1
Alliance Portfolio Size & Diversity: the Moderating Role of Alliance
Experience on Firm Performance
Master’s Thesis
Ma Business Administration: Strategic Innovation Management David Kuiper (S1731920) Davidpeterkuiper@gmail.com Date: 23-06-2014 Word count: 11.824 Supervisor: Dr. P. de Faria Co-assessor: B. Bos
This study examines how alliance portfolio size and diversity influence firm performance. Addressing the call for multi-domain alliance portfolio research this study offers a portfolio diversity construct that includes functional and industry diversity. I further examine whether firms learn to manage functional and industry portfolio diversity. Building on prior work it is argued that higher functional diversity will positively affect firm performance; industry diversity will show an inverted U-shaped relation to firm performance; and high levels of experience are expected to enhance the positive effect of diversity on firm performance. To test the hypotheses information from the Thomson SDC database, Orbis database, and CompuStat database was used to create a panel dataset (1998 – 2006). This dataset contains alliance and financial data on 50 large firms in the Biotech and Pharmaceutical industry. Bringing together literature on portfolio size, diversity, and experience, the results of this study show that portfolio size has a regular U-shaped relationship with firm performance. Industry diversity was found to have a linear negative impact and limited evidence was found for a negative influence of functional diversity on firm performance. Past experience moderates the functional diversity-performance relationship positively, but it does not affect the industry diversity-performance relationship. I suggest that firms should carefully consider size and diversity as more is not always better. Furthermore, as not all domains of diversity can be learned by experience other learning mechanisms should be considered.
Key words: Alliance portfolios, portfolio size, functional diversity, industry diversity, alliance
2
ACKNOWLEDGEMENTS
I see this master thesis as the final proof of competence for obtaining a Master of Science degree in Business Administration. During the master program, Strategic Innovation Management, my interest for collaborative efforts between firms was triggered. As a result I wrote this master thesis on the influence of alliance portfolio size and diversity on firm performance. The setting for the study was the Biotech and Pharmaceutical industry. This industry and its firms are well-known for their high rate of collaboration, making it a very suitable setting for my study. I have enjoyed doing the research and it has helped me to further understand the dynamics involved in inter-firm collaboration.
I want to use this opportunity to show my gratitude to the people that have directly and indirectly helped me writing this thesis. First of all, my thesis supervisor Pedro de Faria, Pedro has always given me very useful and elaborate feedback. During the meetings Pedro de Faria proved to be very knowledgeable, professional and social. This mix of professionalism and a personal touch motivated me and gave me confidence. Second, I want to thank Aneta Oleksiak because we have discussed and resolved many difficulties together. Lastly, I want to thank my girlfriend and my family for their everlasting support.
INTRODUCTION
In today’s business landscape firms are increasingly involved in multiple strategic alliances with different partners at the same time (Gulati, 1998). In many key industries like telecommunications, electronics, pharmaceuticals, and biotech strategic alliances have become essential to the strategy of firms (Wassmer, 2010). Although popular as a potential value-creating strategic option, alliances still face high failure ratings (Reuer, 1999; Lunnan & Haugland, 2008). It is therefore important to understand the factors influencing the success of alliances and alliance portfolios. Examining and understanding the complex interplay of these factors will allow managers to make better decisions to enhance portfolio performance. Furthermore, it enables firms to construct their portfolios in a more effective and efficient manner which in turn results in enhanced overall firm performance.
3 Within the alliance portfolio research stream many sub-fields have emerged examining different characteristics of alliance portfolios. Numerous studies have focussed on portfolio size, mostly finding curvilinear relationships with firm performance (Deeds & Hill, 1996). However, more recent research suggests that portfolio size alone is not a sufficient predictor of firm performance and argue that other configuration factors should be taken into account (Ahuja, 2000; Baum, Calabrese & Silverman, 2000). A prominent stream of research in the portfolio configuration research that is argued to complement portfolio size literature focuses on alliance portfolio diversity (APD) (Duysters & Lokshin, 2011). Prior studies in this field examined the diversity of a portfolio along dimensions such as functional purpose, industrial background, geographical location, and governance form. Most empirical studies find contradicting results on how APD influences firm performance (e.g. Baum et al., 2000; Duysters & Lokshin, 2011). Furthermore, till date the majority of studies conceptualized APD along a single-domain of diversity, missing out on important factors of other diversity domains. This calls for a multi-domain approach that conceptualizes APD along multiple diversity domains (Wassmer, 2010) to develop a more comprehensive overview of the diversity-performance relationship.
Nearly in complete isolation of the APD literature, another stream of alliance research focuses on the effect of past alliance experience on performance. Typically, such studies are based on organizational learning theory and try to answer the question whether experience allows firms to learn to manage alliances (Anand & Khanna, 2000). So far only Duysters & Heimeriks (2012) addressed the question whether firms can learn to manage APD. They conceptualized APD by combining multiple diversity domains into one measure. They find empirical evidence for a positive moderating effect of past alliance experience on the diversity-performance relationship. The downside of this approach is that it does not allow them to examine the effect of alliance experience on specific diversity domains separately.
4 hypothesized to moderate the diversity-performance relationship in such a way that higher levels of experience will increase the positive effect of diversity on firm performance.
In order to test for the hypothesized relationships a dataset consisting of panel data on 50 firms that have almost 1500 alliances among them is constructed. The Thomson SDC database was used to construct alliance portfolios of 50 large biotech and pharmaceutical firms within the 1998 – 2006 timeframe. Financial data was collected from the Orbis and CompuStat databases. As result of the Hausman-test a GLS Random-effects statistical method was used to analyse the data and estimate the models.
The results of this study show that portfolio size has a U-shaped relationship with firm performance instead of the expected inverted U-shaped relationship. Limited evidence was found to claim that functional diversity has a linear negative impact on firm performance in contrary to the hypothesized positive impact. Industry diversity was predicted to have an inverted U-shaped relationship but the empirical evidence shows that the relationship is negative and linear. Past experience is found to significantly benefit functional diversity positively and not to affect industry diversity. This indicates that experience influences the different domains that constitute APD dissimilarly.
My study contributes to the existing literature on alliance portfolios by combining the research streams on portfolio size and multi-domain portfolio diversity. This allows for a better understanding of the intricate interplay between these portfolio configuration factors. Furthermore, this study is one of the first to bring together the two emergent literature streams on portfolio diversity and alliance experience. It elaborates on prior work (Duysters & Heimeriks, 2012) and sheds new light on the moderating effect of past alliance experience on the diversity-performance relation. These are interesting contributions because they show that the diversity domains should be considered separately in order to determine the specific effect experience has on each of the diversity domains individually.
