FDI Escapism: the effect of home country risks on outbound investment in the global economy

1.   Introduction

The recent systematic transformation through liberalization and globalization, increase in investments and trade transactions, as well as the growing interest in the internationalization of firms to remain competitive and become part of the global supply chain, led to the increasing scholarly research papers which seek to explain internationalization process along with the motivation, location, process, etc. With internationalization, firms are able to mitigate the risk of being based in unpredictable and underdeveloped environments which are viewed as comparative disadvantages. But many of these studies have focused on institutional void (Doh et al., 2017; Stoian & Mohr, 2016; Gammeltoft et al., 2010; Liang et al., 2012) and misalignment (Barnard & Luiz, 2018; Witt & Lewin, 2007). However, investments escape from the home country through outward FDI indicates an important but under-explored concept in international business (IB) and economics literature (Witt & Lewin, 2007). This study examines the effects of home country composite risks on firms' internationalization motive through outward FDI and determine which components of home country risk "pushes" firms to initiate the FDI escapism phenomenon in global market. Specifically, firm's strategy behavior for internationalization of global market as response to home country risks which include economic risks, financial risks, and political risk, are determine with emphasis on escape FDI.

According to Cuervo-Cazurra & Ramamurti (2014), to overcome competitive disadvantages, firms may strategy exiting the country through outward FDI due to home country instability, regulatory institution or bad image regarding firm's competitiveness, a phenomenon referred to as escapism FDI. Increase in country risk can lead to political instability, economics risk such as higher inflation, recession, financial risk such as currency fluctuation, expropriation, government default in bond and financial commitment, etc. We argue that home country escapism FDI are not only due to institutional void and misalignment, but also by means of components of country risk via political, economic and financial risks. This corroborates with García-Canal & Guillén (2008) findings that firms in a regulated environment reacts differently to macroeconomic and political risk and seek to expand in countries with stable government with policymaking capacities but shun countries with high level of macroeconomic uncertainty. Although, some developed, developing and emerging economies may have better institutions and economies, other specific domestic risk factors can precipitate management decision to initiate an exit strategy. Therefore, this study also argues that the FDI escapism strategy are not only limited to the developed economies (Kottaridi et al., 2019); developing economies (Barnard & Luiz, 2018), emerging economies (Doh et al., 2017; Witt & Lewin, 2007; Stoian & Mohr, 2016), but to firm in any market economy with internationalization drive. Thus, this study examines the FDI escapism phenomenon in global perspective. Our arguments are built on growing literature from different market economies, where firm utilized outward FDI as a strategic means to exit a competitive disadvantage economy owing to their political, financial, and economic instability. And the OLI paradigm (Dunning & Lundan, 2008) theoretical framework provide a clear understanding of firm's ownership and locational advantages of internalization, and how its interaction supports and promotes escapism motives.

Despite the numerous constraints of economic development such as poor infrastructure, technological backwardness, poverty, and environmental degradation, as well as political instability, outward FDI reached US＄12,670 billion between 2011–2019 from the previous value of US＄11,356 billion in 2018, see Figures 1 and 2. Globalization of the world economy and financial markets certainly increased cross-border activities, and brought considerable growth and development to national and regional economies. Therefore, this study examines how home country risks affect internationalization motives of firm's investment outflow for 127 countries for the period 2003–2016. The panel data framework is adopted, and the study employs the System Generalized Methods of Moment (SGMM), the Difference Generalized Methods of Moment (DGMM), the Fixed Effect (FE) and the OLS pooled regression (OLS) techniques to explore the significant of this relationship. The empirical results have several contributions to the body of literature on FDI and political risk. Thus, we found that home country risk contributed by high political risks and low financial risks components, has moderate negative effects on direct investment abroad. Besides, firm faced with these instabilities (political and financial risks) may seek opportunities elsewhere through diversification of investment or relocation of investment in the form of escaping FDI.

The rest of the paper is structured as follows. Section 2 outlines the theoretical framework and discusses the structure and theories that support the study. Section 3 provides a review of related literature on home country risk and outward FDI as well as the study hypotheses. Section 4 describes the data and the econometric approaches adopted. The results are presented and discussed in Section 5. Section 6 concludes with policy implications and suggestions for future research.

