The play with exploration and exploitation on firm performance : how alliance portfolios could assist multinational enterprises from emerging markets in their efforts for internationalisation

(1)

The play with Exploration and

Exploitation on

Firm Performance:

How alliance portfolios could assist Multinational Enterprises from Emerging Markets in their efforts for internationalisation. Faculty of Economic and Business MSc Business Administration Track: International Management Alexandra M. Heckman Student number 10474552 Supervisor Carsten Gelhard Date 23/07/17 Word count 13,785

(2)

pg. 2

Statement of originality

This document is written by Student Alexandra Heckman who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

pg. 3

Abstract

A heightened firm performance calls for firms to engage in ambidextrous activities – the exploitation of existing competencies, and the exploration of new competencies. We posit that ambidextrous firms have less difficulty internationalising, as they can leverage their resources to enter both emerging- and mature markets. Nevertheless, there are many assumptions on the differences in internationalisation paths of firms originating from emerging and developed markets, as there remains a research gap concerning the antecedents of ambidexterity. This research gap is elucidated by examining how the institutional setting of a firm influences the relationship between ambidexterity and firm performance. Our analysis consists of 1511 MNEs originating in emerging and developed markets which formed cross-national alliances in the period 2006-2015. The findings indicate that ambidexterity does indeed lead to a higher firm performance, while the home country of a firm does not influence this relation. By delving into the debate concerning the differences home markets can have on firm performance, we derive insights into the implications for firms internationalising.

(4)

pg. 4

1. Introduction

It is remarkable that Multinational Enterprises based in Emerging Markets (eMNEs) have been expanding globally at relatively high rates during recent years. This growth is demonstrated both in the increasing Outward Foreign Direct Investment (OFDI) flows and the growth in cross-border alliances. Developed countries continue to be the major OFDI home economies, as indicated by the 60% the OFDI has contributed to the world’s total flows, though, the increasingly growing impact of emerging markets is noticeable as in 2014 they contributed 39% to the total OFDI flows (Rudello, Savarese, & Belussi, 2016). Athreye and Kapur (2009) indicated that the increasing outflows of FDI originating from emerging markets can be linked to the remarkable shift in the successful internationalisation eMNEs are experiencing in recent years. This is also highlighted by the increasing number of eMNEs presented in the 2016 Forbes ‘Global 2000’ list (Forbes, 2016).

The internationalisation of Multinational Enterprises (MNEs) is a heavily researched topic in the international business literature, particularly due to the radical changes portrayed in the global investment trends. Traditional theories related to internationalisation, such as the eclectic theory of Dunning (1988) sketch out certain paths dMNEs should follow to achieve success. The eclectic model, also known as the OLI model, indicates three main drivers as being influential to the firm performance which are O-advantages (ownership), L-advantages (location), and I-advantages (internalization). Hereby, it is stated that MNEs enjoy superior technological and managerial resources, ownership advantages, which are leveraged to enter foreign markets by for example making use of (O)FDIs. In comparison to multinational enterprises originating from developed economies (dMNEs), eMNEs are often construed as being ‘backward’ – often mentioned in relation to the lacking of proprietary technology. MNEs coming from emerging markets do not obey this established view, as their O-advantages are often not developed enough to be leveraged. Their rate of oFDIs has, however, increased

(7)

pg. 7 in the recent years indicating that the conventional view regarding internationalisation of firms does not explain the internationalisation success of eMNEs.

Moreover, Lavie (2006) indicated that the assumption that MNEs, in general, should own or wholly control (valuable) resources leading to a competitive advantage is false, as access to resources whereby the firm can employ the resources and secure the benefits, is enough for MNEs. It is therefore questionable whether traditional models discussing internationalisation can explain the successful internationalisation of eMNEs (Peng, 2012). It is arguable that eMNEs represent a novel group of multinationals which have adapted their pattern of international expansion. For this reason, many scholars argue that a new theory or an expansive theory is necessary to explain this renewed form of internationalisation (Guillen & Garcia-Canal, 2009; Luo & Tung, 2007).

Whilst there are many theories on the cause of this accelerated growth in the internationalisation of the eMNEs, compared to the growth experienced by the of dMNEs, a noticeable research gap exists concerning the interfirm resources MNEs have access to through their alliances and how this may influence their performance. The level of alliance formation has experienced growth as well, complementary to the increase in eMNEs internationalizing. This can be related to the competitive environment, as alliances allow for flexibility by entering new markets, extending competitive advantages, limiting the amount of market uncertainty, and sharing the costs and risks between alliance partners. The benefits related to alliances have motivated firms to engage in them, and have led to alliance networks being a focal point in the development of the business.

When eMNEs internationalise they face the tension of (i) implementing their home-grown competitive advantages in similar (emerging) economies, as well as (ii) further developing their competencies to enter more advanced countries. The continuous exploration (building new competencies), and exploitation (developing current competencies) indicates the

(8)

pg. 8 necessity of eMNEs having to be ambidextrous to succeed within the home country and abroad (He & Wong, 2004). An analysis of a firms’ alliance portfolio, defined as being the firms’ capabilities of accessing and deploying various resources in inter-firm relations (ibid.), will, therefore, shed more light on the influence of differencing alliance portfolios on the internationalisation process of eMNEs and dMNEs.

Various studies have analysed the effect of alliance portfolio on firm performance. Scholars have focused on the influence of structural characteristics, such as portfolio size, partner diversity, alliance intensity, and geographic aspects of firm performance (Lavie & Miller, 2008; Jiang, Tom Tao & Santoro, 2010; Saebi, 2011). Another stream has implemented a firm-level perspective on their research by analysing internal aspects (e.g. management) which could influence firm performance. Though, one gap in this research concerning alliance portfolio’s is the influence of ambidexterity on the internationalisation of firms. Whilst Yamakawa, Yang and Lin (2011) examine the influence of the alliance portfolio on the success rate of firms, they focus on dMNEs who largely obey by the traditional theories, instead of focusing on a new generation of multinational enterprises. Furthermore, the tensions firms experience in finding the balance between exploration and exploitation is of additional interest to explaining the success of eMNEs, and contrasting this with the way dMNEs internationalise. This paper advances the current alliance portfolio research by analysing the striking differences in internationalisation processes of established MNEs from developed markets and MNEs coming from emerging markets. Therefore, the research question is:

To what extent does the alliance portfolio differ between Multinational Enterprises from emerging (eMNEs) and developed markets (dMNEs) and

(9)

pg. 9 The findings address the following key aspects in the current literature.

