1. Introduction

Multilatinas are becoming global leaders in various industries: food and beverage (e.g., Bimbo), meat processing (e.g., JBS) and manufacturing of steel tubes (e.g., Tenaris). Research on the topic remains limited and it mostly focuses on multinational enterprises (MNEs) and emerging economies’ multinational enterprises (EMNEs) from Asia. Nevertheless, academic awareness of Multilatinas has increased recently (^{Cuervo-Cazurra, 2016a}; ^{Aguilera et al., 2017}).

Multilatinas initial strategies for internationalization are commonly not the same as those of other MNEs; moreover, they usually do not possess similar size, goals, structure, scope, and experience. Thus, they must be examined aside and different methods to analyze their processes are essential. In the case of Colombian companies, with 80%, the preferred entry methods in international markets are exports (^{Gonzalez-Perez & Velez-Ocampo, 2014}), while other MNEs are more inclined to tactic knowledge acquisitions and direct exposure which makes them to select entry modes like partial ownership. The above-mentioned can be engendered because Multilatinas generally present a lack of knowledge and experience to internationalize while other MNEs represent the opposite (^{Nelaeva & Nilssen, 2022}).

There is a direct and strong correlation between internationalization and performance of EMNEs (^{Bıçakcıoğlu-Peynirci & Morgan, 2021}), nevertheless, while EMNEs tend to obtain their profit from internationalization through monopoly-based activities and fail in establishing competitive advantages based on knowledge, MNEs are inclined to represent the opposite (^{Buckley & Tian, 2017}).

To measure a firm’s degree of internationalization several indicators have been created and numerous variables are considered, one of the most commonly used indicators was proposed by the ^{UNCTAD (2020)} and is composed of “three ratios: foreign assets to total assets, foreign sales to total sales and foreign employment to total employment” (^{Trentini, 2021}). However, as it was established by ^{Gonzalez-Perez and Velez-Ocampo (2014)} 80% of the Colombian firms start their internationalization process through exports, consequently, aspects such as foreign assets are not generally expected, therefore, the indicator just as others that do not consider the specific case of Multilatinas, could present unexpected variations that change the measure of internationalization.

The foregoing illustrates the differences between multinational enterprises (MNEs) and emerging economy multinational enterprises (EMNEs), highlighting the limitations of existing indicators in measuring the internationalization of Colombian Multilatinas. These differences are essential in defining a unique identity for these firms, emphasizing the significance of developing new indicators and approaches to accurately assess their internationalization.

For the purposes of this study, two key properties of composite indicators are worth highlighting. Firstly, composite indicators enable the proper ranking and consolidation of numerous variables within a single value. This consolidation facilitates a more comprehensive analysis of Multilatinas’ internationalization. Secondly, composite indicators allow for the identification of key constituents influencing the analyzed phenomenon (^{Lafuente, et al., 2022}; ^{Lafuente, et al., 2022a}).

Although various indicators that measure the internationalization of companies have been created, there are few studies that analyze how can Colombian Multilatinas that possess numerous differences with companies from other countries quantify it. As follows, the present study proposes a composite indicator to measure their level of internationalization by answering the research question: What factors explain the heterogeneous degree of internationalization of Colombian Multilatinas?

To answer the research question, the four biggest Colombian Multilatinas included in the Ranking of Multilatinas 2020 proposed by América Economía were analyzed. The sample includes ISA, Grupo Argos, Grupo EPM and Empresa de Energía de Bogotá.

Therefore, this study includes a knowledge transfer product to society, whose main contribution is found in the identification of measurable factors and variables that can guide the decision-making of the stakeholders involved. Likewise, the contribution of this composite indicators allows to delimit in a practical way (based on official data published by internationalized companies) the capacities of certain Colombian Multilatinas and its applicability is reproducible with companies from other regions of the world. The research shows that to reach the most optimal levels of internationalization, the Economic Capacity factor is essential, followed by Human Capital, then Technology and finally Sustainability. This finding is derived from the perception of 57 experts from a Colombian Business School and, in any case, can be contrasted with perceptions of other experts in the field of international business.

This article is structured as follows. The first section provides a literature review of the internationalization processes of Multilatinas and introduces existing indicators for measuring internationalization, afterwards, the methodology of multiple case study is selected to analyze the sample and described with other methodological considerations including the selection of cases, data collection and analysis. Ultimately, findings and conclusions are formulated.

