., Hitt, M.A., Hoskisson, R.E., Ireland, R.D., 1991. Synergies and postacquisition performance: differences versus similarities in resource allocations. Journal of Management 17, 173190. Harrison, J.S., Hitt, M.A., Hoskisson, R.E., Ireland, R.D., 2001. Resource complementarity in business combinations: extending the logic to organizational alliances. Journal of Management 27, 679690. Hollender, J.S., 1967. The ne art of valuing and nancing business acquisitions. New York Certied Public Accountant (pre-1986) 37, 589. James, A.D., Georghiou, L., Metcalfe, J.S., 1998. Integrating technology into merger and acquisition decision making. Technovation 18, 563573. Joreskog, K.J., Sorbom, D., 1995. Submodel 2: Causal Models for Directly Observed Variables. LISREL 8, Users Reference Guide. SSI (Scientic Software International), pp. 133-158. Katz, M.L., Shapiro, C., 1985. Network externalities, competition, and compatibility. The American Economic Review 75, 424440. Kauffman, R.J., Li, X., 2005. Technology competition and optimal investment timing: a real options perspective. IEEE Transactions on Engineering Management 52, 1529. King, D.R., Dalton, D.R., Daily, C.M., Covin, J.G., 2004. Meta-analyses of postacquisition performance: indications of unidentied moderators. Strategic Management Journal 25, 187200. Ku, K.-C., Kao, H.-P., Garumurthy, C.K., 2007. Virtual inter-rm collaborative frameworkan IC foundry merger/acquisition project. Technovation 27, 388401. Kusewitt Jr., J.B., 1985. An exploratory study of strategic acquisition factors relating to performance. Strategic Management Journal 6, 151169. Larsson, R., 1993. Case survey methodology: quantitative analysis of patterns across case studies. Academy of Management Journal 36, 15151546. Larsson, R., Finkelstein, S., 1999. Integrating strategic, organizational, and human resource perspectives on mergers and acquisitions: a case survey of synergy realization. Organization Science 10, 126. Lehto, E.L.O., Lehtoranta, M.O., 2004. Becoming an acquirer and becoming acquired. Technological Forecasting and Social Change 71, 635650.  Leroy, F., 2003. Fusions-Acquisitions: Strategie et Mise en Oeuvre. Fusions  Acquisitions: Les des de lintegration. Institut de lentreprise, Paris. Lubatkin, M., 1987. Merger strategies and stockholder value. Strategic Management Journal 8, 3953. MacKinlay, A.C., 1997. Event studies in economics and nance. Journal of Economic Literature 35, 1339. Miller, J.K., 2000. Determinants inuencing the protability of nancial intermediaries; efciency through economies of scale or market power: An empirical analysis. Ph.D. Thesis, Nova Southeastern University, Florida. Paruchuri, S., Nerkar, A., Hambrick, D.C., 2006. Acquisition integration and productivity losses in the technical core: disruption of inventors in acquired companies. Organization Science 17, 545562. Pautler, P.A., 2003. Evidence on mergers and acquisitions. Antitrust Bulletin 48, 119221. Pehrsson, A., 2006. Business relatedness and performance: a study of managerial perceptions. Strategic Management Journal 27, 265282. Prabhu, J.C., Chandy, R.K., Ellis, M.E., 2005. The impact of acquisitions on innovation: poison pill, placebo, or tonic?. Journal of Marketing 69, 114130. Priest, W.C., 1994. The Character of Information: Characteristics and Properties of Information Related to Issues Concerning Intellectual Property. Center for Information, Technology, and Society. Puranam, P., Srikanth, K., 2007. What they know vs. what they do: how acquirers leverage technology acquisitions. Strategic Management Journal 28, 805825. Seth, A., 1990a. Sources of value creation in acquisitions: an empirical investigation. Strategic Management Journal 11, 431446.

ARTICLE IN PRESS
 P.-M. Leger, L. Quach / Technovation 29 (2009) 704713 713

Seth, A., 1990b. Value creation in acquisitions: a re-examination of performance issues. Strategic Management Journal 11, 99115. Shaver, J.M., 2006. A paradox of synergy: contagion and capacity effects in mergers and acquisitions. Academy of Management Review 31, 962976. Silva, L., Iyer, B., 2006. Using Software Stacks to Explain Complementarities: The Case of Mergers and Acquisitions in the Software Industry. HICSS. Singh, H., Montgomery, C.A., 1987. Corporate acquisition strategies and economic performance. Strategic Management Journal 8, 377386. Sorescu, A.B., Chandy, R.K., Prabhu, J.C., 2007. Why some acquisitions do better than others: product capital as a driver of long-term stock returns. Journal of Marketing Research 44, 5772. Tabachnick, B.G., Fidell, L.S., 2001. Using Multivariate Statistics. Allyn and Bacon, Boston.

Tsai, Y.T., Hsieh, L.F., 2006. An innovation knowledge game piloted by merger and acquisition of technological assets: a case study. Journal of Engineering and Technology Management 23, 248261. Tuch, C., OSullivan, N., 2007. The impact of acquisitions on rm performance: a review of the evidence. International Journal of Management Reviews 9, 141170. Varaiya, N.P., 1987. Determinants of premiums in acquisition transactions. Managerial and Decision Economics 8, 175184. Wang, L., Zajac, E.J., 2007. Alliance or acquisition? A dyadic perspective on interrm resource combinations. Strategic Management Journal 28, 12911317. Yoo, Y.J., Lyytinen, K., Heo, D., 2007. Closing the gap: towards a process model of post-merger knowledge sharing. Information Systems Journal 17, 321347.