University of Amsterdam
Master Thesis in International Management
The impact of ambidexterity on internationalization speed,
contingent upon home country market-supporting institutional
strength and firm capital structure
Ziyao Jiang (11374624)
Supervisor: Dr. Carsten Gelhard
Final version
Date of submission: 23/06/2017
Msc in Business Administration – International Management Track
Faculty of Economics and Business
Statement of originality
This document is written by Student Ziyao Jiang 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.
Abstract
Organizations are facing an ever-changing and turbulent environment that becoming dynamic and flexible in organizational structure and resource allocation becomes essential. To fill in the research gaps left by previous studies, this paper studies the impact of ambidexterity on internationalization speed, contingent upon home country market supporting institutional strength and firm capital structure. Using balanced data of 134 multinational enterprises (“MNE”) from various developed countries, spanning over 10 years and of 9 industries, this paper conducts hierarchical regression analysis, multi-group analysis, moderation interaction analysis, conditional effect analysis and correlation matrix analysis to testify the hypothesis.
This paper argues that balanced ambidexterity has positive impact on internationalization speed. It also argues that MNE’s home country market-supporting institutional strength positively moderates the relationship between ambidexterity and internationalization speed. In addition, this paper postulates that a balanced capital structure of firms, as an auto-adjusting mechanism governance structure, positively moderates the relationship between ambidexterity and internationalization speed.
The empirical results give support to the hypothetical impact of ambidexterity on internationalization speed. These results have practical implications for both academia interested in this field and MNE managers. However, due to sample and time constraints, the moderating interactions do not show significant effect. Further researches are recommended to address this issue with improved dataset.
Table of Contents
Statement of originality ... 2 Abstract ... 3 Table of Contents ... 4 1. Introduction ... 5 2. Literature Review ... 6 2.1. Ambidexterity ... 6 2.2. Internationalization speed ... 9 2.3. Home-country contexts ... 12 2.4. Capital structure... 14 3. Theoretical Framework ... 15 4. Research Design ... 214.1. Data and Sample ... 21
4.2. Independent variable ... 22
4.3. Dependent variable ... 23
4.4. Moderator variables ... 23
4.5. Control variables ... 24
5. Results and Analysis ... 25
6. Discussion ... 36
6.1. Main Discussion ... 36
6.2. Limitations and avenues for future research ... 39
6.3. Scientific relevance and managerial implication ... 41
6.4. Conclusion ... 42
Acknowledgement ... 43
Appendix ... 44
1. Introduction
The business world is changing in a dramatic pace nowadays. As global competition intensifies and pace of environmental change accelerates, traditional stereotypical business routines for organizations may no longer apply. The unprecedented environment where multinational enterprises (“MNE”) compete necessitates ambidextrous strategic decisions and organizational structure(March, 1991, Hsu et al., 2013). Firms need to renew themselves by both exploiting existing competencies and exploring potential opportunities (Kitapci and Celik, 2013). The exploration-exploitation paradigm therefore has received much attention in management research (Stettner and Lavie, 2014). As a dynamic capability, how does ambidexterity influence the internationalization process, and under what circumstances and contingencies?
Internationalization being “the most profound business phenomenon of the 20’s century” has eventually led the development of MNEs(Sapienza et al., 2006, Johanson and Kalinic, 2016). As said, the world is truly becoming more interdependent that there is no place to hide. Internationalization is claimed to take place at a greater pace than theories predict (Kalinic and Forza, 2012, Johanson and Kalinic, 2016). Indeed, given the rapidly changing environment and intensified time-based competition in international market(Stalk and Hout, 1990), research on pace of internationalization has profound influence to not only academia but also to practitioners.
Witnessing the internationalization phenomenon worldwide, why firms from certain countries display common characteristics or generally perform better? Does home country have significant explanatory power to these characteristics? The limited
attention paid to home country institutions is astonishing. Little cross sectional evidence has been available on how home-country effects influence performance of firms in various countries and industries(McGahan and Victer, 2010). In addition to the contextual factors, does firm level specificities such as capital structure also influence MNE behaviors and traits? This paper will attempt to investigate the aforementioned issues and strive to reach a conceptual framework regarding how these factors relate to each other, through quantitative research method.
2. Literature Review
2.1. Ambidexterity
Organizations are facing an ever-changing and turbulent environment that becoming dynamic and flexible is necessary to counter the heterogeneous context. Previous research argues that to prosper in the long-run, firms need to simultaneously exploit concurrent strength and explore future new strength(March, 1991, Jan-Erik Vahlnea and Jonssonb, 2016). JanErik and Jonssonb (2016) argue with real-life cases such as Nokia which paid scarce attention to new smartphone design and SAAB which focused too much on innovation but effectiveness. Thus there needs to be a balance between exploitation and exploration, as explained by ambidexterity (March, 1991, O’Reilly and Tushman, 2008). March (1991) defines exploitation activities as “refinement and extension of existing competencies, technologies and paradigms”, while exploration activities as “experimentation with new alternatives”. Exploration normally includes areas such as research, variation, risk taking, experimentation, play, flexibility,
discovery, and innovation(March, 1991). Although ambidexterity is considered to have positive impact on firm performance and survival, the degree of balance between exploration and exploitation is largely contextual (Jonsson and Foss, 2011, Hsu et al., 2013).
Jan-Erik & Jonssonb (2016) continue to argue that ambidexterity as a dynamic capability, adds value to the Uppsala model by explaining how top-management gradually builds on experience of entrance into new markets. For biotechnology firms, for example, exploration may involve acquiring knowledge for potential novel molecular entities(Rothaermel and Deeds, 2004). Ambidexterity as a dynamic capability consists of different dimensions, including structural, which is the major subject of this paper, sequential and contextual(Raisch et al., 2009, Jan-Erik Vahlnea and Jonssonb, 2016). Sequential ambidexterity implies that firms concentrate one major objective after another. Contextual ambidexterity is embedded in the co-workers of firms rather than structural framework of firms. Raisch et al (2009) argue that these individuals would judge what might be the best time allocation among conflicting demands, for flexibility and adaptability. Jan-Erik & Jonssonb (2016) further argue that these dimensions may be combined to form dynamic capability. The question may arise: is this dynamic capability leading to faster international expansion?
Literature strives to study the black-box between internal ambidexterity and external output. Kitapci and Celik (2013) argue that organizational learning capacity serves as mediator between ambidexterity and productivity performance. Organizational learning capacity is defined as the ability to develop new knowledge
while improve current knowledge(Hult GTM et al., 2002). The capacity therefore needs to be cultivated by external acquisition of new skills and internally analyzing these skills with organizational change. These processes need to be nurtured, encouraged and implemented by a visionary manager, in the form of organizational culture(HH and X, 2005, Kitapci and Celik, 2013). Contributed by the knowledge-based view by Grant (1996), literature continues to argue that knowledge acquisition, utilization orientation, information sharing and dissemination orientation are all essential dimensions of learning capacity (HH and X, 2005, Kitapci and Celik, 2013). Nevertheless, few literature links organizational learning capacity to internationalization speed.