5
THEORY & HYPOTHESES
Firms enter into strategic alliances for many reasons such as access to complementary resources (Eisenhardt & Schoonhoven, 1996), capabilities, or knowledge, and to share costs and risks (Hagedoorn, 1993). When effective, alliances can be growth and profitability engines in both domestic and global markets (Ernst, Halevy, Monier & Sarrazin, 2001). This has caused alliances to become an increasingly popular strategy for firms (Dyer, Kale & Singh, 2001). Traditional alliance research has predominantly concentrated on single alliances, mainly focussing on the formation, governance, evolution, and performance of single alliances. However, in today’s business landscape firms are increasingly found to be simultaneously engaged in multiple strategic alliances (Gulati, 1998). Consequently, firms are ever more facing the challenge of managing an entire alliance portfolio (e.g. Anand & Khanna 2000; Gulati, 1998; Ozcan & Eisenhardt, 2009). As a result, more recently a research stream on alliances portfolios emerged, focussing on the engagement of firms in a wide array of strategic alliances (Wassmer, 2010). In line with prior research this study defines an alliance portfolio as ‘all strategic alliances a firm is engaged in at a certain time’ (Hoffmann, 2007; Lavie, 2007). In this study I will delve into the matter of alliance portfolio size, diversity and performance. A multi-domain approach of diversity will allow it to go beyond most prior work that adopted a single-domain approach.
Alliance Portfolios
In the alliance portfolio literature three main streams of research were identified (Wassmer, 2010) including the emergence of alliance portfolios, the configuration of alliance portfolios, and alliance portfolio management. Within the alliance portfolio configuration research stream the study of APD has received strong attention (Duysters & Heimeriks, 2012). APD is a complex concept that can comprise multiple dimensions including (a) a size dimension (Ahuja, 2000; Hoffmann, 2007) (b) a structural dimension (Ahuja 2000; Hoffmann, 2007) (c) a relational dimension (Hoffmann, 2007) (d) a partner dimension (Lavie, 2007). Complexity increases as these dimensions span across different levels of analysis (Wassmer, 2010).
6 Jiang, Tao & Santoro, 2010). These contradicting findings can partly be explained by the argument that maintaining strategic alliances firms incur not only benefits but also costs (Park & Zhou, 2005). The costs and benefits related to alliance portfolios are contingent on how efficient an alliance portfolio is configured (Baum et al., 2000). Consequently, firms face trade-offs when they increase the diversity of their portfolios. More diversity allows for greater search options and access to larger resource pools, leading to possibilities to create more value and to enhance capability development. On the other hand, higher diversity brings more complexity and potential for conflict. This can result in increased managerial and coordination costs (Jiang et al., 2010).
Till date most studies have focussed on a single dimension of alliance portfolio configuration, missing out on other important factors from other dimensions (Wassmer, 2010). Many of these studies conclude that a single dimension, for example size, alone is not a sufficient predictor of performance (Ahuja, 2000). To better understand the impact of portfolio size on performance other configuration factors must be considered. This urges the need for multi-dimension approaches to develop a better understanding of the relationship between portfolio configuration and performance (Wassmer, 2010). Since size alone is not a sufficient predictor of performance, industrial relatedness and functional purpose will be included to conceptualize APD. In this study I therefore define APD as ‘the degree of variance in industrial background and functional purpose of the alliances in a portfolio’.
7 of domains and do not test whether there are differences between the influences of experience on the domains separately.
Portfolio Size
Alliance portfolio size has long been subject of discussion in the alliance portfolio configuration research stream. Some scholars depict that size affects partner behaviour (Amaldoss & Staelin, 2010) where others focus on the size-performance relationship (Shan, Walker & Kogut, 1994; Deeds & Hill, 1996). A variety of theoretical lenses, such as the Resource-Based View (Barney, 1991) and Social Network Theory (Granovetter, 1985), have been used in portfolio size research by scholars. A number of these studies find a significant (positive) relationship between portfolio size and differences in performance (Shan et al., 1994; Deeds & Hill, 1996). Where Shan et al. (1994) find a positive linear relationship, Deeds & Hill (1996) illustrate that there are diminishing returns from adding more alliances to a portfolio, resulting in a turning-point after which adding more alliances will negatively influence performance. In other words, Deeds & Hill (1996) find an inverted U-shaped relationship between portfolio size and firm performance.
This initially positive relationship can be explained by a number of reasons. First, according to the Knowledge-based view (Grant, 1996) heterogeneous knowledge bases and capabilities of a firm determine its performance. Firms with a larger alliance portfolio have access to a wider variety of knowledge and resources from external sources, thus potentially leading to better performance. Moreover, building on the Resource-based view, a larger set of alliances in a portfolio increases the ability of a firm to access valuable resources that would not be available without these specific alliance partners (Das & Teng, 2000). By effectively combining the complementary resources available through partners, the focal firm can enhance its performance and its sustainable competitive advantage (Sivakumar, Roy, Zhu & Hanvanich, 2011). Second, firms with larger portfolios are found to better respond to external changes by effectively responding to uncertainties (Duysters & De Man, 1999). Furthermore, increased portfolio size is argued to help mitigate risks that result from the external environment (Hoffmann, 2007). Third, a larger portfolio increases the opportunity to lower costs for the focal firm through utilizing cost-effective technologies from partners (Bettis, Bradley & Hamel, 1992) and the potential for economies of scale from multiple alliance partners (Gilley & Rasheed, 2000).
8 of alliance portfolio size. These diminishing returns can eventually lead to disadvantages of adding more alliances to the portfolio and result from different factors. First, firms pursue the most promising alliances early on as they will yield most benefits. This leaves less potent alliances options of which the firm can choose from, resulting in less beneficial alliance possibilities later on (Rothaermel, Hitt & Jobe, 2006). In other words, these less potent possibilities bring little valuable knowledge and resources. Therefore, the gains of such collaboration can be smaller than the costs they bring. Furthermore, it becomes increasingly difficult to find partners that bring a variety of valuable knowledge and resources as portfolio size increases. This difficulty to find and select the right partners gives rise to search costs. These search costs can eventually outweigh the benefits of increasing the size of the alliance portfolio, leading to worse firm performance (Laursen & Salter, 2006).
Second, as commonly illustrated in economies of scale theory (Stigler, 1958), there is an optimal size after which economies of scale inflect into diseconomies of scale. Therefore, firms should seek to find the optimal portfolio size because having a too large alliance portfolio will likely result in diseconomies of scale that negatively influence firm performance.