A. An in-depth analysis regarding the influence of alliance portfolios on emerging- and developed-market multinationals will be conducted, which will allow for greater insight into the accelerated growth trends of eMNEs in relation to the surge in OFDIs. Within this thesis, the alliance portfolios of eMNEs is contrasted against those of dMNEs. Whilst there are multiple studies that analysed either the (i) influence of alliance portfolios on MNEs (Luo & Tung, 2007; Guillén & Garcia-Canal, 2009; Duysters & Lokshin, 2011), or (ii) the relationship between firm characteristics, such as the number of alliances, the type of network or the geographic scope on the level of innovation, value creation and profitability of a firm (Lavie & Miller, 2008; Zaheer & Hernandez, 2011), there is to the author’s knowledge no research that looks specifically into the influence that an alliance portfolio composition (emphasis on exploration or exploitation) in a specific institutional context can have on the level of internationalisation of firms. We wish to investigate whether eMNEs are putting more emphasis on developing ambidextrous capabilities in comparison to dMNEs. This could explain why eMNEs are simultaneously entering developed and developing markets, whilst most dMNEs are following the more traditional path and slowly entering less to more distant countries.

B. Secondly, related to the first aspect, this research could contribute to enhancing or expanding current traditional internationalisation theories such as the eclectic or Uppsala model. There are multiple views on the relatability of traditional models on this new group of multinationals. Some scholars indicate that the current theories can still serve as a tool in analysing the internationalisation patterns (Verbeke & Kano, 2012), as there are more similarities (e.g. motivations) than dissimilarities. Nonetheless, other scholars indicate that current eMNEs represent a completely new generation of

(10)

pg. 10 multinationals and therefore call for new theoretical models. In addition, this research might also provide implications for dMNEs and their (slower) rate of internationalisation. It will give insight into the type of alliance portfolio necessary to enhance their global growth and competitive advantage.

C. Furthermore, to the author’s knowledge, there is no study which uses ‘country of origin’ of an MNE as a moderator in analysing the firm performance of multinationals. This is striking as various scholars have highlighted the added advantages a home country can give firms in their global strategy (Porter, 1990; Buckley, Forsans, Munjal, 2012). This paper is structured as follows: In the next section an in-depth literature review will be presented indicating the current level of knowledge related to the general internationalisation of MNEs. The existing theories will provide the foundation for the conceptual model that will be developed to clarify the internationalisation of eMNEs which is discussed in the section II of this paper. The methodology chapter will discuss the data used in this study, whereby a detailed explanation is given of the dataset and the various (control-) variables used. The following chapters will present and discuss the findings which arose out of the analysis, after which the limitations and future suggestions will be brought forward.

(11)

pg. 11

2. Literature Review

This chapter will analyse and evaluate the literature regarding the influence of alliance portfolios on firm performance. This is done by exploring internationalisation on itself, followed by the explanation of alliance portfolios and the influence of its composition (exploration vs. exploitation) on firm performance.

2.1 Internationalisation

Timberlake, Kumar and McLeod (1982) acknowledge that there is a “second-wave” of companies coming from eMNEs characterized by their activity in the global setting – which is most likely a resultant of globalization. This new generation of multinationals is experiencing increased global growth compared to multinationals based in developed countries.

There are various streams in the literature which attempt to explain the paths MNEs take when internationalising. Vernon’s product cycle theory discusses the stages that firms should pass through for successful internationalisation. Although this theory is often linked to high-tech products, it does indicate the general idea that companies go through incremental growth. Key to this hypothesis is the idea that FDI will flow from developed to less-developed countries. This, however, does not reflect the current situation as there has been an increase in Outward Foreign Direct Investment (OFDI) by companies originating from emerging markets, indicating the focus on more outward orientation of emerging markets. Other traditional theories, such as the Uppsala model, indicate a certain path that companies should follow, often starting with inward FDI allowing companies to acquire manufacturing capabilities, followed by export to markets which have not reached a similar level of development (Athreye & Kapur, 2009). This method recognizes the difficulty in acquiring knowledge in foreign markets, and thus maintains a certain set of routines. And finally, the Eclectic Model, also known as the OLI model, is a leading theoretical framework which determines the extent of a company’s

(12)

pg. 12 multinational potential. Hennart (2012) defines it as being a model which “explains the extent and pattern of value added by MNEs outside their national boundaries” (p. 169). It allows for a constructive way of thinking with regards to MNEs and thus is used as a foundational theory in many economic papers (Hennart, 2012; Ramamurti, 2009). The model indicates the choices for companies to make use of FDI and thus exploit their firm-specific-advantages (FSA). The abbreviation OLI stands for Ownership, Location and Internalization; the three foundations of building an internationally active firm, and its influence on the decision-making process related to becoming multinational. Ownership advantages can also be seen as being firm-specific advantages which are both tangible and intangible assets and allow for the implementation of (high-tech) technology. These advantages indicate the reason as to why some firms internationalise, and others are not able to do so, as they allow MNEs to leverage them to enter foreign countries and overcome costs abroad (Dunning, 1988; Rugman & Verbeke, 1990). A firm can acquire the following ownership advantages: (i) asset-based ownership; which suggest that advantages need to be valuable, rare and imitable, (ii) ownership related to the economies of common governance; related to the scale and scope, and (iii) institutional advantages (related to the culture, norms and leadership within MNEs). Location advantages also explained as country-specific advantages, and relate to the country’s endowment regarding natural resources, the cost of labour and the institutional framework. Internalization advantages discuss which firm-specific advantages are used internally, and is thus related to the way a firm chooses to operate in host countries. These advantages often arise from imperfect markets.

2.2 eMNEs vs. dMNEs

Even though some scholars argue that eMNEs and dMNEs do not need differing theories explaining their internationalisation process, there are some distinct differences between eMNEs and dMNEs dealing with (i) firm-specific advantages as well as (ii) motivations.

(13)

pg. 13 2.2.1 Competitive advantage

The creation of competitive advantage for MNEs has been a key topic in the international business literature. Within the literature, we look at both the Firm Specific Advantages (FSAs) and the Country Specific Advantages (CSAs).

2.2.1.1 Firm Specific Advantages

First, eMNEs bring firm-specific advantages (FSAs) along with them that set them apart from dMNEs. These firm-specific advantages, or proprietary assets, are capabilities which can be transferred to any other setting or country, and thus allows the relevant firm to compete with local companies. Rugman and Verbeke (2001) indicate that FSAs can take two main forms. First, it can be a production-related proprietary asset, or secondly an organizational capability increasing the level of efficiency within the MNEs asset base. Verbeke and Kano (2012) identify various FSAs of eMNEs that drive their level of internationalisation:

I. The first key FSA is the fact that MNEs will experience preferential access to home-country location advantages – this includes aspects such as cheap labour and natural resources, likely to be derived from the local network (government connections) and knowledge of the market in the home country. Though, this explanation may not be completely satisfying, for the reason that FSAs related to the location do not form the foundation of longer-term successes of eMNEs.