Stettner and Lavie (2014) address the relationship between ambidexterity, entry mode and alliance, arguing that balancing ambidexterity within a single mode of acquisition or alliance undermines firm performance due to “conflicting routines, negative transfer, and limited specialization.” Balancing across modes, on the other
hand, enhances performance. As mentioned, balancing ambidexterity across modes in
international expansions enhances performance. Different modes of entry are
intrinsically associated with different speed of entry. The question therefore remains as
to how speed is related to ambidexterity. Piao and Zajac (2016) in their paper touch
upon the timing of new exploration and ambidexterity. They propose that firms
engaging in more intensive repetitive exploitation are slower in exploring new market
segment. However, the analysis remains at product level and does not explicitly address
internationalization matters.
investment (“FDI”), entry modes and performance, little has been done to address the relationship between ambidexterity and the speed of international expansion. This paper will therefore seize the opportunity to address the impact of ambidexterity on firm’s speed of international expansion.
2.2. Internationalization speed
The concept of internationalization speed has been under-researched(Casillas and Acedo, 2013, Chetty et al., 2014). Several studies contribute to conceptualization, measurement and validation of internationalization speed. The concept of time compression diseconomies was raised and discussed by Vermeulen and Barkema (2002), proposing that greater speed of internationalization negatively influence performance of foreign subsidiaries, due to bounded rationality and limited cognitive scope. They generally define speed as how many foreign expansions firms undertake in a certain period of time. Internationalization speed here serves as the independent variable. Lin (2012), Hilmersson and Johanson (2015), on the other hand, measure internationalization by the number of legal entities established in foreign markets.
Chang and Rhee (2011) on the contrary posit that the conventional wisdom of time compression diseconomies does not consider the potential risk of being a later mover in fierce international competition. They further argue that speedy FDI expansions increases firm performance, when firms possess strong internal resources and capabilities, and when globalization pressure is high. Internal capability such as ambidexterity here serves as a moderator and speed as independent variable.
Mohr et al (2014) postulate that increasing internationalization speed strengthens the relationship between home-region concentration and performance. This is consistent with Kim and Aguilera’s (2015) semi-globalization view and Vermeulen and Barkema’s (2002) argument on internationalization speed. In other words, internationalization speed is only beneficial when MNEs internationalize rapidly within home region due to lower liability of foreignness, lower transportation and distribution cost. These arguments offer broader understanding of internationalization speed from different perspective, but never address speed as the subject matter.
Casillas and Acedo (2013) propose that internationalization speed is a multi-dimensional concept and categorize these dimensions as: growth in international commercial intensity (exporting intensity), commitment of resources to foreign activity, and dispersion (growth in breadth) of international market. This paper will only focus on resource commitment, since exports and breadth may be biased in reflecting dynamic capability if a firm is not export-oriented, or concentrates only on certain important markets. Chetty et al. (2014) further distinguish time to internationalization (time of establishment to first abroad operation), and speed of internationalization (process over time). This paper will only investigate the latter.
Some literature argue that internationalization process is never linear and incremental, but rather “leapfrogging” in the form of inconsistent steps(Kalinic et al., 2014, Johanson and Kalinic, 2016). They posit that internationalization is better explained by acceleration and deceleration, which are consequences of firms’ knowledge absorbing and integrating capabilities. Johanson and Kalinic (2016)
continue to argue that a change from acceleration to deceleration indicates the process of knowledge absorption and integration, while a change from deceleration to acceleration indicates opportunity exploration. Nevertheless, acceleration and deceleration are highly inconsistent from time to time and may dampen the visibility of ambidexterity impact on expansions.
Mohr et al.(2014) propose internationalization speed as a moderator to home region concentration and MNE performance. In the same vein, Chang and Rhee argue that (2011) firms’ tangible slack resources and intangible capabilities positively moderate the relationship between FDI speed and performance. Although sufficient research addresses the relationship between speed and performance(Li et al., 2015, García-Garcíaa et al., 2017), little research so far considers speed as a dependent variable and associates speed with ambidexterity explicitly. A study by Powell (2014) investigates the relationship between performance and speed, with speed being the dependent variable. He concludes with a U-shape relationship between profitability and entry speed. Ching-I and Chen (2012) argue that greater FDI accelerates investment speed, with the example of Taiwan investment in China. Luo and Peng (1999) suggest that greater experience in foreign market leads to greater capability and knowledge. As knowledge increases, MNE perceives less risk thus gradually increases resource commitment(M and P, 2007). However, these literatures do not explicitly link speed to ambidexterity, which is a dynamic capability. This paper fills these gaps and additionally considers strength of market supporting institution and firm capital structure as moderators.
2.3. Home-country contexts
Although host country institutional contexts have been adequately studied previously(Hitt et al., 2004, Hitt et al., 2016), limited research links home country contextual factors to ambidexterity or internationalization speed. Previous studies suggest that home country institutions play a key role in shaping MNE’s strategic decisions(Kostova et al., 2008, Geleilate et al., 2016). Arguing from institutional-based view, firm's strategy is a product of its home country's institutional profile(Kostova, 1997, Geleilate et al., 2016).
Geleilate et al. (2016) propose that home country capital market development positively moderates the relationship between multi-nationality and performance. Formal institution such as capital market provides MNEs with high level leverage for foreign business(Fernandes, 2011). Strong home country capital market enhances operational efficiency and firm specific assets, thus lowering the cost of internationalization and enabling the firm to be more competitive in international market(Geleilate et al., 2016).
In the same vein, formal institutions, which include intellectual property protection laws and regulations, are important determinants of firm actions and decisions, especially for high technology firms (Batjargal et al., 2013). Garcia-Garciaa et al. in their work (2017) demonstrate that proprietary technological knowledge steepens the relationship between speed and long term performance. Countries with stronger intellectual property protection institution prevent proprietary technology knowledge from being exploited by competitors or new entrance. The stronger
institutional protection thus supposedly in turn steepens the relationship between speed and performance. Nevertheless, ambidexterity remains under addressed.