Third, as portfolio size increases the ability of managers to effectively manage and exploit alliances in order to obtain benefits will decrease as it increases the demand on managers’ attention and bounded rationality (Rothaermel, 2001). This is where the so-called ‘attention allocation problem’ emerges (Laursen & Salter, 2006). This problem is one of the key concepts in Attention-based theory of the firm (Ocasio, 1997). Attention-based theory argues that managerial attention is the most important resource inside a firm and that the allocation of attention to specific activities is the most important factor in explaining firm performance. The implication in the case of a very large alliance portfolio is that the alliances in the portfolio do not receive the attention needed to fully exploit their benefits.
Fourth, bureaucracy and transactions costs increase within larger alliance portfolios according to Rothaermel (2001). This can be ascribed to problems arising from simultaneously coordinating with multiple partners. Initially, at lower levels of alliance portfolio size the benefits are larger than the costs. However, when a portfolio becomes very large it will reach a point after which the transaction costs rise beyond a point where gains from additional alliances are outweighed by their costs (Jones & Hill, 1988).
9 becomes very large. Therefore, I expect that the size of an alliance portfolio will first be positively associated with firm performance, but after a certain point this association will become negative. This results in an inverted-U shaped relationship between alliance portfolio size and firm performance. Therefore, this study hypothesizes the following:
H1: The relationship between alliance portfolio size and the performance of a firm follows an inverted U-shaped curve.
As mentioned before, the relationship between alliance portfolios and performance is argued to be influenced by other factors than only portfolio size (Ahuja, 2000, Baum et al., 2000). Therefore, the functional purpose of the alliances together with industrial background and alliance portfolio size are combined in this multi-domain study as this is argued to give a more complete explanation of the relationship between APD and performance (Wassmer, 2010).
Functional diversity
Alliances in a portfolio can vary in their functional purpose. The functional activities of an alliance can be focussed towards marketing, manufacturing, distribution, or R&D (March, 1991). The purpose of marketing, manufacturing and distribution alliances can be to broaden market reach, enhance value creation, and exploit core competencies (Prahalad & Hamel, 1990). R&D alliances on the other hand, are often seen in industries that are characterized by rapid technological change and high risk (Hagedoorn, 1993). An alliance can serve multiple functional purposes but mostly a focus on one functional purpose prevails. Since alliance portfolios consist of multiple alliances that may represent similar or different functional purposes (Cui & O’Connor, 2012) alliance portfolio functional diversity is defined in this study as ‘the degree to which the alliances in a portfolio vary in functional purpose’. In an alliance portfolio, the more the individual alliances have a different functional purpose, the higher the functional diversity.
10 knowledge in order to generate short-term financial returns (Koza & Lewin, 1998; Rothaermel, 2001; Rothaermel & Deeds, 2004). Exploration activities are more focussed towards enhancing innovation, acquiring specific capabilities and knowledge in order to generate long-term financial results and competitive advantage (March, 1991). March (1991) further argues that it is important for firms to balance exploration and exploitation as both short-term productivity and long-term innovation are essential for organizational success and survival (March, 1991: 87). Relying too much on either exploration or exploitation can have negative consequences for firms. Organizations that focus on exploration but disregard exploitation can end up with undeveloped ideas and unrealized opportunities. Consecutively, focussing solely on exploitation and neglecting exploration may deplete firm’s opportunities and leave their competencies out-dated (March, 1991).
A recent study of Cui & O’Connor (2012) proposes a somewhat different and less positive view. They argue that from an absorptive capacity viewpoint functional diversity may increase the difficulty of information and resource sharing. This will eventually lead to worse innovative performance as difficulties with information and resource sharing lowers the contribution of resource diversity to innovative performance. Although, information and resource sharing apply strongly to innovative performance it is not certain what their effect is on financial performance (Cui & O’Connor, 2012). Furthermore, Oxley & Sampson (2004) found that increased functional diversity hinders inter-partner communication, resulting in managerial difficulties and higher transaction costs (Faems et al., 2010). This implies that increased functional diversity possibly negatively influences firm performance.
Firms that simultaneously focus on exploration and exploitation are expected to perform better compared to firms that solely focus on exploration or exploitation (Tushman & O’reilly, 1996). Therefore, firms should seek to balance alliance partners with different functional purposes in an alliance portfolio. This will allow firms to leverage current knowledge, resources and opportunities and simultaneously explore new knowledge and opportunities. Jiang et al. (2010: 1139) add to this that as a firm increases the functional diversity of its portfolio, it builds a more balanced portfolio that incorporates core and noncore activities, gains access to supplementary and complementary assets, and expands its knowledge base and market reach. This balanced approach will lead to enhanced overall firm performance.
11 & O’Reilly, 1996) it is expected that the benefits of increased functional diversity outweigh the potential costs. Therefore, I propose the following hypothesis:
H2a: Greater alliance portfolio functional diversity is positively associated with firm performance.
Industry diversity
Alliances can be formed with partners who are active either in the same or in different industries. Depending on differences in industrial background the degree of industry diversity of alliance portfolios can be determined. In this study alliance portfolio industry diversity is defined as ‘the degree to which the alliances in a portfolio vary in industrial background.’ Allying with partners from the same industry (intra-industry alliances) brings a mass of both benefits and drawbacks. This also holds for partnering with firms that have a dissimilar industrial background (inter-industry alliances).
Partners that are active in the same industry are often competitors (Jiang et al., 2010). Cohen & Levinthal (1990) point towards benefits of intra-industry alliances and argue that partnering with competitors, thus firms from the same industry, bring greatest learning opportunities through imitation and greater absorptive capacity because of overlap in knowledge, experience, background, and technology bases. Other studies show that allying with firms from the same industry reduces coordination costs (Luo & Deng, 2009). However, corresponding with Social Network Theory (Granovetter, 1985), it is argued that when a portfolio consists of very similar partners, adding another similar partner will hardly have additional value (Luo & Deng, 2009) because a similar partner is unlikely to possess new valuable knowledge (Grant, 1996). Therefore, problems of redundancy are expected to occur as new combinations of existing knowledge will start to deplete (Vasudeva & Anand, 2011). Furthermore, conflicting interests and learning races can arise when allying with competitors and this will increase monitoring and safeguarding costs (Doz & Hamel, 1998).
12 monitoring and safeguarding costs. This, however, does not automatically imply that overall costs are lower because other costs arise as more complex coordination mechanisms are needed (Rothman & Kraft, 2006). These coordination costs increase because of complexity to coordinate scarce resources among multiple partners and the difficulties associated with aligning strategic goals with partners (Hoang & Rothaermel, 2005).