2.   Theoretical framework

2.1.   Dunning eclectic paradigm (OLI), internationalization theory and FDI escapism conceptualization

According to Sharma & Erramilli (2004), internationalization theories are classified into three paradigms, these include the market imperfection paradigm (Vernon, 1966; Hymer, 1960), the behavioural paradigm (Håkansson, 1987; Johnson & Wiedersheim-Paul, 1975), and lastly the market failure paradigm (Dunning, 1977; Buckley & Casson, 1976; Anderson & Gatignon, 1986). Over the year, students of economics and academic scholars have showed that the OLI eclectic paradigm and the theory of internationalization are the major classical theories underpinning international economics (IE) and business (IB) which most literatures on multinational foreign activities rely on. Both the OLI and internationalization theories are firm's level construct, which based their assumption on firms possessing ownership advantages before internationalization. In order word, these frameworks suggest how multinational strategic motive in outward investment drives, as well as the factors influencing their decisions.

Dunning eclectic paradigm (1981) theory utilizes the integration of Ownership, Location, and Internalization advantages (OLI) to describe a popular approach to the study of MNEs' decision as well as activities across border. The basic assumption shows that, investment can be explained by three factors which includes, ownership advantages of firms (O)—describe firm's capabilities and internalization advantages (which are also attributes of firms), the locational factors (L)—indicates where to produce particularly the host countries location-specific conditions, and lastly the internalization factor (I)—explains why firm engages in FDI instead of using proprietary assets (Dunning, 1993a, 1993b). The eclectic paradigm locational factor examines the determinants of FDI including policy induced which gives rise to both institutional and policy interventions (Dunning & Lundan, 2008), hence MNEs' decision processes through institutional context are significant. But owing to ownership data scarcity, most empirical study employs the eclectic paradigm theory to study the location advantages. MNEs initial expansion and growth explained by OLI framework, combines the comparative advantage of nations and competitive advantages of firms (Dunning & Robson, 1987).

However, the application of Dunning (1981) has received huge criticism over the years due to its generality. But despite these criticism, Dunning (1981) theory remains one of the most effective and significant paradigms that facilitate and promotes enquiries raised by economists, illustrating the alliances of various approaches, and simplifying the challenges raised at different stages of analysis (Cantwell & Narula, 2001). Brouthers et al. (2009) empirical paper utilized Dunning's OLI framework and the neural network (NN) to predict the Central and Eastern European countries (CEE) markets choices for EU firms and the market selection decisions. The expanded model was used to showed strong predictive power in explaining the international market selection. Analyzing transactional cost using of OLI model, assist firms make rational decisions (Whitelock, 2002). However, Rugman (1981) suggest that in order to compare the different service sectors and explore the inseparable service effects, there is the need to expand the eclectic theory. This effect was later investigated by Domke-Damonte (2000) with the use of the resource-based theory supported by the OLI model. Examining the political dimension of MNEs behavior was examined by integrating and expanding the political aspect of MNEs into the Dunning's eclectic paradigm of ownership, internationalization, and the location advantages (Boddewyn, 1988). The Dunning's eclectic paradigm model reveals that the market and efficiency seeking are the Chinese provincial firm's main motive of investing in foreign countries (Chen, 2015). With these application specifics of Dunning OLI paradigm framework, this study examines the effects of home country risk on outward investment flow. This will provide an insight to understand firm's internationalization motive through outward investment in the presence of risk perception in global perspective.