As some literature have already proposed, country specific advantage may enhance capability of firms in exploring internationalization opportunities and translate multi-nationality into greater performance(Buckley and Casson, 1976). Geleilate (2016), in his same study mentioned above, also argues that better regulatory quality in home country positively moderates the relationship between multi-nationality and performance. He further argues that in circumstances of under protection by proper formal regulatory framework, fair competition is unlikely and customer option is reduced. He then emphasizes the importance of credibility and enforcement capability of the effectiveness of state regulatory framework (Geleilate, 2016). However, all these studies do not touch upon ambidexterity.
Landau et al. (2016) in their work argue that firms develop institutional leverage capability (ILC), a dynamic capability that integrates home and host country institutional contexts to form competitive advantage. They also propose that firms’ institutional competitive advantage, generated by ILC, strengthens internationalization ability. However, this dynamic capability ILC distincts from ambidexterity and does not link to internationalization speed. Despite most studies investigate the relationship between home country formal institutions and performance(McGahan and Victer, 2010, Marano et al., 2016), few associates formal institutions to speed of internationalization or ambidexterity.
2.4. Capital structure
Capital structure has been insufficiently addressed within the field of international management. Chen et al. (1997) in early studies argue that debt to equity ratio is positively related to the degree of internationalization within MNEs. Due to internationalization, MNEs generally have lower leverage ratio than domestic firms, after controlling for bankruptcy cost and growth opportunities. However, they posit that high leverage restrict growth options of MNEs by incurring more debt obligations. Chkir and Cosset (2001) add on by bringing product diversification dimension to the study. They suggest that leverage increases with both international and product diversification. However, they conclude that profitability of MNEs is negatively related to the debt ratio.
Similar to Chen et al., Singh and Ali (2004) further claim that MNEs with higher international diversification level tend to have greater long-term debt financing. They argue that after controlling for composition of debt financing, agency cost and asset structure, higher level leverage still reduces the overall cost of capital, despite associated increasing equity risk. However, none of these studies associate capital structure with ambidexterity and speed dimension of internationalization.
Some argues that debt financing functions as hedging tool against exchange and political risk in target market by borrowing in local debt market(Burgman, 1996, Manohar Singh and Nejadmalayeri, 2004). Firms also issue debt dominated in foreign currency, to hedge their exposure in the underlying currency(Kedia and Mozumdar, 2002). With decreased political and exchange risk, firms tend to facilitate faster
international expansions to maximize their return of intangible assets and resources. However, it is notable that high leverage increases credit risks and these studies do not associate leverage with ambidexterity.
In short, this paper will attempt to fill in the aforementioned research gaps and contribute to the existing literature, by attempting to answer the following research question: what is the impact of ambidexterity on the internationalization speed of MNEs, contingent upon the home country market supporting institutional strength and firm specific capital structure.
3. Theoretical Framework
Literature emphasizes the importance of simultaneous exploration and exploitation, and ambidexterity as a dynamic capability is found to have positive impact on performance and firm survival(Hsu et al., 2013, Jan-Erik Vahlnea and Jonssonb, 2016). Teece et.al (1997) offer the definition of dynamic capability as a company’s ability to “integrate, build and reconfigure internal and external competencies to address rapidly changing environments”. Key components of dynamic capability, according to Helfat et al (2007), are organizational structure and processes that nurture learning and innovation, business design, decision frames and heuristics that inform long-term investment. As mentioned early, organizational learning capacity as a dynamic capability, serves as mediator between ambidexterity and performance.
One pre-requisite for firms conducting rapid international expansion is the organizational learning capacity. MNEs acquire and merge foreign entities to acquire
explicit knowledge such as strategic assets, and tacit knowledge such as local customer knowledge. After the acquired knowledge is fully digested, firms continue to expand thus to avoid time compression diseconomies. Referencing arguments from Kiptapci and Celik (2013), this paper infers that ambidexterity facilitates organizational learning capacity, thereby influencing expansions speed. O’Reilly and Tushman (2008) regard ambidexterity as firm’s ability to “reconfigure existing assets and develop new skills to address emerging threats and opportunities”. The complementary combination of exploration and exploitation, together as a dynamic capability, in my expectation, is leading to greater speed of internationalization.
Piao and Zajacc (2016) argue in their research that by focusing heavily on repetitive exploitation, firms accumulate little reservoir of novel alternatives of solutions for emerging market segment, thus creating a mismatch between the actual market demands and what the firm could offer. These firms desensitize themselves to the changing environment and therefore will react slower in “estimating, absorbing and exploring” new segments (Piao and Zajac, 2016). This corroborates my expectation that imbalance of exploration and exploitation reduces the internationalization speed. This literature neither goes beyond market segmentation to internationalization level nor testifies the impact of over-exploration on exploitation. Nevertheless, Piao and Zajac (2016) stress that, the synergistic combination of exploration and exploitation across time over the competitive dynamics, may be the “ultimate solution”.
Some may argue that exploration over exploitation by nature enhances internationalization. This paper however will attempt to prove that over-exploration
only boosts internationalization speed in the short term; while in the long term, speed drops since the inefficiency in exploiting existing competencies deprives firms’ capability in deploying exploration resources in foreign markets. The synergic combination of exploration and exploitation is the key to rapid internationalization expansion. Consequently I posit that:
Hypothesis I: Ambidexterity increases firms’ internationalization speed.
Literature argues that strong regulatory environment at home country offer MNEs, from developed countries, safeguarded property rights and reputable brand names in international market(Acemoglu and Johnson, 2005b, Geleilate et al., 2016). Strong enforcement of intellectual property rights supposedly grants MNE strong incentive and protection to conduct exploratory activities. In other words, strength of home country market-supporting institutions constitutes a premise to both ambidexterity and speedy internationalization.
In developed regulatory and enforcement environment, fair market and competitions are ensured as facilitated economic transactions(Acemoglu and Johnson, 2005a). Improved regulatory system results in better monitoring of “anti-competitive conduct” thus leading to a more efficient and competitive environment(Dutz and Vagliasindi, 2000). Hence, stronger intellectual property protection institutions enhance competitiveness and efficiency of firms in international market, thus leading to greater speed of internationalization expansions in the long term.
reputation and credibility. These attributes in turn help firms, using their strong brand in intellectual right enforcement and reputation, in overcoming trust issues, thus reducing cost of liability of foreignness in foreign expansions. With reduced cost and strong brand image in right enforcement, firms would have less restrictions and more resources and flexibility in international expansions (Geleilate et al, 2016).
Countries with stronger formal institution for intellectual property would also facilitate the development of better corporate governance practices, since these countries encourage best practices such as Chinese walls, segregation of duties, patents rights and management of conflict of interests. These practices would in turn enhance the relationship between internal capability and external expansion, since the countries with strong market supporting institution provide a pre-requisite and a platform for firms to exploit their firm specific advantages.