Since both inter-industry and intra-industry alliances come with many benefits and drawbacks firms have to be careful how to configure their portfolio to maximize financial performance. On the one hand lowering diversity by having same industry alliances eases learning but knowledge is likely to be less novel and valuable. On the other hand, increased diversity from inter-industry alliances potentially provides more valuable novel knowledge but learning will be more difficult. Besides learning effects, there are costs associated with both types of alliances which need to be taken into account. Having too many competitive alliances raises monitoring and safeguarding costs, while inter-industry alliances will face higher costs arising from complex coordination mechanisms. Consequently, firms are likely to benefit most from a balanced portfolio that is not focussed too much on either inter-industry or intra-industry alliances. In line with this reasoning I propose the following hypothesis: H2b: Greater alliance portfolio industry diversity is associated with firm performance that first increases and then decreases, forming an inverted U-shape.
Alliance Experience
In the alliance portfolio literature there is an on-going debate whether to include alliances that have become inactive or not (Wassmer, 2010). This debate can be seen as a result of the different theoretical viewpoints scholars take. Especially from a learning perspective past alliances that have become inactive can still play a significant role. The lessons learned through the experience with alliances in the past might well influence future performance. In this study I build on the learning perspective to determine whether experience with past, inactive, alliances allow firms to learn to better manage their current portfolio of alliances. Therefore, past alliances are not seen as an actual part of the current portfolio but are nevertheless believed to affect firm performance.
13 better in future alliances. This perception originates from the traditional learning curve (TLC) perspective (Penrose, 1959; Anand & Khanna, 2000; Hoang & Rothaermel, 2005). The TLC perspective is built on the assumption that companies learn from their experience and enhance performance by repetition of actions (Reuer, Park & Zollo, 2002). It further assumes that learning effects are always positive and it equates learning with experience. This argues for the expected indefinitely positive effects of experience on future alliance performance.
However, within the literature there is also evidence that poses a rather different view on the effect of experience. Draulans, DeMan & Volberda (2003) for example, also acknowledge experience as an antecedent of alliance performance but they add the notion that there is a limit to learning-by-doing. This view is supported by other empirical studies claiming that the value of experience decays over time (Baum & Ingram, 1998; Darr, Argote & Epple, 1995; Argote, Beckman & Epple, 1990). The decay of the value of experience suggests that the benefits from experience do not indefinitely accumulate over time.
Jiang et al. (2010: 1138) already hint towards the possibility that firms can become more adept in dealing with portfolio diversity. Duysters & Heimeriks (2012) are the first to actually link alliance experience with portfolio performance. They find that past alliance experience has a moderating effect on the relationship between APD and portfolio performance. Higher levels of APD are associated with increasing difficulties to manage the alliance portfolio. However, the empirical evidence they find in their study illustrates that these difficulties can be mitigated by higher levels of past experience (Duysters & Heimeriks, 2012). Although they conceptualize APD in a slightly different manner and look at portfolio performance instead of firm performance, I expect that the positive moderating relationship also applies to this research. Therefore, based on the above arguments and reasoning I propose the following hypotheses:
H3a: The positive effect of functional diversity on firm performance is stronger if firms have high levels of alliance experience.
H3b: The positive effect of industry diversity on firm performance is stronger if firms have high levels of alliance experience.
METHODS
14 2836) for the time period of nine years (1998 – 2006). The firms included in the dataset were mainly selected based on their size. Only large organizations in the biotech and pharmaceutical industry were considered. The reason to focus on the largest firms in the industry is that they are a good representation for the total industry and it will likely reduce heterogeneity due to size differentials. Firms that did not have an alliance portfolio of at least two alliances in all years of interest (2002 – 2006) were excluded from the dataset. Further, one firm and its portfolio of alliances was excluded as this firms was acquired by another firm in 2004, making it impossible to find financial data for the years after 2004. This leaves the study with a sample of 50 firms that have a total of 1483 alliances between them. The focus on the biotech and pharmaceutical industry helps to decrease problems that arise due to heterogeneity across industries. The biotech and pharmaceutical industry is argued to be a suitable industry for research in the field of open innovation because it has high levels of inter-firm collaborations (Hagedoorn, 2002).
The data-set includes information from three main sources: The Securities Data Company (SDC) Database on joint-ventures and alliances, the CompuStat Database, and the Orbis Database. The SDC Database contains information on many types of alliances. The database is mainly compiled from publicly available sources. It is one of the most complete databases on alliances and has extensively been used by prior research on alliances (e.g. Sampson, 2007; Anand & Khanna, 200). In order to increase reliability of the SDC data no alliances from before 1991 where considered in this research as coverage of alliances in the SDC Database is more comprehensive from 1990 (Sampson, 2007). The SDC data was complemented with financial data and firm size data from the CompuStat and Orbis databases that both contain highly detailed (financial) information.
The data that was obtained from the SDC Database had to be transferred from text format to Excel format. Furthermore, some of the variables in the dataset had to be modified and/or recoded. Variables consisting of string variables such as ‘yes’ or ‘no’ were recoded into numeric values in order to make them suitable for statistical analysis. The financial data from the Orbis and CompuStat databases were aggregated in order to obtain a full overview of financial performance of the firms in the sample. The reliability of the data obtained from Orbis and CompuStat was verified by examining the financial annual reports of all the firms in the sample. More detail on the process of coding the variables of diversity will be given in a later section.
15 average an alliance lasts for five years (Sampson, 2007) and the first four years are required to build a full overview of a firm’s portfolio starting in 2002. This longitudinal approach allows this study to also examine the effects of prior alliance experience on the relationship between APD and firm performance.
Dependent variable
In line with prior research Net Profit Margin (NPM) will be used to measure firm performance (Faems et al., 2010: Jiang et al., 2010). The NPM of a firm refers to a measure of profitability that is calculated by finding the net profit as a percentage of the revenue. Net profit margin is used as it argued to be very useful to compare corporate performance in the same industry. An average of the NPM of the firms in the sample is calculated over a three year time period in order to decrease the possibility that accidental extraordinary results will influence the results of this study (Jiang et al., 2010).
Independent variables
In this study the independent variables consist of portfolio size together with APD. There are many different ways to measure APD and this can be done on various levels of analysis (Wassmer, 2010). This study measures APD along two dimensions of diversity namely functional diversity and industry diversity. These two dimensions are both measured along the alliance portfolio level of analysis.
First, functional diversity is a result of the degree to which alliances in a portfolio differ in terms of functional purpose. The functional purpose of an alliance can vary between (1) marketing, (2) manufacturing, (3) distribution, and (4) R&D. In line with Jiang et al. (2010) this study codes functional activities as ‘1’ for marketing, ‘2’ for manufacturing, ‘3’ for distribution and other alliance types, and ‘4’ for R&D. Alliance that had multiple functional purposes according to the SDC Database were classified by only one function. The alliance specific text in the SDC database was analysed to determine the main purpose of the alliance and the alliance was classified accordingly.
