Master Thesis
Msc. Business Studies
International Management Track
Author: G.O. Perez Student ID: 10506535 Supervisor: Dr. I. Haxhi Second Reader: Erik Dirksen Final Draft - 28/2/2014
Multilatinas and the Internationalization
process
The impact of trade agreements on pace, rhythm and scope of
internationalization
Table of Contents
Acknowledgement ... 3
Abstract ... 4
1. Introduction ... 5
2. Theory and hypothesis development ... 7
2.1 Institutional and strategic constraints in Latin America ... 7
2.2 Multilatinas and domestic reforms ... 9
2.3 Multilatinas and regional reforms ... 11
2.4 Trade Agreements ... 12
2.5 Pace, Rhythm and Scope of Internationalization ... 16
2.6 Linking Trade Agreements with Pace, Rhythm and Scope of Internationalization ... 18
2.7 Conceptual Framework ... 24 3. Method ... 25 3.1 Sample ... 25 3.2 Dependent Variables... 26 3.3 Independent Variables ... 28 3.4 Control Variables ... 29 3.5 Data Analysis ... 34 4. Results ... 35 4.1 Descriptive Statistics ... 35 4.2 Correlation Analysis ... 37 4.3 Collinearity Statistics ... 40 4.4 Regression Analysis ... 42 5. Discussion ... 46 5.1 Limitations ... 49 5.2 Future Research ... 50 6. Conclusion ... 50 7. References ... 53 8. Appendices ... 60 Page 2 of 60
Acknowledgement
I would like to thank all the people that helped me during the process of writing my thesis. My appreciation goes out to my supervisor, Dr. Ilir Haxhi for his help and guidance as well my family and friends for their unconditional support and patience.
Abstract
In the 1980s most of the Latin American countries were focused on protecting their domestic markets with import substitution policies. By the 1990s regional reforms such as regional trade agreements were introduced to open up these economies and allowing easier access to resources. As a result new opportunities and threats emerged such as first mover advantages in the now open economies and fierce domestic competition due to new entrants. We argue that both opportunities and threats of the regional trade agreements gave the Multilatinas an extra incentive to expand rapidly in a higher scope of countries. The impact of the regional trade agreements on the internationalization process were tested using multiple linear regressions on a total of 291 foreign mergers and acquisitions covering a period of 27 years (1985-2012). The sample used in this study consists of 72 Multilatinas stemming from 6 Latin American countries; Argentina, Brazil, Chile, Colombia, Mexico and Peru. Key findings of this thesis show that indeed a short time after the regional trade agreements were concluded, Multilatinas opted for a rapid pace and steady rhythm of internalization within the regional trade agreement member countries. The contribution of this thesis is therefore not to only show that trade agreements created new opportunities and threats for the Multilatinas but also proves the notion that trade agreements allowed Multilatinas to leapfrog in the internationalization process. Additional results also showed that a combination of prior experience of e.g. domestic mergers and acquisitions gave the Multilatinas an extra advantage to expand more rapidly with a steady rhythm of internationalization.
1. Introduction
Latin American countries as well as other emerging markets have been characterized as “low-income, rapid growth countries using economic liberalization as their primary engine for growth” (Hoskisson et al. 2000). Despite the remarkable growth and opportunities within this region, the Latin American countries do however face different institutional and strategic constraints such as institutional voids and competitive disadvantages (Khanna and Palepu, 1997; Cuervo-Cazurra, 2008). These constraints may have an effect on inward as well as outward foreign direct investment (FDI). Different scholars have focused on inward FDI by developed MNEs into these Latin American countries (Bengoa and Sanchez-Robles, 2003; Zhang, 2001; Tuman and Emmert, 2004).However little research has focused on outward FDI stemming from home grown Latin American Enterprises or the so-called Multilatinas. Multilatinas constitute the enterprises which have their origins in Latin American countries and have enjoyed home country advantages in order to expand their operations to foreign markets (Cuervo-Cazurra, 2008; Olaya et al. 2012; Casanova and Fraser, 2009).
Unlike other emerging market multinational enterprises (EM-MNEs), Multilatinas took a long time to establish foreign operations (Olaya et al, 2012; Cuervo-Cazurra, 2008). Over the years domestic reforms and regional reforms in the Latin American countries have led to a decrease of institutional constraints through economic liberalization and openness of the economies. These reforms have created new opportunities and first mover advantages for the Multilatinas, but to date the effect of the domestic and regional reforms have been measured by looking only at non-equity entry modes such as exporting (Nakata and Sivakumar, 1997; Lieberman and Montgomery, 1988; Pan and Tse, 2000). Some of the most applied models to measure these effects are the “Gravity Model” and the “Hecksher –Ohlin Model” (Carrère, 1997; Baier and Bergstrand, 2007; Frankel et al., 1995; MacDermott, 2007; Martinez-Zarsoso
and Nowak-Lehmann, 2003). Other examples include the effect of regional integration on the location advantages (Buckley et al. 2001), the effect of regional integration on R&D investment decisions (Cuervo-Cazurra and Un 2007) and the impact of trade agreements on the global orientation of EM-MNEs (Banalieva and Sarathy, 2010).
This thesis will further build on the notion of measuring the impact of trade agreements and examine equity entry modes instead of non-equity entry modes. So far no clear answer has been given in previous literature about the impact trade agreements on equity entry modes and in particular merger and acquistions. Considering the other type of equity entry modes, mergers and acquistions are the quickest way for Multilatinas to acquire strategic assets in host countries (Pan and Tse, 2000; Nadolska and Barkema, 2007). In this thesis a total number of 291 foreign M&As by 72 Multitlatinas will be studied through the focus on timing and location of internationalization. The timing and location of internationalization will be measured using the pace, scope and rhythm of the internationalization process (Vermeulen and Barkema, 2002). The research question of this paper is therefore:
What effect has the trade agreements had upon the internationalization pace, rhythm and scope of Multilatinas?
The answer to the research question is to find out if the regional trade agreements indeed had an influence on the pace, rhythm and scope of the internationalization process. This thesis will also look at bit further and look at what opportunities and threats the regional trade agreements have created. Due to the openness of the economies MNEs from other parts of the world are now able to compete with the Multilatinas in their own backyard (Khanna and Palepu, 1997). The advantages of most of these developed MNEs are ready access to international capital markets, topflight executive talent, powerful brands to leverage, and leading-edge technology.
For the Multilatinas it is necessary to overcome these competitive pressures by leapfrogging in the internationalization process (Chang and Rhee, 2011; Luo and Tung, 2007; Banalieva and Sarathy, 2010).
In order to analyze the internationalization process of the Multilatinas, after the trade agreements are concluded the first an introduction will be given to show what the developments and reforms have been taking place in the Latin American countries. Afterwards the internationalization process of the Multilatinas will be explained in how this process will be measured. In this section the hypotheses will be formulated in which the relationship between the trade agreements and the internationalization process will be tested. The hypotheses of these relationships will be tested using multiple linear regressions on 291 foreign mergers and acquisitions made by 72 Multilatinas. These 72 Multilatinas stem from six nations; Argentina, Brazil, Chile, Colombia, Mexico and Peru. To sum up the intended contribution of this paper is to analyze the impact of trade agreements on the internationalization process by Multilatinas. The internationalization process will be analyzed according to the number of foreign mergers and acquisitions (M&As) made by the Multilatinas before and after a trade agreement is concluded..
2. Theory and hypothesis development
2.1 Institutional and strategic constraints in Latin America
The institutional and strategic constraints Multilatinas face are the institutional voids such as limited home market potential, political instability, corruption and inadequate property protection (Khanna and Palepu, 1997). These institutional voids raise the transactions costs and decrease the trust of doing business in the Latin American countries. Due to the institutional constraints, strategic constraints and other reasons such as domestic import substitution policies, Multilatinas are now latecomers to internationalization process
(Banalieva and Sarathy, 2010; Cuervo-Cazurra, 2008). This latecomer status gives the Multilatinas significant competitive disadvantages compared to rival MNEs of developed nations.
