The existence of W-shaped relationship between internationalization and firm performance: An empirical study of enterprises in Cameroon

Article history: Received January 14, 2020 Received in revised format: March 23, 2020 Accepted June 5 2020 Available online June 5, 2020 This study aims at investigating the relationship between the degree of internationalization and firm performance in Cameroon. Based on a panel dataset of 114 firms surveyed by the World Bank from 2009 to 2019, the regression results find evidence that there exists a relationship between internationalization and performance with respect to the W-shape while the model controls the characteristics of the business environment and firm characteristics. This result has confirmed the impact of internationalization on the firm performance going through 4 stages (Wshaped). We draw a number of implications for the study of internationalization in both theory and practice. . by the authors; licensee Growing Science, Canada 20 20 ©


Introduction
Internationalization is considered to drive enterprises with opportunities to find new markets and gain sustainable benefits from overseas markets (Rugman, 1979). Studies have shown that there are some relationships between the degree of internationalization and firm performance such as positive linear relationship (Bausch & Krist, 2007), negative linear relationship (Geringer). Et al., 2000), U-shaped relationship (Altaf & Shah, 2016;Doan & Bui, 2020;Doan, 2020), inverted U-shaped relationship (Chiao & Yang, 2011), S-shaped relationship (Wu et al., 2012). Recently, an M-shaped relationship (Almodóvar & Rugman, 2014) and W-shaped (Fernández-Olmos et al., 2016) were also found. The research results show that internationalization is a quite diverse and complex process. However, most of the empirical studies on internationalization is done in developed countries in Asia, America, and Europe, the mighty continents with many economic powers (Pham & Pham, 2020). At the same time, very few studies have been conducted on the same issue in countries in Africa, which are famous for harsh conditions, poverty, and underdeveloped economies. Therefore, an observational empirical study is that businesses in developing countries in Africa need to be conducted to verify whether the conclusions of previous studies are relevant to the situation and the reality of small, less developed countries?
In recent years, Cameroon -one of Africa's most developed countries and dubbed for being "miniature Africa" -shows some positive signs of an economy going up as the speed increases. The country's growth continued to increase from 1.9% to 5.6% tripling in the period 2009(World Bank, 2017. Thus, in the context of the current international economic integration, the question is whether or not the internationalization level affects Cameroon businesses' business performance? The answer is the content of this article.

Literature review and theoretical background
Various empirical studies have provided evidence about the benefits and costs that internationalization brings, but there is still no clear consensus on the relationship between internationalization and performance. Many studies have shown that there is a positive linear relationship between internationalization and firm performance (Bausch & Krist, 2007;Thomas and Eden, 2004), some studies show that expansion of operations overseas operations will incur many costs, Geringe et al. (2000), Gomes and Ramaswamy (1999) point that an increase in the level of internationalization leads to a decrease in performance. The two-stage relationship is proven by Ruigrok and Wagner (2003), Capar and Kotabe (2003), Hitt et al. (1997), Chiao and Yang (2011); The three stages of internationalization Contractor et al. (2003), Lu and Beamish (2004); The relationship is in four stages (Almodovar, 2012;Lee, 2013;Almodovar and Rugman, 2014). Specifically, in phase 1, companies often start to expand their operations in countries in the region, close to and have many similarities with their home markets. This helps the companies to minimize the costs associated with adapting to the new institutional and cultural environment (Peng, 2002). In the second phase, as companies increase their penetration into foreign markets, foreign responsibilities will arise as the geographical and cultural gap between origin countries and host countries causes the rise in costs. Moreover, businesses already learned a lot of experience from their local market on how to take advantage available in technology, management, and promotion. Furthermore, the increasing level of internationalization leads companies to expand their operations entirely out of the domestic sector and thus face the responsibilities, commitments, and constraints of market entry in the new region. Therefore, whether the research results mentioned are suitable for Cameroon's practical situation needs to be verified. For that purpose, this study was conducted to determine the relationship between the level of Internationalization and the performance of enterprises in Cameroon.
Hypothesis H1: The degree of Internationalization has a W-shaped relationship with the business performance of enterprises in Cameroon.