16 Further, alliance portfolio size is measured by the number of alliances a firm has in its portfolio over a five-year period. The SDC database does not include data on when an alliance becomes inactive, which makes it difficult to assess the precise size of a portfolio. A period of five years is generally accepted as the average length over which an alliance is active in the literature. Therefore, in line with prior research (e.g. Sampson, 2005), this study assumes that an alliance is active for five years. In this way, by accumulating the alliances a firm started over a five year period, alliance portfolio size for a given year can be computed. For example, the portfolio size of a firm in 2002 is the accumulation of all alliances started in the years 1998 – 2002.
Once all variables are coded the next step in the process is to calculate the Blau Index of Variability (Blau, 1977) scores for the domains of APD. The Blau index measures the degree of heterogeneity of the categorical variables as following:
17
Table 1 Variable operationalization and data sources
Variable (variable name)
Hypothesis/
control Conceptual definition
Operational measure Data source Relevant references Firm performance
H1 – H3 Firm performance Net profit divided by revenue (averaged for three years) Orbis & CompuStat Jiang et al. 2010; Faems et al. 2010 Alliance portfolio size H1 Number of alliances in a portfolio given in a particular year Number of alliances over a five year period
SDC Thomson Baum et al. 2000 Functional diversity
H2a Degree to which alliances in a portfolio differ in functional purpose IQV score of functional diversity SDC Thomson Jiang et al. 2010; Industry diversity H2b Degree to which alliances in a portfolio vary in industrial background IQV score of industry diversity SDC Thomson Jiang et al. 2010 Alliance experience
H3a/b Past experience with alliances (in)experienced based on number of prior alliances SDC Thomson Duysters & Heimeriks 2012; Sivakumar et al. 2011 Firm size Control Number of employees
of a firm in a particular year Logarithmic function of number of employees Orbis & CompuStat Wright, Kroll & Elenkov 2002 Moderating variable
18 sample was split accordingly and tests were run on both sub-samples to examine whether a significant difference could be found between the two.
Control variable
Data on firm size, measured by the number of employees, was obtained from the Orbis and CompuStat databases. Firm size is included as prior studies have pointed towards a correlation between firm size and its propensity to form alliances (Harrigan, 1988). The logarithmic function of the total number of employees of a firm in a given year was used as determinant of firm size.
Statistical method
This study adopted a longitudinal approach to collect data from which a panel-dataset was constructed. Panel data can be analysed by different statistical methods such as pooled OLS, Fixed effects, and Random effects. To determine which method was applicable for the dataset of this study the Hausman-test was conducted. When the data passes the Hausman-test, GLS Random-effects in considered to be most efficient and appropriate. In the other case, when the data does not pass the tests, Fixed-effects should be used. The results of the Hausman-tests show (Model 6 X2(6) = 8.61, p = 0.2818) that the dataset of this research passes the test.
Therefore, in my case the most efficient and appropriate statistical procedure method is GLS Random-effects.
EMPIRICAL RESULTS
19
Table 2 Descriptive statistics and correlations
Variables Mean S.D. 1 2 3 4
1 Net profit margin -0.03989 0.3853
2 Firm size 3,7386 0.8887 0.5187***
3 Portfolio size 14.252 14.3147 0.2115*** 0.6448***
4 Industry diversity 0.6621 0.2341 -0.0003 0.3837*** 0.4174***
5 Functional diversity 0.6926 0.2267 0.2974*** 0.4728*** 0.3266*** 0.2999***
N=250, ***Indicates significance at 1%, **at 5%, *at 10% level, two-sided.
Since the most of the variables within the model show (high) correlation I tested for multicollinearity problems. Multicollinearity occurs when two or more predicting variables in a multi-regression model are correlated with each other. The Variance Inflation Factor (VIF) was calculated to determine to what extend the independent and control variables give valid results. The outcome of the VIF test (mean VIF = 1,58 and condition number = 16,0292) shows a VIF score <5 which implies that no collinearity problems are present in the dataset.
Table 3 reports the results obtained from the GLS Random-effects regression analysis. In models 1 to 5 the control variables and independent variables are tested separately. In model 2 and 4 squared terms are added in order to control for curvilinear relationships. Model 6 jointly tests the influence of the two diversity dimensions and portfolio size on firm performance. In models 7 and 8 the sample is split based on the level of alliance experience a firm possess. All models pass the F-test as they have Prob > Chi-squared values <0.05. Furthermore, the models all show very acceptable R-squared and Rho values, indicating that the models have strong explanatory power.
20
Table 3 Random-effects GLS regression results
***Indicates significance at 1%, **at 5%, *at 10% level, two-sided. Standard errors in parentheses
Independent variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 High Experience
Model 8 Low Experience Firm size 0.265*** 0.314*** 0.261*** 0.257*** 0.271*** 0.328*** 0.3408*** 0.3437***
(0.0499) (0.043) (0.044) (0.046) (0.050) (0.0561) (0.0746)
Alliance portfolio size (APS) -0.003 -0.016*** -0.012** -0.0227*** 0.0097
(0.005) (0.005) (0.0047) (0.0193)
APS squared 0.0001*** 0.0001** 0.0003*** -0.0003
(0.000) (0.000) (0.0001) (0.0008)
Industrial diversity -0.241*** -0.405** -0.348* -1.374 -0.2708
(0.066) (0.184) (0.186) (1.4639) (0.2303)
Industrial diversity squared 0.203 0.208 1.0257 0.0522
21 The test provides a value of 1.89 that is significant at the 5% level (p = 0.0312), indicating that there is a monotone or U-shape in place. In summary, it can be concluded that there is no empirical evidence to support hypothesis 1.
Hypotheses 2a and 2b are concerned with (a) the functional and (b) the industry dimensions of diversity. Hypothesis 2a predicts a positive linear association between functional diversity and firm performance and is tested in models 5 and 6. Model 5 tests the relationship in isolation of the other independent variables and indicates that functional diversity might be negatively associated with firm performance. Model 6 however, does not provide significant evidence for this assumed relationship. This indicates that there is no empirical evidence to support hypothesis 2a and only limited evidence to argue for a negative effect of functional diversity on firm performance is found. Hypothesis 2b states that the influence of industry diversity on firm performance is first positive and later negative, forming an inverted U-shape. This is tested in models 3, 4, and 6. Although at varying significance levels all models illustrate similar results regarding the influence of industry diversity. Model 6 illustrates that industry diversity is negatively associated (p = -0.384) at the 10% significance level. The squared term of industry diversity is not significantly correlated at any significance level in all models. This provides strong empirical evidence to reject hypotheses 2b. It further indicates that a linear negative relationship is present between industry diversity and firm performance.