In order to overcome these competitive disadvantages Multilatinas either need to leapfrog and develop intangible assets at home or acquire strategic assets elsewhere (Contractor et al. 2003; Chang and Rhee, 2011; Luo and Tung, 2007). It is indeed the case that in both of these scenarios, Multilatinas will be able to combine new non-location bound firm specific advantages NLB-FSAs with their location bound firm specific advantages at home (LB-FSA) to create new competitive advantages (Rugman and Verbeke, 2004; Dunning, 1997; Porter, 1990).
To reach these goals, Multilatinas do however have need to have ready access to necessary resources, adequate capital markets, information flows and home country specific advantages (Banalieva and Sarathy, 2010). However it is then again precisely the institutional constraints and strategic constraints in the Latin American countries, which prevent the creation of new competitive advantages (Banalieva and Sarathy, 2010; Chang and Rhee, 2011; figure 1: Trade Agreements, Facilitator of the Internationalization process?).
Institutional contstraints Strategic contstraints Trade Agreements First-Mover Advantages Pace Rhythm Scope Foreign M&As Competitive Pressures
Figure 1:Trade Agreements, Facilitator of internationalization process?
Over the years Latin American countries have been tackling institutional and strategic constraints by implementing domestic forms and regional reforms (Devlin and Ffrench-Davis, 1999; Dominguez and Brenes, 1997). These reforms have created new opportunities and first mover advantages for the Multilatinas, but to date the effect of the domestic and regional reforms have been measured by looking only at non-equity entry modes such as exporting (Nakata and Sivakumar, 1997; Contractor et al, 2003; Lieberman and Montgomery, 1988; Pan and Tse, 2000). Some of the most applied models to measure these effects are the “Gravity Model” and the “Hecksher –Ohlin Model” (Carrère, 1997; Baier and Bergstrand, 2007; Frankel et al., 1995; MacDermott, 2007; Martinez-Zarsoso and Nowak-Lehmann, 2003).
In this study the focus will lie on the effects of the regional forms on equity entry modes. It is indeed the case that trade agreements have been used to reduce trade barriers and facilitate the ease of trade, but the question remains, did it also have an effect on the Multilatinas’ internationalization process (Cuervo-Cazurra, 2008). In order to answer to this question, first past crucial events such as the domestic forms and regional reforms will be discussed to understand the current positions of the Multilatinas in their respective home countries (Cuervo-Cazurra, 2008; Lieberman and Montgomery, 1998).
2.2 Multilatinas and domestic reforms
Before the 1980s most of the Multilatinas were state-owned enterprises (SOEs) of home countries which pursued import substitution policies (Dominguez and Brenes, 1997). The import substitution policies are based on the Hamilton’s infant industry argument which encourages domestic manufactured goods over foreign imports (Peng, 2010; Braer, 1972). In order to implement these policies, the Latin American governments imposed steep tariffs and non-tariff barriers on imports and limited the number of foreign companies into the respective home markets (Dominguez and Brenes, 1997). The “state-owned” Multilatinas and other
domestic companies were given privileged positions to exploit resources and create monopolies (Peng, 2010; Braer, 1972). The downside of the import substitution policies was however the inability to create economies of scale, which led to limited growth and performance of the “state-owned” Multilatinas (Braer, 1972; Olaya et al., 2012).
Countries such as Brazil, Argentina and Peru sought to create synergies between the individual countries through trade agreements. The introduction of trade agreements would shift the Hamilton’s infant industry policies towards more Ricardian comparative advantage policies (Peng, 2010). According to theory of comparative advantage, nations enjoy relative advantage in one economic activity; in this case this would mean establishing complementarily agreements between nations by dividing labor between vertical lines, instead of trade in domestic finished goods (Peng, 2010; Braer, 1972). At the firm-level the trade agreements would lead to the creation of complementary firm specific advantages (FSAs) and the reduction of transaction costs between the host and home country (Rugman & Verbeke, 2007).
However instead of the trade agreements and regional integration, the focus was first on the domestic reforms. The implementation of the Washington Consensus reforms led to economic liberalization and opened the economies to foreign business in the respective domestic markets. This implied that major ‘state-owned’ Multilatinas were privatized and more foreign MNEs were allowed to establish operations in these countries (Kuczynski and Williamson, 2003). The Washington Consensus reforms also led to increased competition, acquisitions, joint-ventures and even bankruptcy (Aykut and Goldstein, 2006). The survivors of this period were forced to be leaner and meaner in order to bear the challenges of competition and internationalization (Aykut and Goldstein, 2006; Olaya et al. 2012). This period also marks the end of most of the ‘state-owned’Multilatinas and the birth of the ‘privatized’ Multilatinas (Cuervo-Cazurra, 2008)
2.3 Multilatinas and regional reforms
Before the 1980s the trade agreements, such as the Central American Common Market (1960), Latin American Free Trade Association (LAFTA) and the Andean Community (1969) had limited effect on trade flows, due to the focus of inward and protectionist import substitution policies. Apart from the import substitution polices, the Old Regionalism Era consisted of planned and political allocation of resources, border barriers and state-driven markets (Frankel et al, 1995; Eden, 2006; Mendoza et al., 2002).
By the mid-1980s the first regional reforms of the Washington Consensus were implemented, which also made way to the later known New Regionalism Era. The Washington Consensus reforms were in fact the transition period of the Old Regionalism towards the New Regionalism Era. The regional reforms of the Washington Consensus were this time focused on the ease of doing business in the region. The reduction of trade barriers and the economic liberalization of these regional reforms had a significant positive impact on the inter-regional trade (Braer, 1972; Devlin and Ffrench-Davis, 1999). The positive impact may have also been boosted by outward FDI of the newly privatized’ Multilatinas. Due to the domestic and regional reforms the Multilatinas are given an extra incentive in seeking new opportunities and resources in the region, but also need to deal with more competitive pressures in their home market (Aulakh et al., 2000; Cuervo-Cazurra, 2008; Mendoza, et al. 1999; Chang and Rhee, 2011).
In the beginning of the 1990s the “New Regionalism Era” was initiated with the signing of the Mercosur Customs Union (1991) between Brazil, Argentina, Uruguay and Paraguay. Contrary to the Old Regionalism, the New Regionalism Era was characterized by export orientation, market allocation of resources, deeper regional integration and equal rules for all nations (Mendoza et al, 1999). The start of the New Regionalism Era also led to a series of multilateral and bilateral agreements to be signed such as Central American Common Market
(1991), NAFTA (1994), Grupo de los Tres (1994) Andean Community (1996), ALBA (2006), UNASUR (2008) and more recently the Pacific Alliance (2012) (Eden, 2006; Peng, 2010; Devlin and Ffrench-Davis, 1999).
2.4 Trade Agreements
Trade agreements can be classified in terms of the degrees of economic integration, these are ; 1) free trade areas (FTA), 2) customs unions (CU), 3) common markets (CM), 4) economic union (ECU), 5) and total integration (TI)/ political union (PU) (Banalieva and Sarathy, 2010; Peng, 2010; figure 2: Degree of Economic Integration). The goal of each level of economic integration is to further reduce the trade barriers among its members, with a customs union implementing a common external trade policy towards non-members, a common market adding the free movement of goods, capital and labor across borders to the trade focus of the CU, the ECU adding the harmonization of economic policies to the CM, and TI/PU attempting to integrate the political and economic affairs among the trade agreements member countries (Banalieva and Sarathy, 2010; Peng, 2010). In Latin American countries there is a mix of free trade areas and customs unions (Ffrench-Davis, 1999; Mendoza et al. 1999).