II. Furthermore, MNEs experience FSAs related to the derived firm capabilities in process innovations allowing firms to produce goods at a lower price without the risk of reducing the product quality. Although, eMNEs enter new markets in order to acquire new FSAs, and not necessarily exploit current FSAs, these newly acquired FSAs will eventually be used in the home country, allowing these firms to enter developing and developed markets simultaneously providing low-cost goods at a high quality (Madhok & Keyhani, 2012; Leo

(14)

pg. 14 & Tung, 2007; Rugman, 2009). Therefore, it can be stated that eMNEs have a completely different starting point when entering the “global” market in comparison to dMNEs.

This view on firm-specific advantages espouses the resource-based view (RBV) of MNEs. The RBV, a strategic management model, indicates how firms compete in term of resources and (internal) capabilities to give rise to sustainable competitive advantage, and consequently an increased firm performance. The framework conceptualizes firms as being heterogeneous entities consisting of idiosyncratic bundles. Heterogeneity denotes that any MNE can compete in the global marketplace and, at least, break even (Peteraf, 1992). The difference is that firms in possession of marginal resources are most likely to just break even, whilst firms with superior resources are more likely to be more profitable. These superior resources are to be found internally and must be ‘imperfectly mobile’ which indicates that these are resources which are more valuable within the MNE that employ them that they would be elsewhere (ibid). The RBV view argues that (internal) resources are key in achieving a competitive advantage, however, as already mentioned eMNEs are (i) either lacking O-advantages and can therefore not exploit nor leverage these to enter new markets or (ii) do possess ownership advantages but these are unlike the ones we are used to seeing in dMNEs (Ramamurti, 2012). Nevertheless, Lavie (2006) indicated that specifically alliances influence the level of competitive advantage of a firm. This is for the reason that ‘strategic’ alliances allow firms with limited resources to create a uniquely heterogeneous resource base, which may, in turn, lead to an increased firm performance (Saebi, 2011). In addition to this, ownership of resources is not necessary anymore to achieve competitive advantage, instead, MNEs just need access or the right to use and employ the resources (Lavie, 2006).

2.2.1.2 Country Specific Advantages

There is a long tradition related to examining the influence nations can have on specific firms. Porter (1990) indicates that countries have increased in importance when facing global

(15)

pg. 15 competition, whereby competitive advantages are created and sustained through a localized process. Hereby it is of importance to grasp that the differences in countries such as culture, values, and institutions, contribute to the competitive success whereby a nation will excel in industries that match the home environment. The differences in national capabilities will, therefore, influence the global strategy of a MNE. Buckley et al. (2012) indicate that the macroeconomic environment can grow into sources of competitive advantage, in the case that firms can internalise these and convert these to mobile assets.

2.2.2 Motivation

EMNEs and dMNEs have differing motivations to enter foreign markets. Dunning (1988) indicated that there are four types of motivations to establish a competitive international position which is the following: (i) Market-seeking; related to increasing the current sales particularly when the home market is saturated or when host-markets are more appealing, (ii) Resource-seeking; related to cheaper natural resources or labour, (iii) Strategic asset-seeking; discusses the growth of the firm due to access to knowledge or technology which allows for the development of current competencies, and lastly (iv) Efficiency-seeking; which highlights the efficiency of current processes related to for example scale of economies, risk reduction or geographic distance.

Table 1 Distinction motivation eMNEs and dMNEs

Dunning’s motivation typology (Dunning, 1988) Explanation Distinction in motivation of eMNE & dMNE

Market-seeking Exploitation of a foreign market which is of appeal to

the MNE Yes

Resource-seeking Acquiring specific resources at a lower price point than

in the home country No Strategic asset-seeking Acquiring (technological) knowledge which attempts to

complement instead of exploiting current competencies. Yes Efficiency-seeking Improving the current level of efficiency in established

(16)

pg. 16 Table 1 indicates that the main differences between motivations are related to market-, strategic-, and efficiency seeking. The difference in market-seeking motivation is related to the reason behind going, whilst dMNEs may go abroad due to the saturated home market, eMNEs will often go abroad due to political instability, increasingly difficult governance conditions, or any other type of regulatory restraint (Guillen & Garcia-Canal, 2009). Strategic asset-seeking was already highlighted when discussing FSAs, as eMNEs mainly strive towards creating new FSAs instead of exploiting the advantages they already have (Guillen & Garcia-Canal, 2009). This strengthens their competitive position in the global market, as well as in their home country due to the springboard perspective (Luo & Tung, 2007). The difference in efficiency seeking strategy is initiated by both eMNEs and dMNEs to overcome export barriers, another factor for eMNEs is that it allows for risk diversification (Gaur & Kumer, 2010; Guillen & Garcia-Canal, 2009). This table indicates that even though both eMNE and dMNE internationalise, the motivation behind internationalizing differs, which is of interest within this study.

2.3 Alliance portfolio’s

Competitive pressures and technological innovations make it is of importance for firms to develop the necessary competencies to achieve a competitive advantage. MNEs can achieve such a competitive advantage by making use of alliances in order to share risks, secure market advantages, limit transaction costs, as well as having access to resources. Alliances are defined as being a voluntary arrangement among independent firms to exchange or share resources and engage in the co-development or provision of products, services, or technologies (Gulati, 1998). The alliance portfolio itself, also known as an egocentric alliance network, is the collection of simultaneously operating direct alliance relationships of the focal firm (Castro, Casanueva, Galan, 2014). As Castro et al. indicate an alliance portfolio may constitute of new alliances which are being formed, as well as terminated strategic alliances. The portfolio itself

(17)

pg. 17 is affected by exogenous factors such as the industry or intensity of competition in which it is active and endogenous factors such as the relational network. However, key is that there is a simultaneous development between the firm(strategy) and the alliance portfolio.

Heimeriks, Klijn and Reuer (2009) indicate that there has been a significant growth in alliance activities in the past two decades. This can be explained by the fact that it allows MNEs to “overcome resource limitations, crack new markets and hedge against environmental uncertainties by creating options to expand” (ibid.). Firms can no longer focus on solely one alliance profile due to the globalization of markets, and the increasingly competitive market. They are facing (market) pressures to form alliance portfolios for the three following reasons dealing with (i) resource endowment, (ii) competitive positioning, and (iii) increased flexibility. An alliance portfolio allows MNEs to enhance their resource endowment, as the network of alliances will give firms the strategic advantage to access resources and knowledge from other markets. Therefore, the alliance portfolio will allow for the quality of information, as well as the quantity and variety of this. The alliance portfolio will also give firms a boost to improve their current competitive positioning with regards to their competitors. Saebi (2011) indicates that the positioning of a firm in a network of alliances substantially influences its competitiveness, as well as its stability within the inter-organizational field. And finally, it allows the firms to access foreign markets due to increased flexibility and adaptability to the differences in the global market (Lavie & Miller, 2008).