Experience was predicted to moderate the diversity-performance relationship in such a manner that when a firm has high levels of experience the positive effect of portfolio diversity on firm performance will be stronger. To test for a moderating effect the sample was split between firms with high experience and firms with low experience, resulting in models 7 and 8. The outcomes concerning industry diversity are not significant, indicating that firms do not learn to manage industrial diversity by experience. The results of functional diversity on the other hand do show a significant difference between the high experience and low experience firms at the 5% level (p = 0.2693 vs. p = -0.1765). Higher levels of functional diversity lead to higher firm performance for firms with high levels of experience. This implies that firms learn to manage functional diversity by experience but do not learn to manage industry diversity by experience. Therefore, it can be concluded that there is empirical evidence to support hypothesis 3a and not support hypothesis 3b.
22 reliability of the findings a consistency check was performed using the logarithmic function of turnover as dependent variable. The results of the consistency check showed no major differences in comparison to the model with three year average net profit margin as dependent variable. The only noticeable difference that could be observed was that functional diversity is not significantly correlated with firm performance in any of the models of the consistency check. In the following section the results will be discussed in more depth.
DISCUSSION & CONCLUSION
Strategic alliances have become an essential part of the strategy of firms in the biotech and pharmaceutical industry. The sample of this study confirms this claim and shows that on average firms are increasingly engaged in more and more alliances. This increase in portfolio size will likely lead to higher levels of portfolio diversity and requires firms to learn to manage this increased diversity. Prior studies on the diversity-performance relationship have shown mixed outcomes (Wassmer, 2010) and were often single-domain studies. Studies that combine multiple domains of diversity and other configuration factors are lacking. It has further not often been examined how and to what extent experience influences the diversity-performance relationship. I have used a sample consisting of almost 1500 alliances divided over 50 large biotech firms to examine the relationships between portfolio size, diversity and performance in a multi-domain study. Moreover, it was examined whether firms learn to manage diversity by experience. The multi-domain approach of this study has yielded a number of interesting new insights.
23 that stresses the negative effect of (too) large alliance portfolios. It confirms that large portfolios result in lower firm performance that can largely be explained by higher levels of bureaucracy and transaction costs (Rothaermel, 2001), attention allocation problems (Laursen & Salter, 2006), and the fact that most potent alliance possibilities are pursued first (Rothaermel et al., 2006) leaving less potent possibilities later on. What cannot be explained is that after a certain point adding more alliances does not influence firm performance negatively anymore but the effect becomes nearly neutral. Possibly scale-effects rule out the negative relationship but further investigation to explain this counter intuitive relationship is desired in future research.
Second, many researchers found direct positive relationships between functional diversity and different measures of firm performance. Jiang et al. (2010) found that as a firm increases the functional diversity of its portfolio, it builds a more balanced portfolio that incorporates core and noncore activities, gains access to supplementary and complementary assets, and expands its knowledge base and market reach. Although this balanced approach is believed to lead to enhanced overall firm performance, this study suggests a different relationship. The empirical evidence of this study finds limited evidence to claim that higher levels of functional diversity negatively influence firm performance. These results are in line with the studies of Oxley & Sampson (2004) and Faems et al. (2010) and can be explained by the arguments that increased functional diversity hampers inter-firm communication and leads to managerial difficulties and higher transaction costs, lowering firm performance. Future research could more deeply examine how the configuration of different functional purposes (R&D, marketing, manufacturing, distribution & other) of alliances in a portfolio influences firm performance. This potentially allows firms to make a better combination of alliances with specific functions in their portfolios to maximize firm performance.
24 evidence for the latter arguments. Increased industry diversity is found to negatively impact firm performance. This implies that firms in the biotech and pharmaceutical industry do not benefit from increasing industry diversity of their portfolios. Future research could enhance the understanding of industry specific relationships by comparing empirical data from different industries.
Finally, the moderating effect of past alliance experience on the diversity-performance relationship has so far only been examined by Duysters & Heimeriks (2012). They conceptualize APD as a combination of multiple domains and only test for the effect of experience on this combined measure of APD and portfolio performance. They find evidence for the existence of a positive moderating effect. This study has elaborated on their study and tested the effect of past experience on the separate domains that constitute APD. I found empirical evidence indicating that it is possible for firms to learn to manage functional diversity by experience. In contrast, the industry diversity-performance relationship is not moderated by past experience. It is uncertain why this difference exists, a plausible explanation is that functional diversity is in fact less diversified than industry diversity. There are less different functional purposes an alliance can serve than there are different industries from which a partner can be selected. This makes it harder to generalize lessons learned from an alliance with a partner from the one industry to the alliance with a partner originating from another industry. In order develop a better understanding of how and to what extend past alliance experience influences the different domains of APD, future research should set out to develop a study that include all possible domains of APD.
The empirical findings of this study contribute to existing literature in a number of ways. Baum et al. (2000) argued that portfolio size alone is not a sufficient predictor of performance. Instead, a more comprehensive approach is needed that includes other configuration factors. In addition, recent studies have started acknowledge that APD is a multi-domain construct that encompasses more than one determinant of diversity (Jiang et al., 2010). This study has tried to fill this gap and adopted a collective approach combining both research streams on alliance portfolio size and APD and contributes to the understanding of the delicate inter-play of these configuration factors. It acknowledges that size and the different domains of diversity play a significant and specific role in explaining firm performance.
25 relationship. This study contributes to the literature by providing a deeper insight into this relationship. I conclude that experience has dissimilar effects on different domains of diversity. Firms can learn to manage functional diversity by experience, but this does not apply to industry diversity. Therefore, in future research this diversification between diversity domains should be taken into account. Where Duysters & Heimeriks constructed one combined measure of diversity, I argue that this gives an unclear and incomplete presentation of the influence of experience and I propose that the underlying domains of diversity should be tested separately. This allows scholars to find the precise effect of experience on the diversity-performance relationship for all the different domains that can constitute APD.
The implications for managers are twofold. Firstly, managers should understand that more is not always better. Especially in the case of portfolio size and diversity it is tempting to only think of the benefits more alliances could bring to a firm’s portfolio. This study, however, clearly underlines that the costs of large and diversified portfolios exceed the benefits. Therefore, managers should seek to balance diversity of their portfolio and specialise their alliance activities in order to lower costs and increase benefits. Balancing diversity and specializing can help to overcome complex coordination issues and difficulties aligning strategic goals and is therefore most likely to increase the performance of the alliances.