Figure 2: Degree of Economic Integration (Peng, 2010)
The structure of the trade agreements can be further explained from a double-diamond perspective (Porter, 1990; Hodgetts, 1993). The Porter’s Diamond consists of the following elements; related and supporting industries, country factor endowments, domestic demand conditions and firm strategy, structure and rivalry (Porter, 1990). By taking a double diamond
analysis, the home market is extended to include other trade agreement member countries. This extension creates a larger combined market size with new opportunities for the Multilatinas (Hodgetts, 1993; Banalieva and Sarathy, 2010; Figure 3; Double Diamond Bilateral Trade Agreement Mexico-Colombia).
Domestic Demand Conditions Country Factor Endowments Mexican Government Bilateral Trade Agreement (MEX/COL) Related & Supported Industries Domestic Demand Conditions Country Factor Endowments Colombian Government Related & Supported Industries
Figure 3: Double-Diamond Bilateral Trade Agreement Mexico-Colombia (Hodgetts, 1993). Trade Agreement Tariffs
The macro-economic effects of the regional trade agreements can be further determined by the looking at the tariff structure (Mendoza et. al, 1999). The trade agreement tariff structure depends on the type of trade agreements signed by the member countries. For a free trade area (FTA) the tariff structure is based solely on increasing trade liberalization between trade agreement member countries. In a customs union, the tariff structure is transformed to include a common external tariff for members outside the trade agreement market area (Banalieva and Sarathy, 2010; Peng, 2010).
In Latin America there are two major customs unions Mercosur (1991) and the Andean Community (1995) (Figure 4: Average National Tariff Mercosur and Andean Community Member Countries). The addition of a common external tariff in a customs union leads the tariffs of individual member countries to fluctuate around it. (Mendoza et. al, 1999). Furthermore a reduction in tariffs between member countries also gives Multilatinas an extra incentive to consider location and timing of FDI (Dunning, 1998; Lieberman and Montgomery, 1988).
Tariff Liberalization in Mercosur Tariff Liberalization in Andean Community
Figure 4: Average National Tariff Mercosur and Andean Community Member Countries (Mendoza et. al, 1999)
Trade Agreement Experience
The number of years of membership in a trade agreement allows the Multilatinas to gain more experience and create learning-based advantages (Banalieva and Sarathy, 2010; Liebermann and Montgomery, 1988). A longer period of time within a trade agreement market area also allows the Multilatinas to gain trust in doing business, form alliances and develop NLB-FSAs (Banalieva and Sarathy, 2010; Rugman and Verbeke, 2009). In fact firms that have accumulated experience of trade agreements and develop internal resources and capabilities can benefit from earlier expansions into new host countries (Gaba et al., 2002; Chang and Rhee, 2011). The trade agreements may indeed facilitate the process of internationalization, but it does not automatically guarantee success and a high rate of
expansion abroad (Nadolska and Barkema, 2007). Therefore Multilatinas need to develop internal resources and capabilities to overcome the risks and uncertainties in the initial phase of market entry (Gaba et al., 2002; Chang and Rhee, 2011; Zaheer, 1995). The essential question in this situation is to ask how to deal with first mover and late mover advantages created by the trade agreements and how to respond in terms of timing and location of expansion (Nakata and Sivakumar, 1997; Hodgetts, 1993; Liebermann and Montgomery, 1988).
The decision to move quickly after a trade agreement is signed will give the Multilatinas first mover advantages; the downside is however it will be more likely to encounter risk and uncertainties due to the Liability of Newness (LON) and Liability of Foreignness (LOF) (Gaba et al., 2002; Chang and Rhee, 2011; Zaheer, 1995). The first mover advantages include pre-emption of the lower cost of necessary resources, development of buyer switching costs and competitive advantages (Contractor et al, 2003; Liebermann and Montgomery, 1988). Multilatinas can consider the option to focus first on gaining experience by acquiring domestic firms and afterwards internationalize (Nadolska and Barkema, 2007; Chang and Rhee, 2011). The value of knowledge and e.g. experiences of domestic acquistions are however bounded by time and location (Nadolska and Barkema, 2002). The decision to stay in their home market also gives the Multilatians the option to look at late mover advantages (LMAs) in foreign markets (Liebermann and Montgomery, 1988).
Following the OLI paradigm by Dunning (1997) the location advantages of trade agreements can be further enhanced by the reduction of tariffs and homogeneity of markets. Apart from the type of trade agreements, tariff structure, the ease of doing of business with other host trade agreement member countries may further depend on cultural, administrative, geographical and economic distance (Banalieva and Sarathy, 2010; Breinlich, 2008;
Nadolska and Barkema, 2007; Ghemawat, 2001). By taking all of these other factors into account, it is essential for the Multilatinas to develop a well-defined strategy to overcome obstacles in the internationalization process (Nadolska and Barkema, 2007; Gaba et al., 2002; Chang and Rhee, 2011; Zaheer, 1995). It is indeed a challenge for Multilatinas to decide where and when to internationalize despite of the time and location effects of internal resources and capabilities and the competitive pressures in their home market (Chang and Rhee, 2011). In order to capture the timing and location strategies of the Multilatinas, the pace, rhythm and scope of internationalization will be measured.
2.5 Pace, Rhythm and Scope of Internationalization
The pace, rhythm and scope of the internationalization process can be best described by the model of Vermeulen and Barkema (2002). The model of Vermeulen and Barkema (2002) analyzes the implications the speed and diversity of the expansion process on absorptive capacity. In order to capture and analyze the time and locations effects of the internationalization process the dependent variables pace, scope rhythm will be used (Nadolska and Barkema, 2007; Vermeulen and Barkema, 2002). The pace and rhythm of internationalization will measure the time effects of the internationalization process and the scope of internationalization will be measure the location effects of the trade agreements.
According to Vermeulen and Barkema (2002) the pace and scope of operating in multiple countries exceeds the absorptive capacity of the focal MNE and hence, negatively impacts firm performance. As late movers in the internationalization process, Multilatinas may benefit from the reduction of trade barriers and regional reforms (Luo and Tung, 2007; Matthews, 2006). Furthermore these Multilatinas also enjoy a couple of late mover advantages (LMAs) (Matthews, 2006; Peng, 2010). Late movers have the opportunity to free ride on first mover investments and the first mover’s difficulty to adapt to market changes (Lieberman and Montgomery, 1988). By capturing these late mover advantages the
Multilatinas are able to leverage their own expansion strategy based on prior international investments and benefit from the leapfrog effect (Luo and Tung, 2007; Nadolska and Barkema, 2007). By “leapfrogging” Multilatinas are able to internationalize at a higher pace (Matthews, 2006). However rapid expansion does have implications for corporate integration and the absorptive capacity (Vermeulen and Barkema, 2002; Dunning, 1987).
Pace of Internationalization
The pace and timing of internationalization might be determined by the different opportunities and constraints to enter new host countries. The relative homogeneity and a greater market size of a trade agreement may increase the ease of doing trade and therefore stimulate rapid expansion (Banalieva and Sarathy, 2010; Chang and Rhee, 2011). A higher pace of expansion will also allow Multilatinas to benefit from first mover advantages (FMAs) such as the pre-emption of the now lower cost of necessary resources and buyer switching costs (Contractor et al, 2003; Lieberman and Montgomery, 1988). The experience of being within a trade agreement may give the Multilatinas additional capabilities to overcome risk and uncertainties of the internationalization process Banalieva and Sarathy, 2010; Rugman and Verbeke, 2009).