2. Literature review

Internationalization is a term that has been discussed for years, nonetheless one of the most used definitions by scholars was proposed in 2009, herein, internationalization represents “the degree in which a firm’s sales revenue or operations are conducted outside its home country” (^{Li, 2009}).

The internationalization of Multilatinas, is an important topic of study because these companies have been occupying an important role in the international arena due to their competitiveness in cost and knowledge-intensive activities (^{Duque-Grisales et al., 2019}). As claimed by ^{Duque et al., (2021)}, the quantity of publications regarding this topic has increased since 2015; hence, the authors highlight that almost 60% of the research papers on this topic have been published since that year.

The internationalization process of Latin America companies involves the assignment of financial resources to balance the lack of familiarity within the target markets (^{Manotas & Gonzalez-Perez, 2020}). A study conducted by ^{Gomez-Trujillo (2021)} on eleven of the Colombian Multilatinas analyzed by the Boston Consulting Group showed that most of them have consolidated their internationalization through acquisitions likewise, organizations tend to choose a place geographically close to Colombia to conduct this procedure. ^{Gonzalez-Perez and Velez-Ocampo (2014)}, delve into the geographical aspects and affirmed that Multilatinas enjoy advantages when operating within the region. Similar levels of purchasing power, social and economic development, along with the low geographical, institutional, and physical distance, contribute to reducing the cultural impact of the internationalization process.

Before delving into the entry modes of Latin American companies, it is imperative to describe certain typical characteristics of this process. ^{Gonzalez-Perez and Velez-Ocampo (2014)} found that internationalization of EMNEs is carried out in sectors with less technological intensity. They also highlight the role of the importance of management leadership as a primary driver of internationalization, with access to information on target markets being especially important.

Multilatinas possess different strategies for internationalization. Firstly, ^{Gonzalez-Perez and Velez-Ocampo (2014)} note that firms from developing countries tend to maintain their production at their primary headquarters rather than establishing production facilities in multiple countries. Market selection is driven by different factors with the ease of doing business, market size, and international experience being the most important, nonetheless, the geographical distance is a key factor too. Another remarkable phenomenon in Latin America is the conversion of state-owned monopolies into privately held Multilatinas that continue to have relationships with the government and use those connections to expand internationally (^{Aguilera et al., 2017}). The authors also present another method where companies benefit from the advantageous regulations to generate profits, which are then invested in foreign markets to mitigate the risk of regulatory changes. Finally, domestic market growth and deregulation are also associated with companies expanding their scope to foreign markets.

In the first strategy mentioned, government influence performs a key role in the internationalization of EMNEs. ^{Nuruzzaman et al., (2019)} argue that direct or indirect government support can reduce the regulatory obstacles of companies leading to changes on prices and improved competitiveness. Thus, institutional support positively influences the internationalization of EMNEs. ^{Finchelstein (2017)} by demonstrating the connection between state actions and the internationalization of Multilatinas. His study shows the case of three Latin American countries where State actions have influenced the path of companies and their internationalization endeavors.

In terms of the entry modes, Latin American firms are more likely to export before they advance to the subsequent stage of internationalization, then, their strategy increases gradually to have more control and commitment of resources while also acquiring knowledge of the market. In other words, the companies tend to follow the internalization process theory, it is substantial to notice that the internalization process theory establishes that the firms increase their exposure and presence in foreign markets gradually as their uncertainty is reduced, nonetheless, there are some firms that do not follow the same path, since they increase their operations abroad through acquisitions (^{Gonzalez-Perez & Velez-Ocampo, 2014}). In numbers, the authors suggest that direct exports are the most frequent method of foreign operations with approximately 60% followed by indirect exports with 25.25%, M&A with 13.13%, FDI with 5.05%, franchise with 4.04% and license with 3.03%.

Nonetheless, it is significant to notice that Latin America provides a contrast when it comes to Born Global companies which are firms that since their creation expect to receive a portion of their income from international market operations, consequently, these organizations do not follow the internalization process theory, but they start their operations in an international market since the beginning of their life. It is mentioned that factors such as innovative capacity and the mindset (international oriented) of the directors are of deep importance for the internationalization process of these companies (^{Escandon-Barbosa et al., 2019}).

Due to the importance of internationalization for companies, governments and scholars, the measure of it has become increasingly significant. Thus, many indicators of this phenomenon and multidimensional indices that aim to analyze the organizations’ presence abroad have been developed. Individual indicators are ordinarily used for the construction of composite indices such as the proposed by the ^{UNCTAD (2020)} that will be discussed subsequently (^{Szymura-Tyc, 2013}).