Secondly, experience alone does not allow a firm to learn to manage all forms of portfolio diversity. For some domains of diversity experience allows firms to better manage the diversity of their portfolio, but for other domains different mechanisms should be in place. Managers should be aware of the different domains of APD and recognize that each should be treated differently. They have to consider utilizing more deliberate learning mechanisms in order to maximize learning benefits when experience does not allow for learning. Building alliance capability or establishing dedicated alliance functions are examples that can foster the learning process (Dyer et al., 2002) and help firms to better manage portfolio diversity.
26 industry. Third, this study has conceptualized APD along two dimensions of diversity (functional & industry). A more comprehensive approach including all possible dimensions of diversity would have given a much more detailed and complete overview of the diversity-performance relationship. However, due to constraints in time and resources other dimensions have not been considered in this study. Future research can build on the multi-dimensional construct used in this research and elaborate on it. Fourth, the alliance experience variable was a measured over only one timeframe (1998 – 2002) and the evolution of this variable over the remaining timeframe (2002 – 2006) was not taken into account. A more comprehensive understanding of the effect of alliance experience could be obtained by including the evolution of alliance experience over time. Therefore, future research could elaborate on this study and develop a more sophisticated measure of alliance experience. Finally, because this is an empirical study it is focussed on the ‘how’ question and does not extensively address the ‘why’ question. Hence, future research should focus on qualitative data in order to develop a deeper understanding of why certain relationships exist.
In conclusion, where most prior literature focussed on one of the three topics of this study, I combined portfolio size, APD, and alliance experience in one comprehensive study. The call for research on portfolio size in addition to multi-domain diversity construct research was addressed. I extended this research by incorporating an emerging stream of literature on alliance experience. The results of this empirical study show that increase in portfolio size and diversity have a negative impact on financial performance. Furthermore, diversity should not be seen as a combination of the domains that constitute the diversity but all domains of diversity require a different approach. Functional diversity can be learned to manage by experience, but for industry diversity other mechanisms need to be utilized. Altogether, these findings shed new light on the interplay of portfolio size and diversity and the moderating role of experience. This study can help firms to optimize the learning and resource benefits while minimizing complex coordination costs that encompass managing a diverse alliance portfolio. Lastly, it uncovers many new interesting opportunities to further investigate in future research.
REFERENCES
27 Amaldoss, W., & Staelin, R. (2010). Cross-Function and Same-Function Alliances: How Does Alliance Structure Affect the Behavior of Partnering Firms?. Management Science, 56(2), 302-317.
Anand, B. N., & Khanna, T. (2000). DO FIRMS LEARN TO CREATE VALUE: THE CASE OF ALLIANCES. Strategic Management Journal, 21(3), 295.
Argote, L., Beckman, S. L., & Epple, D. (1990). THE PERSISTENCE AND TRANSFER OF LEARNING IN INDUSTRIAL SETTINGS. Management Science, 36(2), 140-154.
Barney, J. (1991). Firm Resources and Sustained Competitive Advantage. Journal Of Management, 17(1), 99.
Baum JAC, Ingram P. (1998). Survival-enhancing learning in the Manhattan hotel industry, 1898–1980. Management Science, 44, 996–1016.
Baum, J. C., Calabrese, T., & Silverman, B. S. (2000). DON'T GO IT ALONE: ALLIANCE NETWORK COMPOSITION AND STARTUPS' PERFORMANCE IN CANADIAN BIOTECHNOLOGY. Strategic Management Journal, 21(3), 267.
Bettis, R. A., Bradley, S. P., & Hamel, G. (1992). Outsourcing and industrial decline. Executive (19389779), 6(1), 7-22.
Biemann, T., & Kearney, E. (2010). Size does matter: How varying group sizes in a sample affect the most common measures of group diversity. Organizational research methods, 13(3), 582-599.
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: a new perspective on learning and innovation. Administrative science quarterly, 35(1).
Cui, A. S., & O'Connor, G. (2012). Alliance portfolio resource diversity and firm innovation. Journal of Marketing, 76(4), 24-43.
Darr, E. D., & Argote, L. (1995). The Acquisition, Transfer, and Depreciation of Knowledge in Service Organizations: Productivity in Franchises. Management Science, 41(11), 1750-1763.
Das, T. K., & Bing-Sheng, T. (2000). A Resource-Based Theory of Strategic Alliances. Journal Of Management, 26(1), 31-62.
Deeds, D. L., & Hill, C. L. (1996). STRATEGIC ALLIANCES AND THE RATE OF NEW PRODUCT DEVELOPMENT: AN EMPIRICAL STUDY OF ENTREPRENEURIAL BIOTECHNOLOGY FIRMS. Journal Of Business Venturing, 11(1), 41.
28 Duysters, G., & De Man, A. (1999). A Network Approach to Alliance Management.
European Management Journal, 17(2), 182.
Duysters, G., & Lokshin, B. (2011). Determinants of Alliance Portfolio Complexity and Its Effect on Innovative Performance of Companies. Journal Of Product Innovation
Management, 28(4), 570-585.
Duysters, G., Heimeriks, K. (2007). Alliance Capability as a Mediator Between Experience and Alliance Performance: An Empirical Investigation into the Alliance Capability
Development Process. Journal of management studies, 44(1), 25-49.
Duysters, G., Heimeriks, K. H., Lokshin, B., Meijer, E., & Sabidussi, A. (2012). Do Firms Learn to Manage Alliance Portfolio Diversity? The Diversity-Performance Relationship and the Moderating Effects of Experience and Capability. European Management Review, 9(3), 139-152.
Dyer, J. H., Kale, P., & Singh, H. (2001). How to make strategic alliances work. Sloan Management Review, 42(4), 37–43.
Eisenhardt, K. M., & Schoonhoven, C. (1996). Resource-based View of Strategic Alliance Formation: Strategic and Social Effects in Entrepreneurial Firms. Organization Science, 7(2), 136-150.
Ernst, D., Halevy, T., Monier, J.-H. J., Sarrazin, . 2001. A future for e-alliances. McKinsey Quarterly, 2, 92–102.
Faems, D., de Visser, M., Andries, P., & van Looy, B. (2010). Technology Alliance Portfolios and Financial Performance: Value-Enhancing and Cost-Increasing Effects of Open
Innovation. Journal Of Product Innovation Management, 27(6), 785-796.
Faems, D., Van Looy, B., & Debackere, K. (2005). Interorganizational Collaboration and Innovation: Toward a Portfolio Approach. Journal Of Product Innovation Management, 22(3), 238-250.