Alliances can take on different forms and are therefore seen as flexible arrangements between companies, ranging from the creation of completely new entities, such as joint ventures, to acquiring technological capabilities in the form of R&D alliances. Goeltz (2014) indicates that the governance structure of alliances can be classified as equity-based joint ventures, a minority equity arrangement, or a non-equity agreement. The difference between these three forms is that joint ventures are, as stated before, new entities, whilst minority equity

(18)

pg. 18 arrangements can vary in governance structure such as the before mentioned R&D alliance, and non-equity agreements can be categorised as downstream agreements related to for instance marketing alliances (ibid.).

2.3.1 Alliance strategy

We note that firms can implement multiple strategies that will lead to a more superior performance when analysing competitive strategies that MNEs are likely to apply. By making use of exploration and exploitation, MNEs are able to secure their competitive advantage through developing their alliance portfolio. Whether a firm will enter into an alliance depends to a large extent whether a firm seeks to either (i) exploit or (ii)explore, which is equally as relevant when discussing the strategies MNEs are undertaking. There are various motivations for firms to enter alliances, though the most common reason is to pool complementary resources, secure necessary resources, as well as spread the risks and costs of projects (Koza & Lewin, 1998).

Exploration focuses on prospecting possible future landscapes, and thus delves into opportunities to increase the level of returns. This requires MNEs to learn from their partner alliances or to acquire knowledge to create new competencies. Therefore, an absorptive capacity is necessary to identity possible opportunities. In this way firms’ awareness in relation to learning opportunities are raised, also allowing for a more proactive attitude towards exploration. In contrast to this, exploitation inquires the further development of current competencies by looking at ways to increase efficiency and reducing costs (Koza & Lewin, 1998; Williamson, Ramamurti, Fleury, & Fleury, 2013). Therefore, exploitation alliances delve into leveraging a firms’ existing resources for immediate gains (Krammer, 2016). A firm often puts more pressure on exploitation due to organizational inertia as stated by Lavie and Rosenkopf (2006). This occurs in the case that the speed of reorganization is lower than the rate of changes in environmental conditions (ibid.). Exploitation is shaped by the routinized

(19)

pg. 19 behaviour of firms, as in the case of an issue a localised, internal response is sought after. In this case, uncertainty is reduced due to consistent actions which consequently lead to reproducible outcomes.

The exploration – exploitation view is also supported by Porter and his competitive strategy as indicated by Goeltz (2014). Porter theorised about generic strategies – cost leadership, differentiation and focus – which should allow for a competitive advantage. He explored the difficulties in combining both strategies such as (i) cost leadership, related to exploitation, which focuses on learning to continue lowering costs, standardization and efficiency, whilst (ii) differentiation, related to exploration, emphasises (product) innovation and a premium image. However, in today’s dynamic and turbulent environment, firms need to be able to make use of both exploitation and exploitation in order to stay competitive. In the case that firms overemphasize exploration, there is a risk of overspending already scarce resources with a relatively low payback, whilst when firms solely focus on exploitation, it may impede long-term performance due to the reduction of learning new competencies (Uotila, Maula, Keil & Zahra, 2008). Therefore, ambidexterity, being able to find a balance between exploration and exploitation, is of benefit for MNEs.

2.3.2 Ambidexterity

Organizational ambidexterity is related to the manner in which eMNEs are fulfilling their objectives associated with both exploitation and exploration, and consequently also the firm performance. Ambidexterity is an unusual (competitive) advantage as it allows organizations to develop products for emerging markets simultaneously as well as compete in mature markets (Preda, 2014). The tension between both aspects highlight the inherent trade-offs which occur within firms.

Yamakawa, Yang and Lin (2011) differentiate between exploratory and exploitative activities by stating that R&D alliances and technical alliances are used for exploratory

(20)

pg. 20 purposes, whilst firms’ activities related to increasing the efficiency such as licensing marketing and supplying alliances are exploitative. There are various ‘tensions’ MNEs have to deal with regarding both the exploration and exploitation of alliances, which is also an indication of the complexity of the markets multinationals are operating in. March (1991) argued that when firms are being faced with limited resources, there will be a trade-off with regards to the allocation of the resources to either exploration or exploitation activities. This is also linked to the path dependency of firms, for the reason that “a firm’s previous investments and its repertoire of routines constrain its future behaviour” (Lavie & Rosenkopf, 2006). Finding a balance between these two polar opposites appears to be complex, as the exploration of new competencies conflicts with the exploitation of existing skills (March, 1991). Thus, the key is for a firm to “engage in enough exploitation to ensure the organization current viability” as stated by Levinthal and March (1993).

2.3.3 Ambidexterity in Emerging & Developed markets

Emerging markets, also known as developing markets, are defined as a “societ[ies] transitioning from a dictatorship to a free market-oriented economy, with increasing economic freedom, gradual integration within the global marketplace, an expanding middle class, improving standards of living and social stability and tolerance, as well as an increase in corporations with multilateral institutions” (Van de Kuil, 2008). The traditional view on emerging markets is that they are slow to reach the maturation stage to internationalise successfully due to the constraints of their institutional environment. These markets, however, are currently characterised by their fast growth and continuous changes, but also by their high uncertainty, institutional voids and hyper-competition (Lapersonne, Sanghavi, De Mattos, 2015). The (institutional) transition that emerging markets are undergoing is visible in the Outwards Foreign Direct Investment (OFDI) flows which traditionally were dominated by MNEs from the developed countries, though nowadays there is a trend in which eMNEs are

(21)

pg. 21 becoming significant investors abroad (Buckley et al., 2012). This is exemplified by the increase in China’s overseas investments which increased 13,7 times between 2004 and 2013 (WRI, 2016). This emerging market indicates an example in which the outward FDI is close to overtaking the inward FDI indicating the accelerated internationalisation eMNEs in China are experiencing.

The question remains as to how the alliance portfolios of eMNEs influences the rapid internationalisation and how this again is related to firm performance. Especially since eMNEs might not have the innovative capability to exploit disruptive new technologies in comparison to dMNEs as stated by Lynch and Jin (2016). However, it is of importance to remark that successful eMNEs often have (privileged) access to the main country-specific assets (CSAs). Various studies undertaken in emerging economies have indicated that the business environments have transformed, and firms indicating ambidextrous aspects demonstrated a superior performance in comparison to firms which either chose a pure exploration or exploitation approach (Gopalakkrishna & Subramanian, 2001; Acquaah & Yasai-Ardekani, 2008). Though, this indicates the dual difficulty eMNEs are facing as they should both further develop their FSAs so that they can be utilised abroad, as well as use current FSAs so they can continue to make use of their access to CSAs (Guillén & Garcia-Canal, 2009; Tulder, 2017).