As for developing economies, Geleilate et al. (2016) argue that strong enforcement of property rights does not significantly impact internationalization performance. This is because EM MNEs tend to develop internationalization with non-market resources such as networks, affiliations and state-interventionism (Gaur et al., 2014, Geleilate et al., 2016). This paper thus will only focus on firms from developed countries. In short, I postulate that:
Hypothesis II: The MNE’s home country market-supporting institutional strength positively moderates the relationship between ambidexterity and internationalization speed.
informal and formal institutions, corporate level governance such as capital structure plays essential role guiding economic activities. Manson (2011) argues that some managers as shareholders’ agents tend to be unfavorable of exploration due to distant and uncertain return. Debt intrinsically contains more restrictive covenants such as meeting liquidity requirement and financial obligation that closely monitors managers(S and MR, 2008). In case of failing to meet obligations, bankruptcy may lead to decrease in compensation of managers and employment contract termination (Choi et al., 2016). Choi et al. (2016) add on to Manson by suggesting that managers may also over-explore at the expense of shareholders, in order to foster their skills and human capital for private benefit. They argue that debt and equity financing offer complementary governance mechanisms with respect to ambidexterity. Equity financing favors exploration due to hierarchical control and flexibility for experimentation, while debt financing imposes governance on innovation with bankruptcy risk and cash flow obligation.
In lack of complete information of what constitutes optimal balance of exploration and exploitation, the countervailing forces of debt and equity form autonomous adaption. When the exploration-exploitation ratio deviates, managers and shareholders adjust the capital structure commensurately to balance the trajectory(Choi et al., 2016). Given the autonomous mechanism, they propose that when firms are over-explorative, leverage will increase thereafter as equity becomes too costly to fund ongoing innovations. This is because equity holders will significantly discount the firm value for over-exploration. In addition, credit providers will be unwilling to provide
additional funds to the innovations and R&D. Without balancing exploitation, firms will face increasing risk that credit providers may demand higher return. Similarly when firms are over exploitative, equity financing becomes expensive because equity holders may offer low valuation of enterprise value. Equity investors may regard the firm as uncompetitive without future appreciation potential.
Therefore, balanced capital structure constitutes a premise for both ambidexterity and speedy internationalization. Some may argue that debt inherently favors slower internationalization due to financial obligation and risk of bankruptcy. However, this paper will attempt to prove that over the long run, debt provides close monitoring and autonomous adaption mechanism to managers and shareholders. The balance mechanism, as an interaction factor with ambidexterity, is also a key to rapid internationalization expansion in the long run. I therefore posit that:
Hypothesis III: The balanced capital structure of firm positively moderates the relationship between ambidexterity and internationalization speed.
To illustrate the hypothesis proposed, the following conceptual model is constructed: Figure 1: Conceptual Model
Ambidexterity (balance of exploration and exploitation)
Greater speed of
internationalization of MNE
MNE’s home country market-supporting institution strength
Firm’s balance of capital structure
4. Research Design
4.1. Data and Sample
The overall sample population is based on all public international corporations across developed countries over the period of 2007 to 2016. Their internationalization ratio of foreign sales to total sales must be above 0. This implies that the sample population firms must have international transactions or operations, in order to be assessed on the internationalization aspect. The sample firms exclude those already ceased operation, or acquired by other firms between the prescribed period. By implying these criteria, this paper ensures that the firms are comparable and active in the period examined (Georgakakis and Ruigrok, 2017). Before data analysis, normality and outlier checks are conducted to ensure the quality of data and the subsequent outcomes.
This paper will only consider number of acquisitions, mergers and joint ventures per year as speed of internationalization, as these deal types normally involve internationalization on regular and readily examinable bases. The impact of exploration and exploitation takes effect on firms over a long time. The 10 year time span reduces the bias of irregular expansion and justifies hypothesis through long-run effect.
As argued by Geleilate et al.(2016), strong enforcement right does not impact internationalization significantly in developing countries. In addition, firms from emerging economies expand rapidly into world market so as to overcome their late-comer disadvantage(Peng, 2012, Lebedev et al., 2015), not for economic but for strategic concerns. Hence only developed counties are considered.
Services (“WRDS”) and Capital IQ, which are award-winning research platforms and business intelligence tools for global corporates. For operationalization of ambidexterity, capital structure and some control variables, this paper uses the financial information from WRDS and Capital IQ. As for the speed of internationalization, this paper collects data from Zephyr, a comprehensive database for merger and acquisition deals. This paper also uses the reports published by World Economic Forum (“WEF”) to assess the score of intellectual rights protection in each individual country. [See Appendix A for a complete list of variables]
4.2. Independent variable
H1 lays down the core relationship of this paper by analyzing the impact of ambidexterity on internationalization speed. Instead of using values which differ firm by firm, this paper refers to Swift’s (2016) method by deriving intensity, which provides comparable values for individual firms. This paper employs the absolute difference between explorative intensity and exploitative intensity as the measure of ambidexterity. By deriving and subtracting the average value of ambidexterity among all dataset, a higher value then would indicate greater degree of imbalance between exploration and exploitation. To separately operationalize exploration and exploitation, this paper follows Hsu et al.’s (2013) rationale by categorizing firms’ research and development (“R&D”) and marketing aspects as explorative, and manufacturing and producing aspects as exploitative. Accordingly, this paper measures exploration as the sum of R&D Expense and Marketing Expense, divided by Total Sales. As for exploitation, I
follow Basu et al (2008) by using Cost of Goods Sold (“COGS”) as indicator of manufacturing and producing expenses. They argue that COGS represents the aggregate figure that includes all costs incurred in producing the goods. Accordingly I measure exploitation as COGS divided by Total Sales [See Appendix A for detailed operationalization of variables].
4.3. Dependent variable
Foreign entities acquired and merged between two specific moments can be an indicator of speed of internationalization, in terms of resource commitment (Asmussen et al., 2009, Casillas and Acedo, 2013). Accordingly, to calculate the internationalization speed, this paper adopts the method similar to Vermeulen and Barkema (2002) and Lin (2012)’s, by measuring the average number of foreign acquisitions, mergers and joint venture deals per year.
4.4. Moderator variables
This paper will also investigate how the relationship between internationalization speed and ambidexterity varies in interaction with factors nested at different layers of context, namely firm level and country level. Therefore moderator variables are included to explain the impact of institutional strength and firm specific capital structure, on the relationship between ambidexterity and speed. This paper will focus on the strength of home country market-supporting institution as the contextual factor. I follow several previous literatures by using Global Competitive Index, published by WEF, as the
source of assessing Intellectual Property Protection scores(CLAESSENS and LAEVEN, 2003, Nunnenkamp and Spatz, 2004, Kang and Jiang, 2012). The score of each individual country ranges from 1-7, with 7 being the highest score.