Rhythm of internationalization
The greater the CAGE distance, the greater the challenges Multilatinas will face in pursuing rapid expansion and enjoy first mover advantages (Ghemawat, 2001; Lieberman and Montgomery, 1988). The different obstacles which the Multilatinas encounter and hinder the internationalization process is measured by the rhythm of the internationalization (Nadolska and Barkema, 2007; Gaba et al., 2002; Chang and Rhee, 2011; Zaheer, 1995).The rhythm of internationalization captures how evenly distributed or concentrated the foreign expansion of the Multilatinas are in terms of timing (Vermeulen and Barkema, 2002).
Figure 5: Rhythm of internationalization
In a trade agreement of countries which are characterized as to have a large diversity and high tariffs, Multilatinas may have an irregular rhythm of internationalization (Banalieva and Sarathy, 2010). Due to the different challenges and constraints, it will be mostly likely for Multilatinas to make good and bad investments. As a result the rhythm of internationalization will show higher peaks and valleys of expansion (Vermeulen and Barkema, 2002).
Scope of Internationalization
The scope of internationalization is measured through the number of countries of internationalization (Vermeulen and Barkema, 2002). The benefits of a large scope of internationalization is risk diversification across different countries and the enhancement of the knowledge base, capabilities and competitiveness trough experiential learning (Lu and Beamish, 2004; Chang and Rhee, 2011, Hitt et al, 1994). Vermeulen and Barkema (2002) argued that to having operations in multiple countries in a short period of time do however exceeds the absorptive capacity and corporate integration of the firm.
2.6 Linking Trade Agreements with Pace, Rhythm and Scope of Internationalization
The number of years of membership in a trade agreement allows the Multilatinas to gain more experience and create learning-based advantages (Banalieva and Sarathy, 2010; Liebermann and Montgomery, 1988). A longer period of time within a trade agreement market area also allows the Multilatinas to gain trust in doing business, form alliances and
develop NLB-FSAs (Banalieva and Sarathy, 2010; Rugman and Verbeke, 2009).On the other hand if Multilatinas act quickly it will be able benefit from first mover advantages, such as the pre-emption of the now lower cost of necessary resources and buyer switching costs (Nakata and Sivakumar, 1997; Lieberman and Montgomery, 1988; Contractor et al, 2003).
The crucial strategic decision for the Multilatinas is to decide where and when to internationalize after a trade agreement is signed. If Multilatinas decide not to enter certain host countries quickly it may able to late mover advantages (LMAs) in later point in time and benefit from the “leapfrog effect” ((Luo and Tung, 2007; Matthews, 2006). Then again these Multilatinas do however need to consider the fact that competitive pressures will continue in their home market and can ultimately become a major threat for their operations (Khanna and Palepu, 1997; Chang and Rhee, 2011).
Trade Agreements and Pace of Internationalization
According to Chang and Rhee (2011) the effectiveness of rapid FDI expansion depends on the competitive pressures a focal firm faces. After a trade agreement is concluded, Multilatinas do not only have the opportunities to expand in to host countries, but also have to deal competitive with a more competitive environment in their home market. This makes the pace of internationalization of utmost importance to determine how fast the necessary resources acquired in host markets.
From a resource-based view perspective Bertrand and Zitouna (2006) note that trade agreements affects foreign M&A incentives depending on the degree of knowledge gaps and lower trade costs: If knowledge gaps are high and trade costs are low there is inverted U-shaped relation between trade costs and incentives to merge.
This proves again the notion that trade agreements do not only increase the competitive pressures at home but also raise incentives to merge and acquire foreign companies. This is
because of higher access to lower cost of resources and first mover advantages created by the trade agreements (Contractor, 2003: Lieberman and Montgomery, 1998). Bertrand and Zitouna (2006) and Contractor et al, 2003 also note that degree of internationalization follow an inverted U-shaped relation if access to resources are low, this stage of internationalization is precisely what is provided by the trade agreements.
The trade agreements also allow the Multilatinas to leapfrog the first stage of internationalization and expand more rapidly and capture first mover advantages (see Contractor et al, 2003; Lieberman and Montgomery. 1988; Hitt et al, 1994). On the other hand late-movers are confronted with market saturation characterized by lower incentives to merge and higher cost of resources (Bertrand and Zitouna, 2006; Contractor et al, 2003; Lieberman and Montgomery, 1988). This provides Multilatinas who act quickly after a regional trade agreement is concluded to expand more rapidly until market saturation is reached.
∗ Hypothesis 1a: Trade agreement experience is negatively associated with the pace of internationalization
To further facilitate the access and incentives to acquire the necessary resources and first-mover advantages, the tariffs and trade costs between the trade agreement member countries should be low and trade barriers should be lifted (Bernard & Zitouna, 2006; Breinlich, 2008; Liebermann and Montgomery; 1988). However if the tariffs remain high after a trade agreement is concluded, the domestic market of the Multilatinas will remain protected and the opportunities in other trade agreement member countries will be scarce. In this case high tariffs will slow down the pace of internationalization and constrain the access necessary resources and first-mover advantages (Breinlich, 2008; Liebermann and Montgomery, 1988).
We therefore propose a relationship in which a reduction of trade agreement tariffs allows Multilatinas to expand more rapidly into host countries.
∗ Hypothesis 1b: Trade agreement tariffs are negatively associated with the pace of internationalization
Trade Agreements and Rhythm of Internationalization
Despite of the competitive pressure at home Chang and Rhee (2011) show that Multilatinas which possess heterogeneous and strong non-location bound firm specific advantages (NLB-FSAs) have the opportunity to expand with a regular rhythm of internationalization (Rugman and Verbeke, 2004). These capabilities also aid in overcoming the challenges posed by Liability of Foreignness (LOF) and Liability of Newness (LON) (Gaba et al., 2002; Chang and Rhee, 2011; Zaheer, 1995).
The rhythm of internationalization measures how evenly distributed the foreign expansions are in terms of timing. As a result the rhythm of internationalization shows how well the Multilatinas have been dealing with the different obstacles of foreign expansion. When the argument of the first mover advantages and knowledge gaps is applied to the rhythm of internationalization process, the initial phase of internationalization can be best characterized by a regular rhythm of internationalization with ready access to pre-empt the necessary resources and the establishment of high buyer switching costs (Nadolska and Barkema, 2002; Lieberman and Montgomery, 1988; Vermeulen and Barkema, 2002). After valuable resources and first-mover advantages are captured, the Multilatinas enter a saturation period characterized by high number of takeovers and long periods of inactivity.
In this period Multilatinas do not only capture late mover advantages in selected markets but also deal with higher coordination costs of corporate integration (Vermeulen and Barkema,
2002; Dunning, 1987). These factors ultimately lead to an irregular rhythm of internationalization (Lieberman and Montgomery, 1988; Vermeulen and Barkema, 2002). ∗ Hypothesis 2a: Trade agreement experience is positively associated with the rhythm of
internationalization
When trade costs are low and trade barriers are lifted, the access to first-mover advantages and incentives to merge are facilitated (Bernard & Zitouna, 2006; Breinlich, 2008; Liebermann and Montgomery; 1988). As mentioned previously the reduction of trade barriers and tariffs also allows Multilatinas to focus more on developing, acquiring and integrating new firm specific advantages (Rugman and Verbeke, 2004). Furthermore this situation also allow Multilatinas to develop a better strategy in deciding when and where to internationalize. On the other hand high trade agreement tariffs does not only constrain the Multilatinas in gaining access to first-mover advantages but are also makes it more likely to make bad investments (Vermeulen and Barkenma, 2002). These bad investments are more likely to be made, due the competitive pressures in the home market and the pressures to internationalize after a trade agreement is concluded (Banalieva and Sarathy, 2010; Chang and Rhee, 2011). As a result the rhythm of internationalization will be more irregular with higher peaks and lower valleys (Vermeulen and Barkema, 2002). We therefore test the following relationship between trade agreement tariffs and the rhythm of internationalization. ∗ Hypothesis 2b: Trade agreement tariffs is positively associated with the rhythm of
internationalization
Trade Agreements and Scope of Internationalization
According to Chang and Rhee (2011) a rapid pace of expansion is an effective strategy for Multilatinas facing competitive pressure to internationalize. Chang and Rhee (2011) also note that Multilatinas in this situation should also focus on international diversification and a high
scope of internationalization rather than to expand gradually into new markets (Johanson and Vahlne, 1977).