(22)

pg. 22

3. Theoretical Framework

The literature review has indicated the multiplicity of views scholars have regarding the idea that eMNEs are internationalising faster in comparison to dMNEs. In addition to this, the previous chapter highlighted the forms an alliance portfolio can have (exploration, exploitation or balanced) and the influence this can have on the level of internationalisation and consequently firm performance. This thesis delves into the question surrounding the firm performance of MNEs related to their level of ambidexterity.

3.1 Ambidexterity and firm performance

Firm performance has been key to strategy and organizational research. Especially now that EMNEs are internationalizing faster, and with the recent surge in alliance formation, it is interesting to analyse how this influences the firm performance. As previously discussed, the configuration of an MNEs alliance portfolio may be of influence on the firm performance. An alliance portfolio can either be focused on exploration – thus seeking new competencies, exploitation – the further development of current competencies, or a balance between these two. This perspective allows us to view exploitation as firms forming alliances with firms within the same industry group, and exploration as firms forming alliances with firms outside of the industry group. Both exploration and exploitation are founded on differing structures and processes, but this is also related to aspects such as corporate culture. He and Wong (2004) indicate that exploration is seen in firms characterised by organic structures, autonomy and loosely coupled systems, whilst exploitation is associated with mechanistic structures, path dependence and routinization. The difference between the two is also reflected in the firm performance as explorative firms will enjoy a more variable performance and exploitative focused firms a stable and certain firm performance. March (2003) stresses the difficulty in finding a balance between exploration and exploitation and states that “balancing [these two]

(23)

pg. 23 is involved in balancing unity and diversity, just as in balancing competence an imagination, or in balancing reliability and risk”. In the case that there is too much focus on exploration this could ultimately lead to organizational chaos, and too much focus on exploitation could lead to inertia of the firm (Levinthal & March, 2003).

Therefore, being an ambidextrous organization demands MNEs to “not only have the operational capabilities and competencies to compete in existing markets, but also the ability to recombine and reconfigure assets and organizational structures to adapt to emerging markets and technologies” (O’reilly and Tushman, 2008). Key of ambidextrous firms is that they are capable of simultaneously exploiting competencies and exploring opportunities, allowing them to adapt faster to the local market and accomplish longevity. We, therefore, propose that:

H1: International ambidexterity is positively associated with firm performance

Whilst, ambidexterity is linked to an increased longevity of firms and consequently a higher firm performance, Leventhal and March (1993) indicate it is difficult for firms to “engage in enough exploitation to ensure its current viability, and devote enough energy to exploration to ensure future viability”. Finding that specific balance between these two aspects is hard to determine. In addition to this hardship, firms are facing difficulties with allocating their resources to a specific activity, and are therefore often constrained in allocating their resources to a single activity which consequently may lead to either excessive exploration or exploitation.

3.2 The origin of a firm

Whilst it is often assumed that Globalization leads to the convergence of cultural, political, social and economic aspects of life, MNEs remain dependent upon their local environment to sustain their competitive advantage (Noorderhaven & Harzing, 2003). Firms draw from their home country factor endowments, allowing for perfectly unique competitive advantages. It can, therefore, be argued that the origin of a firm will consequently influence its global strategy. This indicates that the firm performance, and consequently the level of internationalisation of

(24)

pg. 24 firms, is determined by the institutional fabric of a nation – often mentioned as ‘the rules of the game’. These ‘rules’ of the game are enforced by formal actors such as the government, however, in the case that a firm is supported by the government it will allow firms to offset their liability of foreignness abroad (Buckley, Clegg, Cross, Lio, Voss & Zehng, 2007).

It is recognisable that eMNEs make use of international expansion as a springboard to get access to resources necessary to increase their competitive advantage. Key within this springboard strategy is that firms link their international expansion with their home base (Luo & Tung, 2007). Examples of this are companies which either reorganized their production base to meet increased global sales for high-end products or rebranded their home-made products after using foreign technologies (ibid.). This related to the previously mentioned understanding that the international growth and success of eMNEs is dependent on their home-base performance, and thus eMNEs capitalise on this. In the ideal situation, firms will have a balanced alliance portfolio, however, in most cases, this is difficult to acquire due to various (institutional) tensions. EMNEs, for example, are characteristically both asset-, and opportunity seeking to fill their resource void and lessen the home-base institutional constraints (ibid.). Which is why it can be posited that in the case of an intensified and uncertain environment eMNEs will engage in activities such as exploration, strategic renewal and innovation (Lapersonne, Sanghavi & De Mattos, 2015). This is a clear example whereby the home (institutional) environment influences the strategy implemented by MNEs. Therefore, this thesis suggests that the country-of-origin influences the firm performance of MNEs:

H2: The relationship between international ambidexterity and firm performance is positively moderated by firm´s country of origin

(25)

pg. 25 3.3 Conceptual model

To summarise the hypothesis states in the subchapter above, the conceptual model is shown below in figure 1. Figure 1 Conceptual Model Firm performance Origin of MNE Alliance portfolio H2 H1 Control variables: - Firm level: Firm size and Firm age - Industry level: Industry Exploitation Exploration

(26)

pg. 26

4. Methodology

The goal of this thesis is to analyse whether there is a difference in alliance portfolio configuration of eMNEs and dMNEs. This research will make use of quantitative research methods in which it would be able to find a general trend related to the impact alliance portfolio’s related to the exploration and exploitation have on the firm performance of MNEs. This study made use of the Thomson One Database to identify alliances portfolios in emerging and developing markets.

3.1 Data and sample

To answer the research question, a dataset was constructed covering alliance activity in the manufacturing division. This study extracted alliance data using the Thomson One database which includes alliance cases from 1988 onwards, and is deemed one of the comprehensive databases regarding alliance information (Sampson, 2005), and has been used in various other research papers concerning this topic (Jiang, Tao & Santoro, 2010). Furthermore, the Orbis database, provided by Bureau van Dijk, is used to include both financial and company information regarding the firm size and year of incorporation. This global company database includes over 100 million companies of listed companies within a 10-year window. Combining the data aggregated in the Thomson One database with the data from Orbis allows for an overview of alliances in the BRIC countries and the US.

The choice to target any firm in the manufacturing sector (Standard Industrial Classification [SIC] 20-39) is because it was of importance to achieve an objective view of MNEs and their alliance portfolio behaviour – and therefore not only MNEs which have already been deemed successful. Also, this sector includes both low- and high-technology industries such as wood manufacturing industries (SIC 24) and the medicinal chemical industry (SIC 28). Furthermore, the manufacturing division is one in which alliances are formed

(27)

pg. 27 intensively which would give more meaning to the collected data and variables. This division, characterised by continuous changes can be viewed as increasing the level of uncertainty. Goultz (2014) indicated that alliances formed in dynamic industries are directed at reducing the complexity.