This paper follows Choi et al.’s (2016) approach in assessing capital structure, computed as total debt (long-term debt and current portion of long-term debt) divided by book value of total assets. Similar to the approach to ambidexterity, by deriving and subtracting the average value of leverage among all dataset, a higher absolute value then would indicate greater degree of imbalance between debt and equity.
4.5. Control variables
In testing of this model, several control variables are added due to their potential impact on internationalization speed. Industry is included to control for the effect of it on internationalization speed and pattern. This paper will use Global Industry Classification Standard (“GICS”) to group sample companies(MSCI, 2015). The GICS categorizes 11 industry sectors consisting of energy, materials, industrials, consumer discretionary, consumer staples, healthcare, financials, information technology, telecommunication services, utilities and real estates. This study uses industrials as the baseline industry and constructs dummy variables for each individual industry. Financials are excluded from the experimental sample because of their service based nature, and their manufacturing or producing aspect is generally missing. Real estates are also excluded due to their missing R&D and manufacturing data in the company financials.
Large firm size is normally associated with more intensive international expansions. It is computed as the logarithm of assets in a given year(Lin, 2012). Firms with poor performance tend to take greater risk and hence speedier expansions, motivated by the belief that such risk could lead to future better performance(Greve, 2003, Lin, 2012). Hence this paper includes firm performance as a controlled variable, computed by three-year average of Return on Assets (“ROA”)(Lin, 2012). Companies with higher current ratio can adapt risky environment in the foreign market(Nohria and Gulati, 1996, Lin, 2012). This paper therefore will also control for current ratio, computed as current assets divided by current liabilities.
Following Powell’s (2014) approach, contextual factor such as home-country market competition intensity is also controlled as higher intensity supposedly leads to speedier foreign market entry. Other contextual factors that may potentially impact internationalization speed are tax rate and domestic market size. These factors may force domestic firms to internationalize as means to overcome domestic disadvantage. This paper uses WEF to assess these factors as contextual controlled variables.
5. Results and Analysis
A closer inspection of the correlation matrix compiled [see Table 1] reveals a negative medium effect on the relationship between ambidexterity and internationalization speed (PC = -0.235, p<0.01). An in-depth analysis of the relationship between them will be illustrated in later sections. The independent variable is positively correlated to some control variables such as the energy industry (PC = 0.295, p<0.01) and healthcare
industry (PC = 0.316, p<0.01), as energy and healthcare sectors tend to be less ambidextrous. Ambidexterity imbalance is also negatively correlated with telecommunication industry (PC = -0.261, p<0.01). In addition to ambidexterity, the dependent variable is correlated to firm size (PC = 0.286, p<0.01), information technology industry (PC = 0.337, p<0.01) and intellectual property protection (PC = 0.188, p<0.01). Their relationships will be further addressed in later sections.
Some of the control variables are also correlated [Table 1]. Domestic market size is positively correlated to firm size (PC = 0.410, p <0.001), domestic market competition intensity (PC = 0.694, p <0.001) and tax rate (PC = 0.371, p <0.001). Firm size is positively correlated to market competition (PC = 0.311, p<0.01) and tax rate(PC = 0.330, p<0.01). Market competition in home country is also correlated to tax rate (PC = 0.352, p<0.01). The potential explanations for these correlations will be addressed in the discussion section.
Table 1: Correlation Matrix
Correlations Matrix 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1. Ambidex 2. Speed -.235* * 3. IPP -.005 .188* 4. Capital -.042 -.131 -.069 5. Energ .295** -.085 -.148 -.088 6. Materi -.162 .097 .016 -.180* -.099 7. Teleco -.261* * .066 -.078 .103 -.068 -.093 8. C.Disc -.150 -.106 -.009 -.049 -.106 -.145 -.099 9. C.Stap .024 -.053 .129 .141 -.076 -.105 -.072 -.112
Hierarchical multiple regression is performed to investigate the degree of impact of ambidexterity on internationalization speed, after controlling for industry, size, performance of firms, current ratio, home country market competition intensity, overall tax rate, and domestic market size [see Table 2]. In the first step of hierarchical multiple regression, all control variables mentioned above are entered as predictors. This model is statistically significant F(13, 120) = 2.969; p < 0.01 and it explains 24.3% of the variance in expansion speed.
In step 2, after the entry of independent variable ambidexterity, the variance explained by the model becomes 29.1%, F(1, 119) = 7.993; p < 0.01. The introduction of ambidexterity explained additional 4.8% variance of internationalization speed, after controlling for the control variables mentioned (R2 Change = 0.048; F(1, 119) = 7.993; p < 0.01). It can be assumed that for every 1 unit increase in ambidexterity imbalance, namely the targeted firm is less balanced between exploration and exploitation, speed would drop by 3.673 units per year (B = -3.673, p < 0.01). The significant level might improve if sample size increases or more industries and countries are included.
10. HealthC. .316** -.146 .151 .065 -.119 -.163 -.112 -.175* -.126 11. Infortech .013 .337** -.053 -.011 -.112 -.154 -.106 -.165 -.119 -.186* 12. Size -.013 .286** .177* -.108 .045 -.130 .150 .194* .077 -.236* * .082 13. Perform. .110 .062 .049 .027 -.154 -.083 .002 -.071 .139 .268** .157 .036 14. Cur.Ratio .099 -.040 -.006 .135 -.103 .060 -.207* -.058 -.182* .238** .169 -.179* .303** 15. Compet. .095 .059 .098 -.130 .023 -.194* -.245* * .249** .003 -.077 .154 .311** -.151 .016 16. Tax -.064 .115 .231** -.091 -.002 -.046 .044 .191* .084 -.218* -.046 .330** -.204* -.106 .352** 17. MarkSize .035 .164 .160 -.056 -.012 -.103 -.082 .134 .087 -.098 .146 .410** .027 .191* .694** .371**
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
In step 3, both moderators and the independent variable are standardized to construct interaction variables. After the entry of the interaction terms between ambidexterity and IPP, and between ambidexterity and capital structure, the total variance explained by the entire model becomes 30.2%, F(2, 117) = 0.898. The introduction of the interaction terms between moderators and the independent variable explain additional 1.1% variance of internationalization speed, after controlling for the variables aforementioned (R2 Change = 0.011; F(2,117) = 0.898; p > 0.05). Notable that the significant level of the model changes dramatically to 0.410, which is insignificant. It demonstrates that the addition of the interaction terms does not explain the model to a significant level. Moreover, the significant levels of interaction terms, namely ambidexterity*IPP (p = 0.411) and ambidexterity*Capital (p = 0.433), are both larger than 0.05. This shows that the interaction terms are not significant predictors of internationalization speed.