The first index that will be analyzed was published by the ^{UNCTAD (2020)} and is one of the most used for quantifying internationalization, according since 1990 the organization has issued different indicators, their transnationality index (TNI) uses three variables for measuring the abroad presence of MNEs, those are: foreign assets to total assets, foreign sales to total sales and foreign employment to total employment moreover, their most recent index, called FDI lightness index measures the intensity of foreign operations in relation to domestic or global activities by using “the share of sales generated by foreign affiliates and the corresponding share of foreign assets” (^{Trentini, 2021}). This index was primarily developed to comprehend the international involvement of digital MNEs, it is substantial to highlight that the UNCTAD’s indicators possess an advantage over the others in this list because the Organization relies on mostly available information across countries and companies. Nonetheless, according to ^{Szymura (2013)}, the weakness of their indicators is the lack of consideration of factors that are uneasy to measure.

The second index is focused on Multilatinas, The Ranking of Multilatinas, proposed by América Economía is designed to compute the degree and potential of globalization for Latin-American enterprises using four dimensions: commercial strength abroad, employees abroad, degree of geographic coverage and expansion potential. It is substantial to note that the ranking is created under various initial conditions, it only considers Multilatinas with sales over US$230 million per annum and with presence in at least two countries different from the one of origin (^{América Economía Intelligence, 2021}; ^{Infobae, 2021}).

The third index uses five ratio variables: foreign sales as a percentage of total sales (FSTS), foreign assets as a percentage of total assets (FATA), overseas subsidiaries as a percentage of total subsidiaries (OSTS), top managers’ international experience (TMIE) and psychic dispersion of international operations (PDIO). Through the following operation: FSTS + FATA + OSTS + TMIE + PDIO = DOI, calculates the degree of internationalization for a firm. Therefore, the values of DOI must be in a scale of 0.0 which means no international involvement to 5.0 meaning total international involvement. The advantage of this linear combination is the reduction of the spread of systematic and random errors, nevertheless, the author acknowledges the limits of his method and the speculative nature of the measure (^{Sullivan, 1994}).

The fourth index is named internationalization index (INT) and was built under the consideration of numerous variables such as outward internationalization forms and markets (OIFM), inward internationalization forms and markets (IIFM), export share in sales and markets (EXSM), import share in sales and markets (IMSM) and firm international experience and markets (FIEM), each with a different weight within the measure. The index was designed to provide a more effective method for appraising the internationalization degree of companies at an initial stage (^{Szymura-Tyc, 2013}).

The fifth measure of firm internationalization is named ratio of international market shares (RIMS) and is defined as the ratio between the market share for all the countries of the world (excluding the one used in the denominator) and the market share for the country where the country has its greatest market share. Consequently, RIMS range from zero for domestic firms to one for maximally internationalized firms (^{Marshall et al., 2020}).

Considering the previously mentioned information, it is crucial to emphasize that there is a consistent pattern among most indicators, despite their inherent differences. One key aspect shared by them is the recognition of internationalization as a multidimensional construct (^{Hsieha et al., 2019}). By acknowledging its multifaceted nature, researchers have sought to capture the complexity of internationalization by utilizing a variety of variables in their analyses.

Another noteworthy aspect observed in certain indicators is the utilization of heterogeneous weights for the variables. This approach offers distinct advantages as it accounts for the varying degrees of importance that different dimensions may hold in the context of internationalization. By incorporating such heterogeneous weights, indicators aim to provide a more accurate representation of the phenomenon. Notably, the use of endogenous weights proposed by experts has proven effective in developing composite indicators (^{Lafuente et al., 2022}).

Ultimately, it is important to note that “there is neither single indicator nor an index that satisfactorily measures the overall degree of the internationalization of a firm” (^{Dörrenbächer, 2000}), the aforementioned because as ^{Sullivan (1994)} stated, “the reliability of measuring the degree of internationalization of a firm remains speculative”.

3. Methodology

This study is based on a multicriteria approach based on ^{Baena-Rojas et al., (2018)} ; ^{Baena-Rojas et al., (2016)}; ^{Baena-Rojas et al., (2022)} to construct the composite indicator. Then, according to literature review, various indices try to capture the essence of internationalization, they are both univariable and multivariable, even though the choice remains a debate among scholars. This is why ^{Ietto-Gillies (2009)} concludes that internationalization remains a multidimensional concept that depends on the analyzed framework.