Methods of research
The research utilizes available secondary data sources conducted by the WorldBank (WB) by collecting surveys on business managers. The data set used in this study surveyed businesses operating in Cameroon in 2009 and 2016 to look at a large scale with a sufficiently large number of observations and to attach time to estimation results to be as accurate as possible. In addition, selected businesses are representative of the region in which they operate. Data used for the study is panel data, including 228 observations for 114 enterprises over the two years, 2009 and 2016. To estimate the impact of the independent variable on the dependent variable, the study used panel data with two models; the fixed effect model (FEM) and the random effect model (REM), with the estimation equation as follows:

Dependent variable -Return on sales (ROS):
There are many ways to measure the performance of enterprises. In particular, performance measurement based on accounting-related indicators is the most common such as ROA (Hitt et al., 1997;Geringer et al., 2000), ROE (Grant, 1987;Thomas & Eden, 2004), ROS (Contractor et al., 2003). The article uses the ROS index because the number is not affected by inflation or accounting principles, which will more accurately reflect the company's performance.
Independent variable -Degree of Internationalization (DOI): Many studies use a variety of methods to measure Internationalization. In particular, the FSTS -Overseas revenue over total revenue -is the most commonly used (Grant, 1987;Geringer et al., 1989;Riahi & Belkaoui, 1998). Ruigrok and Wagner (2003) also pointed out that this is a measure of the most core aspect of a company's Internationalization-financial aspect. The formula for calculating FSTS: = Firm size (SIZE): measured by taking the natural logarithm of the total number of employees (Altaf and Shah, 2016).
Years of Operation (AGE) was measured by the years of operation of the enterprise from 2009 to 2016.
Foreign equity share (FOREIGN): calculated through a percentage of a company's capital owned by foreign individuals, organizations, or international companies (Krammer et al., 2017).
Managerial Experience (EXPER): measured by the number of years of management experience as of 2009 and 2016 (Tu, & Uyen, 2017).
Manager gender (GEN): is a dummy variable, getting value equal to 1 if the manager is Male and 0 if Female. According to Felson and Gottfredson's (1984) research, men are more active and have more conditions to interact with outside society than women.
Access to finance (FINANCE): measured by the manager of the business answering the question on the Likert-5 level of "At what level, access to finance is a hindrance to operations of the business?" (0: unimpeded → 4: Extremely obstructed).

Bribery (BRI):
An unofficial payment for a particular benefit, but contrary to the public interest or in violation of the law. According to De Jong et al. (2012), BRI has a nonlinear relationship with the performance of enterprises, meaning, in the first stage, to quickly meet the requirements, businesses accept to pay out a negative fee to shorten the waiting time and take advantage of valuable opportunities, so BRI will positively impact the profitability of the company.
According to Le (2008), BRI is a measured by the ratio of the amount of bribery that businesses have to spend on average and the total value of corporate assets multiplied by 1000, but because of the data suitability used, BRI measured by the amount of bribery divided by the total revenue of the business multiply by 1000.
Competitor (COMPETE): is a dummy variable, receiving value 1 if the enterprise must compete with other informal businesses or businesses have not registered, otherwise receiving the value of 0.
Customs procedure and trade rules (CUSTOM): measured by the manager of the business answering the question on a Likert-5 scale "To what extent, customs and trade laws are the hindrances to business operations" (0: unimpeded → 5: Extremely obstructed).
Lack of educated labor (WORKFORCE): measured by the manager of the enterprise answering the question on the Likert-5 scale "At what level, the lack of qualified labor is an obstacle to business activities" (0: unimpeded → 5: Extremely obstructed). The average percentage of the business's profit-to-sales ratio is 77%, with the lowest value is 3%, and the highest is 99%. Meanwhile, the average Internationalization level of Cameroon businesses is 8%, which means that international sales accounted for 8% of the total sales. Not all enterprises are participating in Internationalization, having the lowest value of 0%, which means that they have not yet involved in Internationalization, and the highest is 100% when these enterprises export all of their goods and services to foreign markets.  Table 2 presents the correlation matrix of variables, including the correlation coefficients between pairs of variables. If the correlation coefficient between pairs of variables has a value of 0.6 or more, then there is a strong correlation or very strong, notice that the problem of multicollinearity may occur. According to the correlation matrix, the correlation coefficient between the dependent variable is ROS with independent variables, and the correlation control variables are weak and medium. Thus, independent variables and control variables can be included in the model to explain the dependent variables, and it can be confirmed that no multi-collinear phenomena occur.  (2019) The regression results for FEM and REM models are presented in Table 3. The first four models regressed according to the FEM fixed-effects model. The 1st to 4th exponent of the independent variable DOI was sequentially added from model 1 to model 4 to examine the impact of the independent variable Internationalization level on firm performance as well as their nonlinear system. In addition to the independent variable, ten other control variables are included permanently for all four models. The REM random effects model is applied to model 5 to model 8.   Similarly, the independent variable (DOI) was multiplied by powers of 2, 3, 4, and was introduced in turn from model 5 to model 8 with the number of control variables estimated in the first four models. After evaluating the FEM and REM models, the study continued to perform the Hausman test to see which model was more appropriate in the two models mentioned above. The Hausman test results gave shown that the hypothesis H0 was rejected at the significance level of 1% (p <0.01), suggesting that the FEM model is more suitable in this study.