In the final model, three out of sixteen predictor variables are statistically significant, with industry (information technology firms) recording a higher beta value (B = 1.763, β = 0.348, p < 0.01), than firm size (B = 0.906, β = 0.272, p < 0.01) and ambidexterity (B = -3.605, β = -0.254, p < 0.01). In other words, Information technology firms are generally 1.763 units higher in internationalization speed than baseline industry, which is industrials. The results imply that information technology firms are speedier in expansions. Moreover, in average, firms with larger size implement more international expansions within a given time. In unstandardized coefficient terms, if a firm’s size increases by one unit in log scale, the speed will
increase by 0.906. If a firm’s ambidexterity imbalance increases by one unit, the speed will decrease by 3.605 expansions per year.
Table 2: Hierarchical Regression Model
Hierarchical Regression Model of Internationalization Speed
Step 1 Step 2 Step 3
β p B β p B β p B Energy -.059 .502 -.426 .022 .811 .156 .013 .888 .093 Materials .159 .092 .885 .138 .133 .769 .132 .152 .736 Telecommunication .020 .834 .150 -.034 .716 -.258 -.022 .816 -.167 Consumer Discretionary -.094 .331 -.499 -.122 .198 -.646 -.128 .180 -.675 Consumer Staple -.069 .461 -.477 -.056 .538 -.388 -.058 .525 -.400 HealthCare. .003 .975 .016 .084 .421 .410 .106 .315 .519 Information Technology .328 .001 1.665 .338 .001 1.713 .348 .001 1.763 Company Size .244 .013 .813 .264 .006 .877 .272 .005 .906 Performance .034 .719 1.165 .048 .605 1.629 .042 .655 1.415 Current Ratio -.107 .272 -.283 -.101 .282 -.269 -.099 .296 -.262 Home Country Market
Competition
-.108 .396 -.579 -.087 .481 -.468 -.103 .407 -.553
Tax Rate .069 .453 1.231 .071 .431 1.255 .081 .371 1.434 Domestic Market Size .121 .344 .229 .109 .379 .207 .115 .353 .218
Ambidexterity -.259 .006 -3.673 -.254 .007 -3.605 Ambidexterity*IPP -.070 .411 -.123 Ambidexterity*Capital .066 .433 .123 R 0.493 0.539 0.549 R2 0.243 0.291 0.302 Adjusted R2 0.161 0.208 0.206
Standard Error of the Estimate 1.660 1.614 1.615 R2 Change 0.243 0.048 0.011 F Change 2.969 7.993 0.898 df1 13 1 2 df2 120 119 117 Sig. F Change 0.001 0.006 0.410
In average, with comparison to industrials, telecommunications (B = -0.167, β = -0.022), consumer discretionary (B = -0.675, β = -0.128), and consumer staples (B =
-0.400, β = -0.058) are less prone to expansions, where consumer discretionary has the lowest speed. Energy (B = 0.093, β = 0.013), materials (B = 0.736, β = 0.132) and healthcare industries (B = 0.519, β = 0.106) are generally greater in speed of expansions, where materials has the highest speed. Nevertheless, note that their relationships with speed are not statistically significant with p > 0.05. With increasing sample size, the significant level might improve.
Firms with better performance, measured by ROA (B = 1.415, β = 0.042), generally show speedier expansions, as stronger performance enhances investor confidence thus attracts substantial financing and shareholder approval, to conduct new expansions. Firms incorporating in countries with higher tax rate show greater tendency in internationalization (B = 1.434, β = 0.081). This is because higher tax rate squeezes profitability and firms tend to transfer and disperse resources to foreign countries with better tax policies.
Nevertheless, the results do not cater to all expectations raised. Firms with higher current ratio (B = -0.262, β = -0.099) are less aggressive in expansions. This phenomenon contradicts to previous arguments raised by Nohria, Gulati (1996) and Lin (2012). Possible explanations could be that certain firms need to keep large portion of current assets such as cash and receivables, to counter liquidity issues instead of implementing expansions.
According to the results, in average, firms operating in countries with high home-market competition intensity are slower in expansions (B = -0.553, β = -0.103). A possible explanation is that countries with low competition intensity have lower
capability in facilitating domestic firms’ expansions in international domain. In other words, competitive environments and markets such as Japan and US enhance firms’ competitiveness. The statistical results did not reveal the significant relationship between home market competition intensity and speed of resource commitment in foreign expansions. Nevertheless this relationship was testified in Powell’s(2014) study, where he argues that greater competition intensity divert resource commitment in expansions.
Firms with larger domestic market size show speedier expansions (B = 0.218, β = 0.115). A possible explanation is that large domestic markets such as US nurtures competitive firms. However, notably that these relationships of control variables with expansion speed are statistically insignificant with p > 0.05. The results and explanations need further research and clarification. The significant level may improve with increasing sample size.
The regression coefficient [See Table 3] for the interaction between Ambidexterity and Intellectual Property Protection is -0.0155 and is statistically close to zero, t (128) = -0.0084, p>0.05. Thus, as elaborated above, it can be interpreted that the effect of ambidexterity on internationalization speed does not really depend on the level of IPP in home country. The regression coefficient for the interaction between Ambidexterity and Capital Structure is 8.332 and is statistically different from zero, t (128) = 0.402, p>0.05. Thus it can be interpreted that the effect of ambidexterity on internationalization speed does depend on the capital balance of targeted firm, but to a very insignificant extent.
Table 3: Moderation Interaction Table Interaction table Coefficient SE t p Intercept 1.7900 0.1550 11.5350 0.0000 Ambidexterity -3.3520 1.4170 -2.3660 0.0195 IPP 0.527 0.207 2.543 0.012 Capital structure -3.315 2.280 -1.454 0.148 Ambidext*IPP -0.0155 1.8420 -0.0084 0.9930 Ambidext*Capital 8.3323 20.7120 0.4023 0.6880 R2 = 0.108, p = 0.0019 < 0.01 F (128) = 4.039
According to R2 value this overall model accounts for 10.83% of variance in internationalization speed. The results is statistically significant (F (5, 128) = 4.039, p < 0.01). The level of overall variance explanation might improve if sample size increases, or with more industries and countries included.
To better understand the moderating effect of moderators, this paper follows Perols et al’s (2013) methodology by adopting a multigroup analysis, where each moderator sample is split into two sub-samples, namely the low value moderator groups and the high value moderator groups. The splitting to sub-samples is executed based on the median value of each moderator. Hierarchical multiple regression is conducted for each sub-sample separately, with the other moderator as an additional control variable [See Table 4].