Hence, according to the previous literature review about internationalization the current composite indicator is modeled considering four factors. All of them, which were selected to capture the most fundamental concepts around the internationalization of Multilatinas. In this case: economic capacity, sustainability, technology, and human capital. Within each of the mentioned factors three variables were analyzed looking to capture all aspects of it.

The variables were selected because of their relationship with internationalization as can be observed in Table 1. Afterwards, adopting an overall average from 57 experts (professors) from a Colombian Business School. Thereby the variables were weighted to compose each factor, then the factors were also weighted to obtain the percentage of each criterion.

In other words, the weights of each factor and variable were supported the participatory method known as the Budget Allocation Process (BAP). In this sense, it was required the active involvement of expert professors in international business. Then, through a survey, they assigned a score ranging from 0 to 100 to each factor and variable based on their expertise, ensuring that the total sum equated to 100 (See Figure 3) (^{Karagiannis & Karagiannis, 2020}; ^{OECD, 2008}).

Similarly, it is displayed a box plot presenting key statistics derived from the survey’s results of the factors. The box plots offer a visual comparison of the medians and variability across the variables Economic Capacity, Sustainability, Technology, and Human Capital. The median values reveal that Economic Capacity tends to possess the highest values, followed by Human Capital, Technology, and Sustainability. Moreover, the interquartile ranges indicate considerable variability for Economic Capacity, while Human Capital and Sustainability exhibit similar levels of variability. In contrast, Technology demonstrates slightly lower variability. Notably, outliers are observed in certain variables, suggesting the existence of potential extreme values. These results offer insights into the distribution of the weights assigned to the factors by each professor (see Figure 1).

Likewise, following the proposed dynamic of Figure 1 it is also displayed the weights of the variables. As it can be seen, the interquartile ranges demonstrate significant variability for the variables across all spheres, except for Human Capital where the professors have assigned comparable weights. It is worth emphasizing the presence of outliers across most variables, indicating the existence of extreme values.

Figure 1: Results of CMCI’s scores on factors 

The results of each variable and factor were weighted to identify the internationalization degree of the companies, therefore, Figure 2 shows how the methodology works and the associated weights conforming to the survey's results. The survey uses accidental sampling with 57 respondents. All this considering the availability of resources at the time of the study. In any event, although it is non-probability sampling, it does not mean that it cannot show and properly compile the opinions of the experts in the weight they assign to each factor and variable (^{Otzen & Manterola, 2017}; ^{Rivaldo et al., 2022}).

Figure 2: Results of CMCI’s scores on variables 

Figure 3: Criteria technique for Colombian Multilatinas composite indicator (CMCI) 

As follows, the composite indicator’s construction process involved the implementation of the following steps:

Source:^{Baena-Rojas et al., (2018)};^{Baena-Rojas et al., (2016)};^{Baena-Rojas et al., (2022)}.

Equation 1: Normalization of values 

Considering the Equation (1), it is coherent to establish that the results of each variable and factor will be transformed into a grade in the range of (0,5) depending on the relation between them and the internationalization concept. It can either be directly proportional (DP) in which case the first part of the equation is used or indirectly proportional (IP) using the second part.

Source:^{Watkins (2022)}.

Equation 2: Weighted Average 

4. Data and Results

4.3 Results

It should be noted that the results show that the weights of the factors have an amplitude of 11.75% considering the economic capacity as the most crucial factor from professor’s perceptions and sustainability as the least significant. The following step consists in using Equation (2) and the results of Table 3 to obtain a clear outcome regarding the level of internationalization for each company, thus weighted averages of each factor are added. By doing so, the results are presented in detail within Table 4 and Figure 4 also.