Taking an institutional-based view (IBV) a higher scope of internationalization with a lower number of years of trade agreement experience will mostly likely constrain Multilatinas in adapting to a diverse formal and informal institution of the host countries (Peng; 2010; Banalieva and Sarathy, 2010; North, 1990; Lu and Beamish, 2004; Zaheer, 1995; Lieberman and Montgomery, 1988). This is because formal and informal institutions take a while to change, after a trade agreement is concluded (North, 1990; Williamson, 2000; Banalieva and Sarathy, 2010).
From a resource-based view the creation of new opportunities by the trade agreements allow for the Multilatinas to move from early internationalizers to mid-stage internationalizers. These new opportunities are characterized by ready access to low cost resources and first-mover advantages (Contractor et al. 2003; Lu and Beamish, 2004; Lieberman and Montgomery, 1988). This effect does not only allow Multilatinas to leapfrog in the internationalization process with a rapid pace and steady rhythm of internationalization but also allow the Multilatinas to diversify its internationalization process in a large scope of countries.
∗ Hypothesis 3a: Trade agreement experience is negatively associated with the scope of internationalization
The scope of internationalization may also depend on the number of countries within a trade agreement. As mentioned by Chang and Rhee (2011) the proposed strategy to counteract competitive pressure in the domestic market is to focus on international diversification. International diversification may be the best option, but Multilatinas also need to take trade
cost into account when making the decision to expand in a higher scope of countries, especially if these countries are not be part of a trade agreement. As mentioned previously the trade agreements may create new opportunities in trade agreement member countries, but if trade agreements tariffs remain high, Multilatinas will be less likely to access the necessary resources and first mover advantages (Breinlich 2008; Lieberman and Montgomery, 1988). We therefore argue that the trade agreement tariffs need to be reduced and become more homogenous for Multilatinas to access the necessary resources in a higher scope of countries. ∗ Hypothesis 3b: Trade agreement tariffs are negatively associated with the scope of
internationalization 2.7 Conceptual Framework
Figure 6 depicts the relationship between the independent and dependent variables. As mentioned before, this thesis will analyze the timing and location effects of internationalization after a trade agreement is signed. In order to establish the right strategic move of internationalization, several elements may play a role in this decision. For instance the decision to move to certain host country after a trade agreement is signed between countries maybe further influenced by diversity of formal/informal institutions and access to resources (Banalieva and Sarathy, 2010; North, 1990; Contractor et al. 2003). Other elements which may influence the strategic decision in the timing and location of internationalization may be common borders, common language and cultural distance (Banalieva and Sarathy, 2010; Breinlich, 2008). All of these elements will be used as control variables, in order to capture other aspects which might influence the pace, rhythm and scope of internationalization. Furthermore the independent variable RTA experience is linked to the dependent variables pace, rhythm and scope of internationalization by the number of foreign mergers and acquisitions over time. For each of the number of foreign mergers and
acquisitions the following control variables are used to check for the influence on the strategy of timing and location of internationalization.
RTA Experience (Independent) RTA Pace (Dependent) RTA Rhtyhm (Dependent) RTA Scope (Dependent) H1a: (-) H2a: (+) H3a: (-) Control Firm Effects • Firm Age
• Experience with Domestic Acquisitions
Trade Agreement Effects • RTA Type • RTA Diversity Country Effects • Cultural Distance • Border Effects • Language Effects RTA Tariffs (Independent) H3b: (-) H1b: (-) H2b: (+)
Figure 6: Independent, Dependent, Control Variables and Hypotheses
3. Method
3.1 Sample
The hypotheses are tested on a sample of 72 Multilatinas, originating from six Latin American countries; Argentina, Brazil, Chile, Colombia, Mexico and Peru. Panel data was gathered from Thomson One database on the number of mergers and acquisitions covering a period of 27 years (1985-2012).The firms in this sample undertook 932 mergers and acquisitions, of which 32 percent were foreign acquisitions (figure 7: Regional Trade Agreements). These 291 foreign acquisitions were used to measure the dependent variables. In Latin America the major trade agreements are either custom unions or regional trade agreements. The major custom unions are Mercosur (1991) and Andean Community (1996) (Peng, 2010). Two of the six nations are not part of these custom unions; these are Mexico and Chile. For these countries the RTA experience is measured using the first free trade
agreements of these countries. Coincidentally the first regional trade agreement of these countries is a free trade agreement between the respective countries itself (FTA Mexico/ Chile) (1992) (figure 7: Regional Trade Agreements).
Regional Trade Agreements (RTA)
Year of entry
Mercado Común del Sur (Mercosur)
Argentina 1991
Brazil 1991
Bilateral FTA (Mexico-Chile)
Mexico 1992
Chile 1992
Comunidad Andina (ANDEAN Community)
Colombia 1996
Peru 1996
Figure 7: Regional Trade Agreements (http://www.worldtradelaw.net/fta/ftadatabase/ftas.asp)
3.2 Dependent Variables
Pan and Tse (2000) describe equity-based entry modes to be composed of joint-ventures, mergers and acquistions or greenfield investments. The main advantage of mergers and acquisitions for Multilatinas is to quickly acquire essential resources in the domestic market as well as in foreign markets (Nadolska and Barkema, 2007). The makes the analysis of mergers and acquisitions an excellent choice to find out how quickly the impact of trade agreements have on the Multilatinas. The number of foreign mergers and acquisition of by the Multilatinas are analyzed over a period of 27 years (1985-2012). After the Mercosur (1991) was signed a total of 931 mergers and acquisitions were recorded. Trade Agreements have an effect on foreign as well as domestic acquisitions (Breinlich, 2008; Nadolska and Barkema, 2007). In this analysis the total number of mergers and acquistions were used to measure the pace, rhythm and scope of internationalization, because experience with domestic M&As can have an effect on pace of foreign M&As (Nadolska and Barkema, 2007).
Pace of internationalization
The pace of acquisitions indicates how many firms the Multilatinas have acquired in a certain period of time. To measure the pace of internationalization, the average number of acquisitions per year is computed after a trade agreement is signed (Vermeulen and Barkema, 2002). For the six nations of this sample this means a different period in time, in which the first trade agreement is signed (figure 7: Regional Trade Agreements).
Rhythm of internationalization
The rhythm of internationalization process was measured through the kurtosis of the first derivative of the number of mergers and acquisitions of the Multilatinas over time. This variable measures how concentrated in time the change in the number of acquisitions is (Vermeulen and Barkema, 2002).The concentration and changes in time are measured by the kurtosis: Kurtosis = ) 3 )( 2 ( ) 1 ( 3 ) ( ) 3 )( 2 )( 1 ( ) 1 ( 4 2 − − − − − − − − +
∑
n n n s x i x n n n n nWhere n= number of observations, x(i) = number of acquistions in year i, and s= standard deviation of the number of acquisitions (Vermeulen and Barkema, 2002). Large peaks in a firm’s expansion patterns, combined with periods of inactivity, result in relatively concentrated distribution and therefore a high kurtosis.