Table 4: Multigroup Analysis for moderators
Moderation Effect - Multigroup Analysis
Moderator Split β-coefficient P-value
IPP Low Group (N=48) -0.165 0.353
High Group (N=32) -0.21 0.387
Capital Low Group (N=67) -0.309 0.597
High Group (N=67) -0.095 0.446
*p-value is significant when < 0.05
protection in a given institutional environment, greater imbalance of ambidexterity leads to lower speed, but with smaller effect (β = -0.165, p = 0.353) in comparison to the original model (β = -0.254, p = 0.007). With high IPP of a given country, greater imbalance of ambidexterity leads to lower speed, with also a smaller effect (β = -0.21, p = 0.387). The results of the study, in the same vein, show that with greater degree of capital balance in a given firm, greater imbalance of ambidexterity leads to lower speed, but with larger effect (β = -0.309, p = 0.353), in comparison to the original model (β = -0.254, p = 0.007). This is consistent with the results from the conditional effect analysis, which will be further explained below. With greater imbalance of capital structure in a given firm, greater imbalance of ambidexterity leads to lower speed, but with much smaller effect (β = -0.095, p = 0.446). However, it is notable that all these results from the multigroup analysis are insignificant, corroborating the results obtained beforehand for insignificant moderating interactions.
According to Table 3, although no significant interactions between both moderators and ambidexterity, it is observed that IPP does influence speed by itself (coefficient = 0.527, t(128) = 2.543, p = 0.012, slightly > 0.01 but < 0.05). It is therefore predicted that for every 1 unit increase in IPP, speed would increase by 0.527 units per year. This relationship can also be observed in the plotting of speed against ambidexterity, with IPP as third variable [see Figure 2]. Assuming at the same level of ambidexterity, companies operating in a country with high IPP have greater speed. No significant influence of capital structure on speed is observed (p > 0.05).
Figure 2: Plot of IPP as moderator
A closer inspection of the conditional effects indicates that the relationship between ambidexterity and speed is only significant among companies with balanced capital structure [see Table 5]. This is also observable in the plotting of speed against ambidexterity, with capital structure as the third variable, where only balanced capital structure makes logical sense [see Figure 3]. This is because in high capital imbalance group, imbalance of ambidexterity leads to higher speed. This is contradictory to the main argument of this paper.
Out of these companies with balanced capital, only those who operate in countries with average (effect = -3.949; SE = 1.755; CI: -7.422 to -0.477) and lower (effect = -3.940; SE = 1.873; CI: -7.647 to -0.233) IPP scores show significant relationships between ambidexterity and speed. Possible explanation may be that firms from high IPP countries have both high ambidexterity and high speed of expansion, thus the moderating effect of IPP is not easily observable. The results show that for
-3 -2 -1 0 1 2 3 4 5 6 7
Low Ambidex High Ambidex
In te rn ation ali zation S p ee d Low IPP High IPP
companies with balanced capital structure operating in average IPP countries, every unit increase in ambidexterity imbalance leads to a -3.949 drop in average expansions per year; for those with balanced capital structure operating in lower IPP countries, every unit increase in ambidexterity imbalance leads to a -3.950 drop in average expansions per year.
Table 5: Conditional Effect Table
Conditional effect of Ambidexterity on Internationalization speed at levels of IPP and capital structure
Capital Structure IPP Effect SE t p
-0.0717 -0.6020 -3.9400 1.8730 -2.1030 *0.0370 -0.0717 0.0000 -3.9490 1.7550 -2.2500 *0.0260 -0.0717 0.6020 -3.9590 2.2620 -1.7500 0.0830 0.0000 -0.6020 -3.3430 1.8020 -1.8560 0.0660 0.0000 0.0000 -3.3520 1.4170 -2.3660 *0.0200 0.0000 0.6020 -3.3620 1.7990 -1.8660 0.0640 0.0717 -0.6020 -2.7460 2.7180 -1.0100 0.3140 0.0717 0.0000 -2.7550 2.3110 -1.1920 0.2350 0.0717 0.6020 -2.7450 2.4000 -1.1520 0.2520 *Statistically significant
6. Discussion
6.1. Main Discussion
The outcome of this paper provides supporting evidence to the first hypothesis, that the complementary and synergistic combination of exploration and exploitation will lead to greater speed of internationalization, in the long term. The main relationship of the research question is therefore empirically proved. However, this study did not find significant moderating interactions of market-supporting institutional strength and firm capital structure on the relationship between ambidexterity and internationalization speed. Despite the conceptual framework stands true, demonstrated by the presumed signum of interaction terms, the significant level is far from expectation. In other words, according to the statistical results, the relationship between ambidexterity and speed does not depend on IPP and capital structure to a significant extent. Thus Hypothesis II and Hypothesis III are not supported to a satisfactory degree, by the empirical evidences of this paper.
Potential explanations for the insignificant results of moderations could be the small sample size, limited number of countries included or incompleteness of certain financial information in databases. IPP only represents market-supporting institutional strength to a limited extent. Other dimensions in addition to IPP could be brought in such as capital market strength, in order to observe significant moderating interactions. In addition, as shown by the statistical outcome, IPP by itself, as a contextual factor, directly influences internationalization speed. In other words, IPP could be a potential independent variable against expansion speed, instead of functioning as a moderator.
This may explain the insignificant interaction effect of IPP. Similarly, the over simplification of operationalization of capital structure as leverage ratio could potentially dampen the moderating effect.
The statistical results unexpectedly reveal the significant relationships between industries and internationalization speed. Industrials as the baseline industry generally include aerospace and airlines, construction and engineering, industrial conglomerates, professional services and transportation infrastructure. Information technology companies show significantly greater speed than industrials. Renowned firms such as Facebook and Uber have been observed to grow dizzyingly from star-up to multi-billion firms within an extremely short period of time. The business nature of information technology firms determines that rapid expansion is essential. Studies report that higher growth offer far more returns to shareholders among information technology firms. Growth also predicts long-term success and matters more than margin and cost structure for information technology firms, according to a study by McKinsey & Company (Kutcher et al., 2014). It was reported that the average growth rate in other industries is considered risky for information technology firms, as more than 90% of these firms with such growth rate would cease to exist within a few years(Kutcher et al., 2014).
Telecommunications, consumer staples and consumer discretionary are generally slower in expansions. It might be interpreted that industries such as telecommunications, media, retail, food and beverages have less need or capability for expansions comparing to industrial industries such as construction, engineering, airlines and infrastructure. On the other hand, energy, material and healthcare industries
are more prone to expansions. Potential explanation for this phenomenon is that industries, such as chemicals, oil & gas, biotechnology and pharmaceutical, are more expansion oriented due to their R&D intensive, technology based and market oriented nature.