Table 3: Values obtained for each variable 

	Weight	ISA		Grupo Argos		Grupo EPM		Grupo de Energía de Bogotá
Economic Capacity (ECAP)	32.28%	Value	Normalized Value	Value	Normalized Value	Value	Normalized Value	Value	Normalized Value
Total Sales (TOSA)	38.86%	9,110,847	2.30	13,990,523	3.53	19,798,912	5.00	5,125,946	1.29
Costs (COST)	32.33%	3,864,858	3.53	10,631,924	1.28	13,508,628	1.01	2,729,542	5.00
Total Assets (TOAS)	28.81%	54,029,723	4.24	50,773,567	3.98	63,778,020	5.00	31,152,646	2.44
Total ECAP			3.3		2.9		3.7		2.8
Sustainability (SUST)	20.53%
Emissions tCo2 (EMCO)	32.67%	45,793	0.03	8,615,133	5.00	2,846,932	1.65	137,764.51	0.08
Gender Equality (GEEQ)	35.81%	0.29	3.37	0.43	5.00	0.11	1.28	0.33	3.84
Investment in Communities (INCO)	31.52%	40,605	5.00	30,000	3.69	25,487	3.14	23,695	2.92
Total SUST			2.8		4.6		2.0		2.3
Technology (TECH)	21.58%
Foreign Shareholders (FOSH)	40.10%	0.09	4.56	0.10	5.00	0.00	0.00	0.00	0.00
Remote Work (REWO)	32.07%	0.69	3.45	1.00	5.00	0.69	3.45	0.85	4.25
Local Suppliers (LOSU)	27.83%	0.90	4.47	0.91	4.45	0.81	5.00	0.96	4.20
Total TECH			4.2		4.8		2.5		2.5
Human Capital (HUCA)	25.61%
Foreign Employees (FOEM)	40.00%	0.60	5.00	0.44	3.67	0.19	1.58	0.43	3.58
Employee Satisfaction (EMSA)	29.67%	0.86	4.73	0.82	4.51	0.86	4.71	0.91	5.00
Average Training Hours (ATHO)	30.33%	67.0	5.00	54.3	4.05	34.0	2.54	20.2	1.51
Total HUCA		4.9		4.0		2.8		3.4

Table 4: Scores and description’s ranges of Colombian Multilatinas composite indicator (CMCI) per factor 

	ISA	Grupo Argos	Grupo EPM	Grupo de Energía de Bogotá
Economic Capacity (ECAP)	3.3	2.9	3.7	2.8
Sustainability (SUST)	2.8	4.6	2.0	2.3
Technology (TECH)	4.2	4.8	2.5	2.5
Human Capital (HUCA)	4.9	4.0	2.8	3.4

Figure 4: Overall score of Colombian Multilatinas composite indicator (CMCI) 

Considering Table 4 and Figure 4 the composite indicator is completed with a depiction of this current study. In this way, the evidence suggests that Grupo Argos is the most internationalized firm and the most prepared enterprise for today's global environments while Grupo de Energía de Bogotá represents the lower internationalized firm within the sample.

Similarly, if the obtained data are compared in terms of strong and weak points with the analyzed cases. For example, it is possible noticed that the value of the weakest factor of ISA (sustainability) is the second largest among the analyzed Multilatinas with 2.8.

Also, Grupo ARGOS has a clear bottleneck in Economic Capacity factor regard all the other considerer firms with 2.9. Meanwhile, Grupo EPM and Grupo de Energía de Bogotá report a relatively balanced structure with a clear high performing in Economic Capacity and Human Capital factor with 3.7 and 3.4 respectively.

It is crucial to point out that considered variables possessed different weights in the calculation. Moreover, the results show that Grupo Argos has the best infrastructure in terms of the analyzed factors on the whole, thus it enjoys a greater opportunity of expansion to other international markets as it follows the international process theory.

Source: Own Elaboration.

Equation 3: CMCI Equation 

The Frobenius inner product notation is used in equation (3) to construct the Colombian Multilatinas Composite Indicator (CMCI). Each variable's weight within the equation reflects how it will affect how internationalization is calculated. Notably, the CMCI computation gives the most weightage to the Economic Capacity (ECAP) composed by TOS, COST, and TOAS. Therefore, any changes in this economic environment would cause the CMCI index to change significantly. Additionally, the Human Capital (HUCA) is another significant element, with FOEM and ATHO showing the highest weight percentages of 40% and 30.33%, respectively. This means that changes in the number of foreign employees and the average training hours would have a significant impact on the total HUCA score and, as a result, the composite indicator as a whole.

In conclusion, the Colombian Multilatinas Composite Indicator emphasizes the significance of competitiveness as a primary driver of success for companies and economies, as stated by ^{Porter (1990)} . It also recognizes that the ability of a firm to reach potential markets and business networks plays a substantial role in achieving competitiveness. The aforesaid, because competitiveness is directly linked to internationalization (^{Elena-Mădălina et al., 2017}), likewise, internationalization positively impacts a firm’s performance and enhances competitiveness. In other words, the expansion of a company’s activities abroad enhances its chances of prevailing in the market (^{Manotas & Gonzalez-Perez, 2020}). Therefore, the selected variables of the composite indicator were chosen since they provide information on the competitiveness of Multilatinas and their capacity to expand to new markets, as these factors are inherently connected.

5. Concluding remarks, implications, and future research lines