Scope of internationalization
The scope of internationalization is measured by looking at the number of countries in which Multilatinas acquire other firms. The scope of internationalization is used to measure the geographic diversity of the internationalization process (Vermeulen and Barkema, 2002). In this sample the Multilatinas had an average geographic diversity of 1.93 countries. This limited number of countries is largely caused by the focus on of our study on only Latin American countries and the large focus of the Multilatinas on the domestic market.
Figure 8: Dependent Variables
3.3 Independent Variables
RTA Experience
The RTA experience is measured looking at the absolute number of years the acquirer nation is part of the RTA (Banalieva and Sarathy, 2010). The number of years a country or acquirer nation is part of the RTA is computed looking at the establishment of major regional trade agreements (figure 7: Regional Trade Agreements). For Argentinean and Brazilian Multilatinas, this means that the RTA experience is measured from (1991-2012). For Mexico and Chile this is (1992-2012) and for Colombian and Peruvian Multilatinas, the number of years in a regional trade agreement is measured from 1996 till 2012.
For the Multilatinas founded after one of the regional trade agreements is established, the RTA experience is equal to the firm age. This means, if a Multilatina is founded in 2000 and the regional trade agreement is signed in 1996, the RTA experience of this particular Multilatina is measured between 2000 and 2012. In addition to the former aspect, another aspect which is kept into account is firm survival. The firm survival is analyzed if a company is acquired or ceases to exist before 2012 the period of RTA experience ends at that particular point in time. To capture the effect of the RTA experience the control variable firm age will be included to look at the overall experience of the Multilatinas before the trade agreements are established.
RTA Tariffs
The RTA Tariffs are measured by looking at the MFN tariffs of the individual countries participating in either free trade agreements (FTA) or customs union (World Bank, 2008). A higher MFN tariffs indicates greater import substitution polices to protect domestic markets (Banalieva and Sarathy, 2010). The downward trend in the graph of the MFN tariffs show a
Variables Definition
Internationalization pace The average number of acquisitions per year
Internationalization scope The number of countries of acquisitions in a given year Internationalization rhythm The kurtosis of the number of acquisitions of the firm over
time
continuous progress in allowing new foreign entrants into the domestic markets and new opportunities in host countries (figure 9: MFN Tariffs).
Figure 9: (Source: http://data.worldbank.org/indicator/TM.TAX.MRCH.SM.FN.ZS/countries?display=default)
The RTA tariffs are also shifting towards a constant number, this indicates that the tariffs within the Latin American countries are becoming more homogeneous (Banalieva and Sarathy, 2010). This creates a more integrated market for doing business, which decreases transaction costs. If we take a look at tariffs of Peru, it can be concluded that it had the second-highest tariff after the New Regionalism Era was initiated. Nowadays it has the lowest tariff of all. This trend does not however hold for the common external tariff (CET) of the entire Andean Community customs union. In this example the Colombian and Peruvian MFN tariffs do show a high difference between each other. As for the Mercosur member countries, Brazil and Argentina the most favorite nation (MFN) tariffs of the respective countries do seem to fluctuate around a common external tariff (CET).
3.4 Control Variables
Firm Age
The firm age control variable captures the overall experience of the Multilatinas in the doing business either abroad or in the domestic market. The firm age of the Multilatinas also gives an overall indication of the first time of internationalization. Furthermore this control variable
3 8 13 18 23 28 1991 1996 2001 2006 2011 Argentina Brazil Chile Colombia Mexico Peru Page 29 of 60
also gives an indication of the number of acquisitions in the pre-RTA period and how this has changed in the post-RTA period. For the Multilatinas established after a trade agreement is established the control variable Firm age is equal to the RTA experience.
Experience with Domestic Acquisitions
The experience of domestic acquisitions is measured by looking at the number of domestic acquisitions by Multilatinas over the past 27 years (1985-2012). This time-frame is chosen to analyze if the experience of domestic acquisitions do have a significant effect on the pace, rhythm and scope of internationalization without of the trade agreements. It is indeed the case that the majority of acquisitions in this sample are focused on the domestic market of the Multlatinas, but this does not automatically mean that the domestic acquisitions have an impact on foreign acquisitions. The experience of domestic acquisitions can be a helpful mechanism to understand and learn from the acquisition process (Nadolska and Barkema, 2007).
RTA Type
Trade agreements can be classified between the degrees of economic integration extending from; from 1) free trade areas (FTA), to customs unions (CU), common markets (CM), economic union (ECU), and total integration (TI)/ political union (PU) (Banalieva and Sarathy, 2010; Peng, 2010; figure 3: Degree of Economic Integration). In this sample there are two types of trade agreements customs unions and free trade agreements. Argentina, Brazil, Colombia and Peru are part of a customs union. Mexico and Chile are part of a free trade agreement. In order to quantify the RTA type, a dummy variable was created according to the degree of economic integration. Therefore member countries of a free trade agreement were set as number 1 and member countries of a customs union were set as number 2.
Border Effects
In the sample of the six nations Mexico is the only country that does not share with one of the other five nations in this sample. Brazil shares the most borders with the other five nations,
but still is more focused domestic market instead of cross-border acquisitions. In order to measure the border effects on the foreign M&As in this sample, a dummy variable was created. As a result a common border between acquirer and target nation is set as number 1 and 0 otherwise. Despite of the lack of common borders with the other five nations, Mexico does however share a common language with four of the other nations in this sample. In this case Brazil shares the most borders with other nations, but lacks the common language. Therefore a dummy variable is created to control language effects.
Language Effects
Most of the countries in Latin America are Hispanophone countries except for countries such as Suriname, Guyana, French Guyana, Belize or Brazil, which are either Dutch-speaking, Anglophone, Francophone or Lusophone countries. In this sample the majority of the Multilatinas stem from Hispanophone countries except for Brazil which is a Lusophone country. In order to control the language effects on the internationalization process; a dummy variable is created and set as number 1 for common language and 0 otherwise.
RTA Diversity
The regulatory environment of the host countries is a crucial element of defining the risks and uncertainties of foreign acquisitions. In order to measure the regulatory environment or institutional distance in this sample the Economic Freedom scores are taken of the six nations from 1985 - 2012. The Economic Freedom scores captures the various aspects of the institutional environment in these countries such as business, trade, monetary, government, fiscal, property rights, investment, financial, labor and corruption (Banalieva and Sarathy, 2010; Figure 10: Economic Freedom Scores).
Figure 10: Economic Freedom scores (Source: http://www.freetheworld.com/datasets_efw.html)
As shown in graph above the Economic Freedom scores of the Latin American countries in this sample were all below 6.5 before the major regional trade agreements were concluded except for Chile. Afterwards the New Regionalism Era was initiated the Economic Freedom scores fluctuated around 6.5. These scores give another positive indication of the focus improvement, coherence and stability after the regional trade agreements are concluded. The effect of the trade agreements on the Economic Freedom scores can also be seen in the remarkable positive trend of the Peruvian scores.
Cultural Distance
The Hofstede (1980) scores on cultural distance also have an effect on the ease of doing business in new host countries. To compute the differences between home country and host countries within this group, each of the six nations was set as the home country. In addition to the six nations of this sample other host countries such as Ecuador, Uruguay and Venezuela were included. The countries are added to this figure because of their membership in one of the customs unions Mercosur or Andean Community. The comparison of the six nations plus other member countries of the Mercosur (1991) and Andean Community (1995) show a relative coherence in cultural distance scores.
2,5 3,5 4,5 5,5 6,5 7,5 8,5 1985 1990 1995 2000 2005 2010 Peru Argentina Brazil Chile Colombia Mexico Page 32 of 60
Home/Host PDI UAI INV MAS Argentina 49 86 46 56 Brazil 69 76 38 49 Chile 63 86 23 28 Colombia 67 80 13 64 Ecuador 78 67 8 63 Mexico 81 82 30 69 Uruguay 61 100 36 38 Venezuela 81 76 12 73
Figure 11: Cultural distance scores
For Paraguay and Bolivia the Hofstede scores (1980) are unknown, therefore estimates were calculated with the aid of the scores of Uruguay and Ecuador. The scores of the individual countries were applied to each of the foreign acquisitions in the respective countries.