In internationalization concept, firms utilize their specific advantage so as to acquire certain advantages in other locations (caves & Porter, 1977; Duning, 1981). However, recent studies have revealed that firm internationalizes in order to get competitive advantage and escape home country risk which can give rise to firms disadvantages in competitiveness (Cuervo-Cazura & Ramamurti, 2014; Khanna & Palepu, 2010). Firms moving to foreign country to seek advantages by evading harsh and poor economic conditions can be referred to as FDI escapism (Cuervo-Cuzurra et al., 2015). While limited FDI escape (occasioned by strain) happens owing to instability and poor institutional reform, extensive escape FDI (occasioned by stress) happens due to institutional changes and societal instability (Barnard & Luiz, 2018). Institutional deterioration (caused by stress) may be due to conditional escape. Through outward FDI, firms can escape or relocate abroad from home country when there are potential risks to investment and misalignment with the country's institution frameworks (Witt & Lewin, 2007; Wu & Chen, 2014). To diversify risks by investing abroad, firms need to identify the key country risk factors as well as the macroeconomic components that drives the external investment. The internalization theory has also been employed to examined outward FDI flow and firms escaping from specific economy due to country's regulatory voids. For example, studies on outward FDI effects using Brazilian and central & eastern European countries regulatory void have been examined (Staian & Mohr, 2016; Stal & Cuervo-cuzurra, 2011). But because most of these studies were on emerging economies, Kattaridi et al. (2019) study extended the FDI escapism concept to developed economies in order to determine whether regulatory context and corporate taxation leads to FDI escapism. The regulatory and structural weakness in the Chinese dairy sector were used as evidence to examine the escape FDI phenomenon, and whether the industry restructuring as well as discrimination, precipitate FDI escape, (Enderwick, 2017).

Despite the not too many empirical literature of escape FDI in behavioral finance literature, the escape FDI phenomenon remain under explored concept in international economics (IE) and business (IB) literatures. But in view of globalization and liberalization of market, which have caused many firms to internationalize their business operations (Asgari et al., 2010), firm's internationalization motives in the presence of different levels of country's risk have been strategic. If the risk is high, MNEs might relocate as escaping FDI. The effects of country risk on FDI flows varies with investor, ditto their responses. For instance, MNEs internationalization through outward flow of FDI can either be influenced by country's instability with relocation as response (escaping FDI) or attaining firm's internationalization motives—market seeking, resource seeking, efficiency seeking and strategy resources seeking FDI. A decrease in home in the components of country's risks, increases the outflow of FDI, which leads to facilitation of trade, bring about foreign exchange for the economy, as well as integrate the economy, etc. But an increase in home country risks hinders investment flow abroad, leads to diversification of investment, reduces production volumes, and increase speculation of domestic economic situation which may increase chaos and create more financial crisis with losses in revenue (Krayenbuehl, 2001). Using the negative binomial regression with Uppsala model on a sample of 166 firms, Jiménez (2010) study examines the nexus between political risk and the scope of internationalisation. Their findings show that firms partially follow the Uppsala model, and that political capability plays a crucial role in firms' internationalization strategy.

Furthermore, a rise in country's instability can lead to relocation of investment to foreign country known as FDI escapism which can further create a net movement of fund out of the economy (Kottaridi et al., 2019). Therefore, if country risk impact negatively, potential investment flow abroad might become escapist FDI. This indicates that the effects of country risk on FDI can either encourage or discourage the flow of outward FDI, which numerous empirical studies reported as either positive or negative (Osabutey & Okoro, 2015). In view of the foregoing, we extend the escaping FDI conceptualization, by examining the effects as well as identifying the significant components of country risk precipitating firm's outward investment as escaping FDI in global perspective.

3.   Review of related literature

3.1.   Country risks and outward foreign direct investment

Many exiting literatures have shown several research studies with plethora of empirical methods the effects of country risks on MNEs internationalisation motives moderated by home country institution. Within this framework, findings suggest that firms with strong labor protection are likely to initiate offshoring in their domestic operation in order to address their home country institutional challenges (Weng & Peng, 2018). However, studies on how country risk affects firm's ownership strategy moderated by firm's home country formal and informal institutions has also been examined (Buckley et al., 2011; Tang & Buckley, 2020; Barnard & Luiz, 2018; Slangen & van Tulder, 2009). The influence of political risk on FDI announcement has also been documented, and findings showed overall negative effects for Taiwanese firms that announces investment during the risky period (Huang et al., 2010). Empirical results showed that among the available determinants of institutional quality, expropriation risk is most crucial. Whilst firm's financial capability plays a crucial role in managing its opportunities and risk, the strength of the relationship between business behaviour & performance and the policy environment also matters (Henisz & Zelner, 2003).