The statistical results also reveal unexpectedly the relationship between firm size and internationalization speed. In average, firms with larger size implement more international expansions within a given time. This is contradictory to the arguments raised by previous literature (Greve, 2003, Lin, 2012). It might be argued from resource based view that firms with larger resource base, international network and experience are less restricted and more flexible in acquisition, merger and joint venture deals and transactions.
The correlation matrix shows that some control variables are correlated to each other. The significant correlation between domestic market size and domestic market competition intensity may be explained by the fact that large and advanced markets such as US and Japan facilitate competition among companies. The observed correlation between market size and firm size may be due to the fact that large firm size leads to greater advantage and higher survival rate in large markets such as US. As for the correlation between market size and tax rate, a possible explanation might be that firms are normally facing trade-offs between large and attractive market with high tax rate such as US, or small market with low tax rate such as Singapore. The correlation between firm size and home market competition intensity may be attributed to the fact that large size is essential to survive intense competition. Firm size and tax rate is
positively correlated, as higher tax rate potentially forces firms to expand aboard as to overcome the institutional disadvantage.
6.2. Limitations and avenues for future research
However, like all research, this study is not without limitation. One limitation is that this paper only uses archival data without delving into the micro-processes through which ambidexterity impacts internationalization speed(Georgakakis and Ruigrok, 2017). Hence I suggest future work to use other research designs such as multiple case studies to complement this work, investigating how micro-processes of ambidexterity impact expansion speed(Gibbeet and Ruigrok, 2010, Georgakakis and Ruigrok, 2017). Second limitation regarding ambidexterity is that the balance configuration of exploration and exploitation only partly represents international ambidexterity(Luo and Rui, 2009, Hsu et al., 2013). Future studies need to examine other different dimensions of international ambidexterity and their corresponding relationships with expansion speed.
Thirdly, as mentioned previously, internationalization speed also consists of different dimensions such as exporting intensity and breadth growth. This paper only tests resource commitment as indicator of internationalization speed. Further research may focus on other dimensions of speed. This paper only investigates time in progressive sense. Future studies could consolidate previous literature by investigating time as in first international operation(Chetty et al., 2014). This paper only considers speed with average expansions in a given time frame. However, it does not consider the
acceleration, deceleration or possible irregularity and fluctuation of expansions within a short time frame, as investigated by other literatures(Vermeulen and Barkema, 2002, Johanson and Kalinic, 2016). Future research may investigate the relationships between these factors with ambidexterity.
In addition, this study did not prove the moderating effects of IPP and capital structure. Nevertheless, it could be due to the small sample size and other unanticipated factors. Thus future studies could retest the moderating effects of the two on major variables with larger and more comprehensive dataset, under various circumstances.
Data collection and operationalization of variables are also subject to limitations. This paper has limited sample size due to time and resource constraints. Future studies may consolidate and further prove the correlations among different variables with larger sample size, covering broader industrial and institutional contexts.
Moreover, marketing and R&D only partly represent exploration as manufacturing and producing costs only partly represent exploitation. Mergers, acquisitions and joint ventures are not holistic indicators for internationalization speed. Future studies could improve this paper by including more aspects for each variable concerned.
Additionally, this paper uses a time span of 10 years from 2007 to 2016, to measure the average internationalization speed. However, due to the financial crisis in 2008 and its disastrous impact on firm financials, all companies in developed markets may be more or less affected by cutting down on investment and expansions. The statistical results thus could be biased to a certain extent due to the macroeconomic
determinant. Future studies may use data from less volatile periods or with longer time span.
Certain industries especially service-based ones such as financials, are not included as control variables as they do not have manufacturing, producing or R&D functions and thus difficult to measure. In addition, the databases used do not always show sufficient information, for example, missing marketing or R&D expenses. Future research could improve by assessing more comprehensive databases such as government or institutional reports. Additionally, this paper only focuses on developed countries. Future studies could fill in the gap by investigating the relationships among different variables in developing countries or in different institutional contexts.
The statistical results of this paper suggest that industry, especially information technology, and firm size, also have significant correlations with expansion speed. Therefore future researches could study the relationships among these variables, with industry or size as independent variable.
6.3. Scientific relevance and managerial implication
The findings of this study have clear and practical implications in further researches and in management and decision making. This paper fills in the gap of understanding the relationship between corporate ambidexterity and expansion speed. Future researchers may use this relationship in their theoretical models or conduct further studies based on the foundation of this paper.
Managers should comprehend the importance of balancing exploration and exploitation, as too much exploration detriments productivity while too much
exploitation hinders innovation. Ambidexterity as a dynamic capability is essential for enhancing shareholders’ value and nurturing long-term development. Although this paper only conducts experiment on structural ambidexterity, it should be noted that other dimensions such as contextual ambidexterity are equally important. Therefore managers should nurture organization learning and develop a flexible and adaptable structure and culture, in order to foster the cultivation of dynamic capabilities(Hsu et al., 2013). However, management should also understand that speed of international expansions needs to be balanced due to time compression diseconomies(Vermeulen and Barkema, 2002). This paper only focuses on speed rather than performance of firms. As illustrated in previous research, high speed may not catalyze or even hinder performance of firms.
The statistics also indicate the positive correlation between IPP and expansion speed. Hence governments need to improve the domestic market supporting institutional environment, such as intellectual property protection regulations, to enhance the competitiveness of domestic firms in the international market.
6.4. Conclusion
To summarize, in this paper I present theoretical constructs and statistical evidences about the impact of ambidexterity of firms on internationalization expansion speed, upon the moderating effects of firm capital structure and market supporting institutional strength. Although previous studies have addressed in detail the relationships between performance and other variables such as ambidexterity, speed and capital structure
respectively, this paper is the pioneer in addressing the relationship between ambidexterity and speed, with empirical evidence. In addition, this paper also includes firm level governance (capital structure) and institutional (intellectual property protection strength) context as moderators.
The findings validate the impact of ambidexterity, as a dynamic capability, on expansion speed. It provides a framework for future studies that investigate the relationship between internal corporate structure and external expansions. For practitioners and managers, these findings suggest that a balanced degree of exploration and exploitation activities is so important that balanced resource management and an adaptive organization structure, are essential for long term development.
Due to sample and time constraints, this paper did not directly prove the moderating effects of capital structure and IPP on speed against ambidexterity. This paper also presents some surprised findings with statistical evidence, that firm size, IPP and industry also have significant impacts on internationalization speed. Future IB research could address these relationships in separate studies. In addition, future studies could delve into the micro-process of the impact of ambidexterity on expansion speed, or investigate other dimensions of these variables.
Acknowledgement
The author would like to express great gratitude to Dr. Carsten Gelhard for his unwavering support and insightful comments.