Results and discussion
According to the Hausman test, the FEM fixed impact model is most suitable for the study; therefore, this model will be used to explain and test the research hypothesis. Table 3 shows that the results of the regression model are completely consistent with the hypothesis of the study with the coefficients of powers from 1 to 4 of DOI are negative ( = -6,09; p<0,01); positive ( = 38,23; p<0,01); negative ( = -60,17; p<0,01) and positive ( = 27,10; p<0,01). This proves that there is a nonlinear relationship between the degree of internationalization and the business performance, namely the relationship in the form of W. Take the first derivative of the Return on sale (ROS) function by the Degree of internationalization (DOI), we have: = -6,09 + 38,23 × 2 × DOI -60,17 × 3 × DOI 2 +27,1 × 4 × DOI 3 = 0,1 = 0 => = 0,52 = 1,04

Degree of internationalization (DOI):
The implication of the above result shows that if the degree of internationalization (DOI) is below 10%, it will reduce the performance of enterprises. If DOI is between 10% and 52%, then the operational efficiency will increase, however, if DOI is from over 52% to less than 104%, it will reduce the operational efficiency and if it is larger or equal to 104%, meaning that when businesses fully export their goods and services abroad, they will achieve growth in performance. Cameroon, although a small developing country in Africa, has had the level of corporate governance reaching the fourth stage, which can be explained by the large and small companies' internationalization activities originating from the market will soon become stronger does not matter little inland activities. For large companies with small domestic markets, the business opportunities in this market have been fully exploited. Therefore, there are few opportunities for these companies to develop similar products or services, and the competition level is fierce. Small firms from the small local market will be very active abroad as a supplier to large companies (Gulati, 1998). On the other hand, companies with large domestic markets are considered as companies with a moderate degree of internationalization because they only started to enter foreign markets after exploiting opportunities in their home market.
Firm size: Negative beta coefficient with a significance level of 1% (β = -0.06, p <0.01) shows that firm size and operating efficiency are inversely proportional to each other. Although the results do not meet the expectations, however, it is consistent with the results of previous studies such as Bigsten and Gebreeyesus (2007); Ramachandran and Shah (1999); Benito-Osorio et al. (2016). The bigger the business, the more complex the operation management, and the higher the management and personnel costs will be. Having to pay a hefty fee for administration and operation, facing the risks of entering the international market while not having many years of experience, the performance will undoubtedly be adversely affected.
Years of operation: years of operation also has an inverse relationship with the performance of the enterprise (β = -0.01; p <0.05). Accordingly, if the business age increases by one year, the operating efficiency will decrease by 0.01 units in the condition that other factors remain unchanged. This is because as companies age, the structure and strategy may become inflexible to adapt in a changing environmental conditions.

Foreign equity shares:
With a 90% confidence, the company's percentage of foreign capital is found to have a positive effect on the performance of the business (β = 0.001, p <0.1 ). Indeed, the larger the foreign-invested enterprise is, the more efficient it is to operate, due to its substantial financial resources ready to support business expansion activities. Moreover, absorbing new experiences, knowledge, and technology from foreign companies to improve productivity and operational efficiency (Almodovar and Rugman, 2014).
Managerial experience: achieving the expected signs and effects (β = 0.007; p <0.01) shows that the more experienced managers are in running a business, the better the performance of the company. This is because of the broad vision, the overview, and the ability to turn around difficulties, thereby making the right decisions and benefits for businesses (Phan Anh Tu and Tran Thi Thu Uyen, 2017).