This study focused on multinationals coming from both emerging and developed markets, which will make up the overall sample in this research. The emerging economy considered within this research are the BRIC - Brazil, Russia, India, China – nations. These countries are considered for the reason that “the four BRIC countries are home to 41% percent of the world population and account for 23% of world GDP between them” (UNCTAD, 2016). We are aware that these emerging markets are not homogeneous, however, they are confronted with similar motives, and experience similar regulatory constraints. The developed market taken into consideration is the US, because they are the largest outward investor in the world (ibid.) Table 2 gives an indication of the OFDI of the selected countries in relation to each other. As this study focuses on internationalisation, only cross-border alliances will be taken into account as they will allow for better insight into the international alliance portfolio of firms.

Table 2 Outward FDI flows, % of GDP, for selected countries, various years (UNCTAD, 2017)

Country 2006 2009 2012 2015 Emerging markets Brazil 50.14 -11.93 -4.33 3.48 Russia 10.69 9.99 5.00 7.27 India 17.17 13.67 5.07 3.78 China 6.18 10.75 9.99 11.03 Developed markets United States 8.02 9.90 9.77 8.35

The data sample consisted of MNEs who have completed a cross-national alliance in the manufacturing division (SICs 20-39) in the period from 2006 – 2015. This timeframe of 10 years allows for a comprehensive sample and increases the likelihood that the information on

(28)

pg. 28 alliances has been included in the Thomson One Database. The sample includes 1511 firms over a period of 10 years, with a total of 3796 alliance cases. Within this sample, there are 1162 eMNEs (30%) from the BRIC countries and 2634 dMNEs (70%) from the US, as shown in Table 3.

Table 3 Dataset

Number of cases

Exploitation Exploration Total

Developed markets 1589 1045 2634

Emerging markets 753 409 1162

Total 2342 1454 3796

Table 4 gives an overview of the complete dataset used within this research.

Table 4 Overview dataset

N Mean Min/Max S.D. Firm age 3796 30.23 0/62 17.771 Firm size 3796 25630.53 1/380200 50683.016 Industry 3723 13.62 1/20 5.222 ROA 3796 -2.95 -10.16/4.82 1.142 Alliance 3796 .38 0/1 .486 Origin of MNE 3796 .69 0/1 .461 3.2 Measures

This section will give an overview of the variables used in this study. An indication will be given of their relevance, the way they are measured and obtained. An overall summary of the variables is also given in Table 5.

3.2.1 Dependent variable.

The dependent variable within this research is the ‘firm performance’. Al-Matari, Al-Swidi and Fadzil (2014) indicate that the firm performance is related to measuring the efficiency and effectiveness of the actions taken by a firm. Within this research, the firm performance is limited to the (quantitative) finance performance related to the Return on Assets (ROA); which

(29)

pg. 29 indicates the net income over total assets at the end of the year and thus can give an indication of annual firm performance.

3.2.2 Independent variable

Ambidexterity – thus exploration vs. exploitation alliance ratio – is taken as the independent variable. Within this study, the level of exploration is defined as alliances formed which are very distinct to the current alliance portfolio of the firm and had a focus on developing new competencies (e.g. R&D alliances). The level of exploitation is operationalised as being the alliances formed which are closely related to the firms’ current alliance portfolio, and have a focus on developing the current competencies (e.g. marketing alliances). This will be looked at by making use of the SIC codes up to three digits which indicate the industry group. Therefore, if a firm moves out of the industry group it is considered as being exploration, and within the industry is considered exploitation. Explorative and exploitative alliances are distinguished by using dummy variables, whereby exploitative is given a ‘0’, and explorative alliances a ‘1’. Therefore, ambidexterity, a perfect balance between exploration and exploitation was given a 0.5. By adopting a combined perspective, thus viewing ambidexterity as being the amalgamation of both constructs, we allow for orthogonal outcomes.

3.2.3 Moderating variable

Origin of MNE. This research will consider the origin of the MNE as the moderator, and thus influencing the relationship between the type of alliance portfolio of a firm and their firm performance. We acknowledge that firms originating in differing countries, draw from a different set of factor endowments, cultural traits, and socio-political infrastructure which gives each firm its own unique set of factor endowments (Sethi & Elango, 1999). The origin of the MNE will be defined as being either an eMNE (Brazil, Russia, India, China) or a dMNE (US). These are given dummy variables, whereby eMNEs receive a ‘0’, and dMNEs a ‘1’.

(30)

pg. 30 3.2.4 Control variables

The control variables used are industry, firm size and firm age as these three influence the firm performance of a company.

Industry. It is important to control for the industry since industries differ among themselves, as they offer market opportunities and impose operational constraints which in turn leads to a greater within-group homogeneity and between group heterogeneity (Sethi & Elango, 1999). These industry differences may therefore possibly influence the Return on Assets of a firm, and subsequently the firm performance within this sample. The ROA is measured as the ratio between net profit and the total assets and is thus highly dependent on the industry as for example, an industry in which there are relatively little assets (e.g. technology industry) will expect a higher ROA in comparison to an industry with a high number of assets. This sample only analyses companies within the manufacturing sector, which includes twenty various industries. This study controls for industry idiosyncratic effects by including all industries within the sample. The dummies created are shown below in Table 4.

Table 5 Dummy variables Industry

SIC Industry group Dummy

20 Food and Kindred Products 1

21 Tobacco Products 2

22 Textile Mill Products 3

23 Apparel and other Finished Products Made from Fabrics and Similar Materials 4

24 Lumber and Wood Products, except Furniture 5

25 Furniture and Fixtures 6

26 Paper and Allied Products 7

27 Printing, Publishing, and Allied Industries 8

28 Chemicals and Allied Products 9

29 Petroleum Refining and Related Industries 10

30 Rubber and Miscellaneous Plastics Products 11

31 Leather and Leather Products 12

32 Stone, Clay, Glass, and Concrete Products 13

33 Primary Metal Industries 14

34 Fabricated Metal Products, except Machinery and Transportation Equipment 15 35 Industrial and Commercial Machinery and Computer Equipment 16 36 Electronic and other Electrical Equipment and Components, except Computer

Equipment 17

(31)

pg. 31

38 Measuring, Analysing, and Controlling Instruments; Photographic, Medical and Optical Goods; Watches and Clocks

19

39 Miscellaneous Manufacturing Industries 20

Firm size. This variable is usually controlled for, as typically, larger firms are often more likely to have the capacity and resources to internationalise. Firm size is considered by measuring the number of employees. The firm size was indicated by taking the average over the 10-year period and taking the natural logarithm transformation.