Home CD(Host) CD(Host) CD(Host) CD(Host) CD(Host) CD(Host) Argentina 0 1.3325917 2.0121162 2.4118474 2.8815686 1.9659150 Brazil 1.3325917 0 1.1579328 1.1633959 0.9454355 1.1348838 Chile 2.0121162 1.1579328 0 1.6661876 2.6399233 0.2827413 Colombia 2.4118474 1.1633959 1.6661876 0 0.8793148 0.6940614 Mexico 2.8815686 0.9454355 2.6399233 0.8793148 0 1.7735829 Peru 1.9659150 1.1348838 0.2827413 0.8033066 1.7735829 0 Ecuador 5.0025174 1.8773994 3.1865261 0.6940614 1.3902468 2.1741627 Uruguay 1.3836066 1.9688720 0.9254705 2.7112155 2.9263617 1.0880241 Venezuela 4.5046568 1.8768289 3.3541503 0.5749690 0.5721268 2.0537370 Average score 2.6868525 1.2730378 1.6916720 1.2115887 1.5199708 0.9751974
Figure 12: Comparison between host and home country
The cultural distance scores between the home country and host countries were measured by taking each of the six nations in this sample as a home country. The average scores show that Argentina (2.68) has the highest scores on cultural difference and Peru (0.975) has the lowest cultural difference. The cultural difference between Peru and Chile (0.28) is the lowest individual score between the countries. The scores on cultural difference do not automatically give an indication of the most favorite target nations or countries which are most likely to engage in signing trade agreements between each other. In fact in the case of Chile and Peru it took a total of 10 years after the Andean Community (1996) was signed for Peru and Chile to sign their first free trade agreement between each other. Furthermore the scores on cultural distance do not fluctuate over the years (Williamson, 2000; Shenkar, 2000). This makes it
possible to assign each cultural distance score to the foreign acquisition without taking the year into account.
3.5 Data Analysis
The three hypotheses of this thesis will be tested using regression analysis on the three dependent variables. Each of the three dependent variables will be tested, by looking at the relationship between each individual dependent variable and the independent variable RTA experience. The optimal goal of the regression analysis is to find the best fitting straight line in this particular relationship.
In order to test the conceptual framework depicted in figure 9, three regression analysis will be tested on the dependent variables pace, rhythm and scope of internationalization. In Model 1 only the control variables are tested on the each of the dependent variables. The control variables are depicted in figure 13.
Figure 13: Control Variables
Subsequently in model 2 the first independent variable is added to test the relationship between RTA Experience and each of the three dependent variables pace, rhythm and scope of internationalization. In model 3 the second independent variable RTA Tariffs is added and in the last model the results are given for the control variables as well as both of the
independent variables RTA Tariffs and RTA Experience (figure 14; Stepwise regression of the dependent variables).
Control Variables Definition
Experience with Domestic Acquisitions The number of Domestic acquisitions Border Effects Countries share the same border Language Effects Countries share the same language
RTA Type Customs Unions or Free Trade Agreement
RTA Diversity Economic Freedom Score
Cultural Distance Kogut-Singh Formula between home & host country Firm Age Number of years since the firm is founded
Stepwise regression on Pace, Rhythm and Scope of Internationalization Model Control variables RTA Experience RTA Tariffs Model 1 X
Model 2 X X
Model 3 X X
Model 4 X X X
Figure 14: Stepwise regression of the dependent variables
4. Results
4.1 Descriptive Statistics
The descriptive statistics provide the mean, standard deviation and range of the dependent independent variables and control variables (figure 15; Descriptive Statistics). The pace of internationalization varied between a minimum of 0.0588 foreign acquisitions per year towards 2.6667 acquisitions per year. The rhythm fluctuated between -2.1094 and 14.1324. The low pace combined with high rhythm of internationalization clearly shows an irregular rhythm, characterized by long periods of inactivity and relatively high number of acquisitions in certain periods of time.
Descriptive Statistics
Variables N Minimum Maximum Mean Std. Deviation
Pace 72 0.0588 2.6667 0.7547 0.5949 Rhythm 72 -2.1094 14.1324 2.4500 3.6710 Scope 72 0.0000 10.0000 1.9306 2.1580 RTA Experience 72 4.0000 21.0000 15.4167 4.2352 RTA Tariffs 72 4.8500 25.8400 12.3738 3.7394 RTA Diversity 72 3.9719 7.9814 6.3890 0.9426 Cultural Distance 72 0.0000 2.9441 0.9355 0.7626 RTA Type 72 1.0000 2.0000 1.6806 0.4695 Border Effects 72 0.0000 1.0000 0.3508 0.4133 Language Effects 72 0.0000 1.0000 0.4669 0.4656 Domestic Acquistions 72 0.0000 25.0000 4.0694 5.2844 Firm Age 72 4.0000 130.0000 65.4583 36.6852
Figure 15: Descriptive Statistics
The scope of internationalization varies between 0 and 10 different countries of foreign acquisitions. The scope of the host countries measured in this sample was only focused on
countries in Latin America. Host countries which are not a member of any trade agreements are also included; such as Costa Rica, Guatemala, Honduras, Nicaragua and Panama. In this regard firms of these countries were mostly acquired by either Mexican Multilatinas are Colombian Multilatinas. An explanation for the latter can be the relative proximity between the home and these host countries.
The independent variable RTA experience varied between 4 and 21 years. The RTA experience is measured according to the number of years the Multilatinas’ home country is part of a trade agreement. For Multilatinas founded after a trade agreement is signed, the RTA experience is equal to the Firm Age. This means, if a Multilatina is founded in 2000 and the regional trade agreement is signed in 1996, the RTA experience of this particular Multilatina is measured between 2000 and 2012. In addition to the former aspect, another aspect which is kept into account is firm survival. The firm survival is analyzed if a company is acquired or ceases to exist before 2012. In this case RTA experience starts with ends at that particular point in time. The Multilatinas with the minimum amount years of RTA experience was established after a trade agreement was established and acquired four years after it was founded. The maximum amount of years of RTA experience account for both the Argentinean and Brazilian Multilatinas which are established before the Mercosur (1991) was signed (figure 16: Regional Trade Agreements).
Regional Trade Agreements (RTA)
Year of entry
Mercado Común del Sur (Mercosur)
Argentina 1991
Brazil 1991
Bilateral FTA (Mexico-Chile)
Mexico 1992
Chile 1992
Comunidad Andina (ANDEAN Community)
Colombia 1996
Peru 1996
Figure 16: Regional Trade Agreements (http://www.worldtradelaw.net/fta/ftadatabase/ftas.asp)
The control variables of this sample start with the RTA Diversity and end with the Firm age. Page 36 of 60
The range of the cultural distance scores varies between 0 and 2.9441. The zero score on cultural distance was used to define the domestic acquisitions of this sample. The domestic acquisitions do not need have cultural distance scored due to the same home and host country. In this case of the foreign acquisitions, the home and host country were checked and scores were added according to the comparison scores in figure 9. For each Multilatina the average scores of its foreign acquisition was computed to get the final score for each firm. The maximum score of 2.9441 gives an indication of the Multilatina with the highest scores on foreign M&As. The control variables firm age has the highest mean (65.4583) and variance (36.6852) of all. The range of this variable is between 4 and 130 years.
4.2 Correlation Analysis
The correlation matrix shows a strong positive relation between the dependent variables pace and the scope of internationalization (0.863). This number show how related the speed and geographic diversity are in the internationalization process. The strong correlation between these two dependent variables can be explained due to the ready access to low cost of resources and first mover advantages in large scope of countries (Contractor et al., 2003; Liebermann and Montgomery, 1988; Hitt et al, 1994).