Managerial gender:
The results show that administrative gender does not affect business performance due to p> 0.1. Thus, it can be concluded that there is no difference between male and female managers in relation to firm performance.
The level of financial access: Coefficient β = 0.06 and p <0.01 show that the level of financial access has a positive relationship with the business's performance. This result indicates that, in practice, not always the obstacles to access to capital are adversely affecting operational efficiency, and this is consistent with the actual situation in Cameroon.

The level of competition:
No relationship between competition and business performance found because the estimated coefficients were not significant (β = -0.03, p> 0.1). Thus, although the negative beta is as expected, the hypothesis that the degree of competition's hindrance harms business performance has been rejected.

Customs procedures and trade laws:
The results have proved that the degree of obstruction of customs procedures and trade rules is always a barrier to the operation of enterprises with a coefficient of β = -0, 06, p <0.05. This is entirely understandable because the commercial laws in the host countries have specific rules and frameworks that every business wants to operate must follow. If these regulations are too restrictive, it will affect the operation of the business in a negative direction.

Bribery:
The regression results show that bribery is related to the performance in U-shape inverse with the correlation coefficient of the BRI variable is β5=0,003, p<0,0, and BRI 2 variable is β6=-0,00002, p<0,01, both variables have 99% confidence. Get the first derivative of ROS according to BRI: = 0,003 -2 × 0,00002 × BRI = 0,003 -0,00004 × BRI In the condition that other factors remain the same, if the ratio between bribery and total revenue is less than 75 ‰ or 0.075 (75/1000), bribery will help increase the operational efficiency of the business. However, when the ratio between bribery and total revenue is greater than 75 ‰ or 0.075, the business performance will decrease with bribery costs. This result is consistent with previous research results of Krueger (1993), suggesting that the cost of bribery works to encourage bribe officials to meet the business requirements or shorten the waiting time as well as cumbersome procedures to reduce opportunity costs.
Lack of educated labor: with β = -0.04, p <0.01; One of the essential factors that make up the operation efficiency of the enterprise is to have high-quality human resources, especially in the trend of international economic integration today. The ability to work in a competitive and evolving technology environment, with the ability to quickly adapt to the rapid changes in science, is a key factor in ensuring the sustainable growth of businesses.

Conclusion
The results of the model analysis with the table data show the relationship between the level of internationalization and the performance of enterprises in Cameroon following the W-shaped. This contributes to the latest research about the fourstage model in the relationship between the degree of internationalization and business performance (Almodóvar & Rugman, 2014;Fernández-Olmos et al., 2015;Ferraris et al., 2016). In addition, the study gave a new insight into the internationalization activities of the small economies from Africa.
Based on the four-stage relationship found between the degree of internationalization and the performance as well as the significant control variables in the model. This research implies that, firstly, businesses need to understand the relationship between internationalization and performance is a dynamic relationship and the impact of the degree of internationalization will cause two degradation in performance, mainly due to the cost of learning and the costs of involving and adapting to new cultural and institutional environments. Therefore, businesses do not rush to stop activities of expanding internationalization when seeing negative signs appear, but persevering in investment, will gain long-term benefits.
Secondly, managers need to focus on improving management capacity, preparing the required knowledge before starting to participate in internationalization by attending seminars, trade fairs, international exhibitions, etc. where they can establish relationships with potential foreign partners (Osei-Bonsu, 2014). At the same time, managers need to measure costs and obstacles in each process of internationalization to have clear directions and strategies in the next growth plans.
Thirdly, on the government's side, legislators need to implement policies to support exporters to promote businesses to participate more actively and effectively in the process of internationalization. At the same time, spend more time to care for improving the factors that are barriers of the business environment in the direction of benefiting businesses such as helping small and medium enterprises to increase access to finance by setting up more funds of their own. Government to support businesses with loans with low lending conditions and interest rates. Besides, the government can open more vocational schools, training, coaching, and improving programs for workers to provide highly qualified workers to promptly meet the needs of enterprises when participating in international competition.