Firm age. Within this research firm age is related to the experience companies have built up over time. Extant literature has stated that firms with a greater experience will enjoy advantages over competitions (Sampson, 2005). Older firms may have more knowledge yet are often more unwieldy, while younger firms may be more flexible but lack the resources. Firm age is measured by taking the difference between the year the firm was founded till the year this study was done - 2017.

Table 6 Variables

Variable Description Source

Firm performance Return on Assets – calculated as: ROA: Net income/ total assets

Orbis

Wharton Research Data Services Ambidexterity Difference in SIC code to 3 digits (200 – 399);

- Exploration: MNE moves out of the industry group thus <200 or >399. - Exploitation: MNE moves within the

industry group thus >200 or <399.

Wharton Research Data Services

Origin of the MNE EMNE: BRIC countries DMNE: US

Thomson One

Industry Manufacturing Sector (SIC 20-39) Thomson One

Firm Size Number of employees Orbis

Firm Age Difference in year of firm founding and year of this study – 2017.

(32)

pg. 32 3.3 Method of analysis

This subchapter will highlight the method used to analyse the hypotheses. The hypotheses are tested using a regression analysis as this will give a clear indication of the predictability of the independent variable (alliance portfolio) on the dependent variable (firm performance). Also, the moderator effect of the MNEs origin is studied by analysing the interaction effect in the regression. The formula used to analyse this relationship is the following:

!"#$ &'#()#$*+,' = ._/ + .₁23 + .₄25' + .₆7"8' + .₉: + ;_<

In this regression model, AP is the ambidexterity, Age is the firm age, size the firm size, and I is the industry. Firm age, firm size and industry, are the control variables, β0 is the constant, β denotes the parameters (regression coefficients), and ε is the error term (difference between real of the dependent variable and predicted value of the model) within this regression.

Hypothesis 1 is tested by making use hierarchical regression analysis which gives an indication of the (strength of the) relationship between ambidexterity and firm performance. Hypothesis 2 takes a moderator into account and thus requires an adjusted regression equation: !"#$&'#()#$*+,' = ._/+ .₁23 + .₄25' + .₆7"8' + .₉: + .₉=#" + .₉ =#" ∗ 23 + ;_<

The relationship between alliance portfolio and firm performance is moderated by a dichotomous variable – origin of the MNE. Here, the interaction effect between the origin of the MNE and the alliance portfolio will be tested to analyse whether this effect can predict the firm performance.

(33)

pg. 33

4 Analysis and Results

This chapter presents the results of the statistical tests which were applied to test the hypotheses. In the first part of the chapter, the bivariate analysis is presented, followed by the regression analysis. The sample used in this research consists of 1511 companies within the developed and emerging markets, whereby 40% engaged in only one alliance and 60% of the sampled firms are involved in two or more alliances.

Before performing the analysis, all variables were checked for a normal distribution. The firm performance (ROA) and firm size showed various outliers. For this reason, we took the natural log for these two variables, leading to a normal distribution.

4.1 Bivariate correlation analysis

This correlation analysis analyses the presence of correlations between the variables considered in this study. This is of interest before performing the regression analysis to validate that there is no presence of multicollinearity in the variables, making the results unreliable. To be cautious, the Variance Inflation Factor (VIF) was looked at - all values registered a VIF-value below 10 indicating there was no multicollinearity present. Table 7 presents the correlation between the variables in the dataset.

Table 7 Descriptive statistics

MEAN S.D. 1 2 3 4 5 6

Firm Age 30.23 17.71 1

Firm Size (log) 8.52 2.35 -.073** 1

Industry 13.6215 5.22 -.024 .002 1

ROA (log) -2.95 1.14 -.011 .089** -.032 1

Origin MNE 1.30 .461 -.001 -.004 .016 -.051** 1

Ambidexterity .38 .48 .025 -.061** -.100** .041* -.042** 1

*p<0,05 **p<0,01 ***p<0,001

The above table indicates that firm age is negatively correlated to firm size (r=-0.073/ p<0.01). Whilst this is an interesting correlation indicating that firm age has a negative effect on firm

(34)

pg. 34 size, this correlation is not a focus point within this research as both are considered control variables. In addition, there is a positive correlation between the firm size and the ROA (r-0.089/ p<0.01), indicating that the growth of firms in size (employees) positively influences the return on assets of a firm. Interesting items highlighted by this bivariate correlation analysis is the negative correlation between the firm size and ambidexterity, the positive correlation between ambidexterity and the firm performance, and the negative correlation between the origin of the MNE and ambidexterity. These relationships are of interest for our first hypothesis which assumes a positive relationship between ambidexterity and firm performance.

However, as previously stated the firm performance of an MNE is influenced by the country in which it operates. This is related to the country-, and firm-specific assets an MNE is faced with. Whilst emerging countries face a lot of regulatory difficulties, they may benefit from being close to (natural) resources benefitting their production processes, and the same contrast counts for MNEs from developed markets. Therefore, Table 8 presents the correlation coefficients for both markets separately.

Table 8 Descriptive statistics (market specific)

De ve lo p ed Ma rk et s Mean SD 1 2 3 4 5 Firm Age 30.24 17.529 1 Firm Size 8.528 2.353 -.063** 1 Industry 13.567 5.285 -.043* .022 1 ROA -2.911 1.140 -.018 .065** -.030 1 Ambidexterity .40 .489 .014 -.060** -.104** .024 1 Em er gi n g ma rk et s Mean SD 1 2 3 4 Firm Age 30.22 18.315 1 Firm Size 8.509 2.341 -.095** 1 Industry 13.743 5.078 .020 -.043 1 ROA -3.038 1.140 .006 .144** -.036 1 Ambidexterity .35 .478 .048 -.062* -.090** .074* 1 *p<0,05 **p<0,01 ***p<0,001

It is striking to see that the correlation coefficients for both markets are quite similar with regards to the variables. However, within the emerging markets, there appears to be a positive

(35)

pg. 35 correlation (r=0.074, p/ 0.01) between the alliance portfolio and the firm performance. Therefore, this analysis partially confirms hypothesis 1a indicating that there is a positive relationship between the alliance portfolio and firm performance in emerging markets.

The correlation analysis allows for insight into the assumptions made in the hypothesis, which will be analysed by a regression analysis to test the assumptions made.

4.2 Regression

To test the relationship between the alliance portfolio and the firm performance multiple regressions were performed as shown in Table 9. Model 1 includes only the control variables, Model 2 the independent variable, Model 3 and 4 analyses the effect of the moderator, and 5 and 6 include post-hoc tests.