Gilley, K., & Rasheed, A. (2000). Making More by Doing Less: An Analysis of Outsourcing and its Effects on Firm Performance. Journal Of Management, 26(4), 763-790.
Granovetter M. (1985). Economic action and social structure: the problem of embeddedness. American Journal of Sociology, 91(3), 481–510.
Grant, R. M. (1996). TOWARD A KNOWLEDGE-BASED THEORY OF THE FIRM. Strategic Management Journal, 17, 109-122.
29 Hagedoorn, J. (1993). UNDERSTANDING THE RATIONALE OF STRATEGIC
TECHNOLOGY PARTNERING: INTERORGANIZATIONAL MODES OF
COOPERATION AND SECTORAL DIFFERENCES. Strategic Management Journal, 14(5), 371-385.
Hagedoorn, J. (2002). Inter-firm R&D partnerships: an overview of major trends and patterns since 1960. Research policy, 31(4), 477-492.
Hoang, H., Rothaermel, F. T. (2005). The effect of general and partner-specific alliance experience on joint R&D project performance. Academy Of Management Journal, 48(2), 332-345.
Hoffmann, W. H. (2007). Strategies for managing a portfolio of alliances. Strategic Management Journal, 28, 827-856.
Jiang, R., Tao, Q., & Santoro, M. D. (2010). Alliance portfolio diversity and firm performance. Strategic Management Journal, 31(10), 1136-1144.
Jones, G. R., & Hill, C. W. (1988). Transaction cost analysis of strategy‐structure choice. Strategic management journal, 9(2), 159-172.
Kogut, B. (1988). Joint ventures: Theoretical and empirical perspectives. Strategic management journal, 9(4), 319-332.
Koza, M. P., & Lewin, A. Y. (1998). The co-evolution of strategic alliances. Organization science, 9(3), 255-264.
Laursen, K., & Salter, A. (2006). Open for innovation: the role of openness in explaining innovation performance among U.K. manufacturing firms. Strategic Management Journal, 27(2), 131-150.
Lavie, D. (2007). Alliance portfolios and firm performance: A study of value creation and appropriation in the U.S. software industry. Strategic Management Journal, 28, 1187-1212. Lind JT, Mehlum H. 2010. With or Without U? The Appropriate Test for U-Shaped
Relationship. Oxford Bulletin of Economics and Statistics 72(1), 109-118.
Lunnan, R., & Haugland, S. (2008). Predicting and measuring alliance performance: A multidimensional analysis. Strategic Management Journal, 29(5), 545–556.
Luo, X., & Deng, L. (2009). Do Birds of a Feather Flock Higher? The Effects of Partner Similarity on Innovation in Strategic Alliances in Knowledge‐Intensive Industries. Journal of Management Studies, 46(6), 1005-1030.
30 Nooteboom, B., Van Haverbeke, W., Duysters, G., Gilsing, V., & Van den Oord, A. (2007). Optimal cognitive distance and absorptive capacity. Research Policy, 36(7), 1016-1034. OCASIO, W. (1997). TOWARDS AN ATTENTION-BASED VIEW OF THE FIRM. Strategic Management Journal, 18, 187-206.
Oxley, J. E., & Sampson, R. C. (2004). The Scope and Governance of International R&D Alliances. Strategic Management Journal , 25(8/9), 723-49.
Ozcan, P., & Eisenhardt, K. M. (2009). ORIGIN OF ALLIANCE PORTFOLIOS:
ENTREPRENEURS, NETWORK STRATEGIES, AND FIRM PERFORMANCE. Academy Of Management Journal, 52(2), 246-279.
Park, G., & Kang, J. (2013). Alliance Addiction: Do Alliances Create Real Benefits?. Creativity & Innovation Management, 22(1), 53-66.
Park, S. H., & Zhou, D. (2005). FIRM HETEROGENEITY AND COMPETITIVE
DYNAMICS IN ALLIANCE FORMATION. Academy Of Management Review, 30(3), 531-554.
Prahalad, C. K., & Hamel, G. (1993). The core competence of the corporation. Organization of Transnational Corporations, 11, 359.
Reuer, J. J. 1999. Collaborative strategy: The logic of alliances. Mastering Strategy, October 4, 12–13.
Reuer, J. J., Park, K. M., & Zollo, M. (2002). Experiential learning in international joint ventures: the roles of experience heterogeneity and venture novelty. Cooperative strategies and alliances, 321, 344.
Rothaermel, F. T. (2001). INCUMBENT'S ADVANTAGE THROUGH EXPLOITING COMPLEMENTARY ASSETS VIA INTERFIRM COOPERATION. Strategic Management Journal, 22(6/7), 687.
Rothaermel, F. T., & Deeds, D. L. (2004). Exploration and exploitation alliances in
biotechnology: a system of new product development. Strategic management journal, 25(3), 201-221.
Rothaermel, F. T., Hitt, M. A., & Jobe, L. A. (2006). Balancing vertical integration and strategic outsourcing: effects on product portfolio, product success, and firm performance. Strategic Management Journal, 27(11), 1033-1056.
31 Sampson, R. C. (2007). R&D alliances and firm performance: the impact of technological diversity and alliance organization on innovation. Academy of Management Journal, 50(2), 364-386.
Shan, W., Walker, G., & Kogut, B. (1994). Interfirm cooperation and startup innovation in the biotechnology industry. Strategic management journal, 15(5), 387-394.
Sivakumar, K. K., Roy, S., Zhu, J., & Hanvanich, S. (2011). Global innovation generation and financial performance in business-to-business relationships: the case of cross-border alliances in the pharmaceutical industry. Journal Of The Academy Of Marketing Science, 39(5), 757-776.
Stigler, G. J. (1958). Economies of Scale, The. JL & Econ., 1, 54.
Tushman, M. L., & O'Reilly III, C. A. (1996). Ambidextrous organizations: managing evolutionary and revolutionary change. California management review, 38(4).
Vasudeva, G., & Anand, J. (2011). Unpacking absorptive capacity: A study of knowledge utilization from alliance portfolios. Academy of Management Journal, 54(3), 611-623. Wassmer, U. (2010). Alliance Portfolios: A Review and Research Agenda. Journal Of Management, 36(1), 141-171.
Wright, P., Kroll, M., & Elenkov, D. (2002). Acquisition returns, increase in firm size, and chief executive officer compensation: The moderating role of monitoring. Academy of Management Journal, 45(3), 599-608.
Books:
Blau, P. M. (1977). Inequality and heterogeneity: A primitive theory of social structure (Vol. 7). New York: Free Press.
Doz, Y. L., & Hamel, G. (1998). Alliance advantage: The art of creating value through partnering. Harvard Business Press