Due to these factors Multilatinas do not only want to expand rapidly but also diversify the internationalization process into a higher scope of countries. Furthermore the control variables experience of domestic acquisitions (0.537) and cultural distance (0.420) also both show a strong positive correlation on the pace of internationalization. Taking both the RTA Diversity as well as the RTA Tariffs control variables, it can be concluded that both of control variables do not have correlations with any of the other variables. The control variables are also correlated; in this case the cultural distance and the experience of domestic acquisitions have a strong correlation. The Cronbach’s Alpha for the 12 items in this study is
0.636. By taking a look at the dependent variables the correlation matrix shows a strong correlation between the pace and scope of internationalization (0.863).
Master Thesis G.O. Perez (10506535) Reliability Statistics Cronbach’s Alpha Cronbach’s Alpha on Std. Items 0.063 0.636
Summary Statistics & Correlation Matrix
Variables Mean Std. Deviation N 1 2 3 4 5 6 7 8 9 10 11
Pace 0.7547 0.5949 72 Rhythm 2.4500 3.6710 72 -.098 Scope 1.9306 2.1580 72 .863** -.003 RTA Experience 15.4167 4.2352 72 .029 .207 .140 RTA Tariffs 12.3738 3.7394 72 .161 .029 .121 -.195 RTA Diversity 6.3890 0.9426 72 .144 -.003 .124 .204 -.042 Cultural Distance 0.9355 0.7626 72 .420** .470** .508** .233* .151 .176 RTA Type 1.6806 0.4695 72 -.191 -.210 -.231 .302* -.087 -.245* -.276* Border Effects 0.3508 0.4133 72 .224 .230 .243* .238* .140 .032 .369** .337** Language Effects 0.4669 0.4656 72 .221 .480** .372** .045 .144 .001 .539** -.224 .234* Domestic Acquisitions 4.0694 5.2844 72 .537** -.107 .485** .356** .049 .193 .255* .061 .295* -.206 Firm Age 65.4583 36.6852 72 -.073 .007 .057 .246* -.313** .024 .033 .027 .122 .138 -.118 **. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
4.3 Collinearity Statistics
The Collinearity Statistics was applied for each of the dependent variables; pace, rhythm and scope of internationalization. The following tables show equal tolerance and VIF scores for each dependent variable. In order to rule out the presence of multicollinearity between the independent variables the scores on Tolerance and Variance Inflation Factor (VIF) scores are checked. VIF scores show values below five and Tolerance score of above .20. This means there are no significant correlations among the independent variables.
Collinearity Statistics of the Pace of Internationalization
Std. Coefficients t
Collinearity Statistics
Beta Tolerance VIF
RTA Experience -0.240 -2.012** 0.600 1.668 RTA Tariffs 0.034 0.334 0.829 1.206 RTA Diversity 0.033 0.326 0.642 1.229 Cultural Distance 0.153 1.133 0.467 2.140 RTA Type -0.036 -0.279 0.524 1.908 Border Effects -0.027 -0.220 0.568 1.760 Language Effects 0.271 2.161** 0.543 1.840 Domestic Acquisitions 0.638 5.569*** 0.618 1.617 Firm Age 0.034 0.327 0.780 1.282
Dependent Variable: Pace
Sign at 0.1 level*; Sign at 0.05 level**; Sign at 0.01 level***; Sign at 0.001 level****
Furthermore the table above already indicates which independent and control variables have a significant effect on the dependent variable. In this case the independent and control variables which have an effect on the pace of internationalization are RTA Experience, Language Effects and Domestic Acquisitions. The range of the Tolerance values is between 0.467 and 0.780. In the collinearity statistics a higher value in Tolerance means a lower value in VIF. In this case the Cultural Distance control variable has the lowest Tolerance value (0.467) and the highest VIF value (2.140). The firm age control variable has the opposite the highest Tolerance value (0.780) and the lowest VIF value (1.282).
Collinearity Statistics of the Rhythm of Internationalization
Std. Coefficients
t
Collinearity Statistics
Beta Tolerance VIF
RTA Experience 0.346 2.771*** 0.600 1.668 RTA Tariffs -0.140 -1.318 0.829 1.206 RTA Diversity -0.078 -0.732 0.813 1.229 Cultural Distance 0.226 1.601 0.467 2.140 RTA Type -0.312 -2.336** 0.524 1.908 Border Effects 0.250 1.952* 0.568 1.760 Language Effects 0.190 1.451 0.543 1.840 Domestic Acquisitions -0.295 -2.400** 0.618 1.617 Firm Age -0.190 -1.737* 0.780 1.282
Dependent Variable: Rhythm
Sign at 0.1 level*; Sign at 0.05 level**; Sign at 0.01 level***; Sign at 0.001 level****
For the second dependent variable which is the rhythm of internationalization the following independent variables and control variables have a significant effect; RTA Experience, Cultural distance, RTA Type and Domestic acquisitions.
Collinearity Statistics of the Scope of Internationalization
Std. Coefficients t
Collinearity Statistics
Beta Tolerance VIF
RTA Experience -0.103 -0.908 0.600 1.668 RTA Tariffs -0.023 -0.232 0.829 1.206 RTA Diversity 0.022 0.229 0.642 1.229 Cultural Distance 0.157 1.212 0.467 2.140 RTA Type -0.101 -0.825 0.524 1.908 Border Effects -0.037 -0.315 0.568 1.760 Language Effects 0.383 3.285*** 0.543 1.840 Domestic Acquisitions 0.595 5.295*** 0.618 1.617 Firm Age 0.109 1.094 0.780 1.282
Dependent Variable: Scope
Sign at 0.1 level*; Sign at 0.05 level**; Sign at 0.01 level***; Sign at 0.001 level****
For the third and last variable which is the scope of internationalization, only the control variables experience with domestic acquisitions and language effects have a significant impact on the dependent variable.
4.4 Regression Analysis
The hypotheses of this study will be tested using multiple linear regressions on the dependent variables. In this sample the 72 Multilatinas undertook a total of 932 mergers and acquisitions, of which 32 percent are foreign mergers and acquisitions. From these 291 foreign acquisitions, the hypotheses will be tested on the dependent variables: pace, rhythm and scope of internationalization.
Regression Analysis Pace of Internationalization
Table 1 provides the statistical results of the hypotheses test. In model 1 of the regression analysis on pace, R-squared is .437 for the control variables. The R-squared is explained by the language effects and number of domestic acquisitions.
Regression Analysis on the Pace of the Internationalization process
Model 1 2 3 4 Sig. F change 0.000*** 0.049** 0.884 0.139 R² 0.437 0.471 0.437 0.472 Adj. R² 0.375 0.403 0.366 0.395 Constant 0.000 0.000 0.000 0.000 Independent Variable RTA Experience -0.240** -0.240** RTA Tariffs -0.015 0.034 Control Variables RTA Diversity 0.064 0.032 0.063 0.033 Cultural Distance 0.091 0.153 0.092 0.153 RTA Type -0.140 -0.049 -0.145 -0.036 Border Effects 0.006 -0.023 0.007 -0.027 Language Effects 0.249* 0.267** 0.248* 0.271** Domestic Acquisitions 0.566*** 0.647*** 0.569*** -0.6458** Firm Age -0.020 0.033 -0.019 0.034*
Sign at 0.1 level*; Sign at 0.05 level**; Sign at 0.01 level***; Sign at 0.001 level****
In model 2 the independent variable RTA Experience is entered. The addition of the RTA Experience increased the R-square to (0.471). Hypothesis 1a predicts that the trade agreement