4.2.1 Ambidexterity and firm performance

A hierarchical regression was performed to investigate the ability of an MNEs level of ambidexterity to predict the firm performance of an MNE. In the first model of hierarchical multiple regression, two predictors of firm performance were entered - the control variables firm age and firm size. This model is statistically significant F(3, 3719) = 11.393; p <.000 and explains 0.9% of variance in firm performance.

In the second model, the independent variable, ambidexterity explains 1.1% of the total variance in the model as F(4, 3718)=102.169; p <.000. By adding the ambidexterity variable an additional 0.2% of variance is explained in firm performance after controlling for firm age, firm size and industry (R2 Change = .002; F(1, 3718) = 6.446; p < .01. In this model, two out of the three predictor models were statistically significant, with firm size recording a higher Beta value (β = .092, p < .000) in comparison to the ambidexterity (β = .042, p < .01). This indicates that in the case that the firm size increases by one unit, the firm performance will increase by .045. The most striking aspect is that firms with an alliance portfolio focused on exploitation score .099 units higher regarding firm performance. This would indicate that firms

(36)

pg. 36 engaging in exploitation seem to have a higher firm performance in comparison to firms engaging in exploration.

4.2.2 Moderator Origin of MNE and firm performance

The third model in the table shown below analyses the influence of the moderator “origin of the MNE” on the firm performance. The analysis indicates that the origin of an MNE has a statistically significant negative effect on the firm performance (B=.125, SE = .040). This indicates that firms originating from emerging markets score .125 units higher regarding firm performance than developed markets. Due to the reason that this analysis yielded the above results, it would be of interest to analyse the moderator effect of origin of the MNE on the relationship between alliance portfolio and firm performance.

4.3 Interaction effects

This subchapter will analyse whether the origin of an MNE moderates the relationship between an alliance portfolio and firm performance by conducting a hierarchical multiple regression shown in model 4. To circumvent potentially problematic high multicollinearity regarding the interaction term, the variables were centred, and an interaction term between the alliance portfolio and origin of the MNE was created (Aiken & West, 1991). Model 4 indicates that the alliance portfolio (B= .095, SE = .039, β=.040, p<.01) positively influences firm performance, whilst the origin of the MNE (B=-.118, SE - .040, β=-.048, p<.001) is negatively associated with firm performance. The created interaction effect was added to the regression model, however, did not present significant results, suggesting that the effect of the alliance portfolio on firm performance is not dependent on whether the MNE originates in an eMNE or a dMNE.

Therefore, we can conclude that the results derived from this analysis do not completely coincide with the assumptions developed in this study. The type of alliance portfolio influences the firm performance of MNEs, however this is not dependent on the origin of the MNE as the results do support this assumption.

(37)

pg. 37 4.4 Supplementary analysis: EMNEs vs. DMNEs

As the interaction term proves to have an insignificant effect on firm performance within this study whilst the moderator ‘origin of the MNE’ has a significant positive effect on the firm performance we believe it is of interest to pose a follow-up question. Therefore, the following question was considered: Is International ambidexterity positively associated with firm performance in eMNEs?

Prior studies have shown that the home-base of an MNE provides for competitive advantage, due to the unique locational advantages MNEs can absorb in their global strategy (Porter, 1990; Buckley et al, 2012). Therefore, this subchapter analyses whether the level of ambidexterity in a firm can predict the firm performance in firms with differing home markets. In this analysis, the entire sample was divided into two subgroups: (i) the MNEs originating from an emerging market and (ii) the MNEs originating from a developed market. Another hierarchical regression was performed of which the results are presented in models 6 and 7 in Table 9 which yielded some interesting results.

The results of model 5, which only looks at MNEs located in emerging markets, suggests that ambidexterity is a statistically significant predictor of firm performance (F(4, 1140)=10.305, p < .000), with an R2 of .028. The firm performance is equal to -3.695 - .005 (industry) + .073 (firm size) + .001 (firm age) + .183 (ambidexterity), whereby alliance portfolio’s focused on exploitation performed .183 units better than explorative firms. However, model 7, indicates that the dMNEs alliance portfolio is not a statistically significant predictor of firm performance as F(3, 2630) = 4.550, p < .01, with an R2 of .005.

These results suggest that eMNEs engaging in exploitation seem to have a higher firm performance in comparison to eMNEs engaging in exploration.

(38)

pg. 38 Table 9 Overview regression results - Direct effects

Dependent variable: ROA

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Control Firm age .000 .000 .000 .000 .001 -.001

Firm size .043*** .045*** .043*** .045*** .073*** .033*** Industry -.007* -.006 -.007 -.006 -.005 -.006 Independent Alliance portfolio .099* .095* .183** .055 Moderator Origin of MNE .125** -.118** Interaction Alliance portfolio x origin MNE .117 Model fit F 11.393 10.169 10.958 8.617 8.071 4.007 R2 _.009 _.011 _.012 _.014 _.028 _.006 ∆R2 .008 .010 .011 .012 .024 .005 *p<0,05 **p<0,01 ***p<0,001

(39)

pg. 39 To conclude, the outcomes of the analysis do support the hypotheses to an extent. There appears to be a positive relationship between ambidexterity and firm performance. However, the origin of an MNE does not influence the relationship between ambidexterity and firm performance. These results are presented in figure 2 shown below.

Firm performance Origin of MNE Alliance portfolio .099* Control variables: - Firm level: Firm size and Firm age - Industry level: Industry Exploitation Exploration Figure 2 Overview findings

(40)

pg. 40

5 Discussion

This chapter will interpret the results presented in the previous chapter and relate them to the current level of understanding regarding alliance portfolios and firm performance. The chapter will then continue to discuss the academic relevance of the findings, followed by the limitations and future research.

5.1 Findings

This study advances the current understanding of the relationship between ambidexterity and firm performance in differing markets. Internationalisation, characterised as being a dynamic phenomenon, is traditionally seen as a process in which an MNE goes through successive stages of international expansion. Literature regarding strategic alliances mainly focuses on two overarching themes (i) alliances within a network, and (ii) alliances as the central aspect of analysis. This research focuses on the latter, however, relates this to a newer trend in the organisational literature; the increasing speed of internationalisation of eMNEs. This trend is related to the configuration of the MNEs alliance portfolio, and whether firms are investing in exploring new competencies, or exploiting the competencies already in their possession. We propose, in accordance by the literature, that the home market of an MNE is of influence on the relationship between their alliance portfolio and the firm performance with the research question:

To what extent does the alliance portfolio differ between Multinational Enterprises from emerging (eMNEs) and developed markets (dMNEs) and

The play with exploration and exploitation on firm performance : how alliance portfolios could assist multinational enterprises from emerging markets in their efforts for internationalisation