Similarly, literature on country political risks and investment (FDI) nexus have also been of great interest to economics scholars and investors, but most of the existing studies are confined to individual country analysis and experiences. For instance, empirical findings showed that at the national level for Poland, a transitional economies CEE country, liberalization, political risk, and economic reforms are crucial drivers of FDI inflows (Avioutskii & Tensaout, 2016). Holburn & Zelner (2010) argues that the response from MNEs to host-country policy risk vary due to the differences in organizational capabilities for evaluating such risk and managing the policy-making process. Their results also suggest that firms from home countries with weak institutional constraints especially on policymakers will be less intuitive to host country's policy risk in their strategies for global expansion. But investing firms tend to expend lower amount of capital on investments located in higher political risks countries and where the political risk is high (King et al., 2021). Study on Pakistan political violence and FDI inflow found a decline in imports due to the disruption of FDI, commerce and trade flow to the country (Latif et al., 2017).

Even so, results showed that political instability in developing economies of SSA such as Nigeria has a significant influence on FDI inflow to the country (Osabutey & Okoro, 2015). But only recently, how country political instability impacts on firms' internationalization motive through outward FDI grabbed the attention of academic discussions, thus, empirical studies in this strand of literature are scanty. For example, study on China global investment and international trade revealed that high-level partnership under belt and road initiative can increase Chinese outward investment flow if firm and country level coupled with other fixed effects level are controlled (Shao, 2020). Research conducted showed that the risk of stock price crash is reduced by outward FDI, and the effects are more visible in host countries with better institutional environments (Liu et al., 2021). Whilst the nexus between outward FDI and risks in countries with abundant resources have positive moderating effects, the empirical analysis for all African countries indicates negative effects (Lu et al., 2017). Whilst there is a gradual decline in the Chinese political risk distribution, outward FDI flow precipitated by the new Chinese reform increases, furthermore, the Chinese outward FDI flow for the period 2006–2017 were found to concentrate in moderate and low risk countries (GaoYan, 2020).

Furthermore, studies based on country risk and escape FDI relationship have also received very little attention, but the significant effects of the phenomenon to country's economy have led to some research motivations. Besides identifying the conditions that give rise to escape FDI, Enderwick (2017) study criticized the concept of "escape FDI" and posit that industry's restructuring and discrimination leads to escape of FDI. An exploratory study conducted reveals that FDI escape involves three cumulative phases of condition escape FDI (stress), limited escape FDI (strain) and extensive escape FDI (failure) which occurs when the standard "rule of the game" breaks down and the productive capacity of the economy is in distress (Barnard & Luiz, 2018). Utilizing social science and historical framework, escape FDI phenomenon occurs when there are break down or shift in institutional conditions which usually occur in coordinated and liberal economies (Kobrak et al., 2018). Besides the dark side of political risks to multinationals, political risk can also be a source of opportunity for firms to invest and expand abroad by lobbying and supplying needed items, thus political risk positive association with firms' internationalization activities (Jiménez et al., 2014)

Furthermore, the misalignment between home country institutional changes and firms' internationalization motives indicates the extents of outward FDI as escape, and likely to increase based on societal coordination in country's political economy (Witt & Lewin, 2007). But building on Cuervo-Cazurra & Ramamurti (2015) research findings, that firms engaging in emerging economies investments are not only motivated by the "pull" factors such as abundant talents or large markets of the host countryy, but also by "push" factors such as home country's weak institutions, we extend the push factor notion to include the home country composite risks which encompasses the economics, financial, and political components, and determine how it affect firms internationalization motive through outward FDI. Considering the risk associated with the complexities of globalized interdependencies and the prominent role play by outward FDI, this study contributes to literature by examining the effects of home country risk on outward FDI for 137 countries for the period of 2003-2016. This research is based on panel data framework using the SGMM, DGMM, and FE estimators which is robust and consistent to econometric issues related to endogeneity.

3.2.   Hypothesis development

Countries risk can incur additional cost of doing business, creates uncertainty regarding the rules and regulations of home country's institutions, which give rise to spending more resources and using valuable time to understand new regulations and how it affects firm. Thus, country risk negatively affects firms. Increase in country risks can hinder investment flow abroad, leads to diversification of investment, reduces firms' production volumes, and increase speculation of domestic economic situation which may increase chaos and create more financial crisis with losses in revenue. However, investment increases when country's risk is low (Osabutey & Okoro, 2015), therefore we test whether firm's investment flow abroad from home country is firm's internationalization motive (Barłożewski & Trąpczyński, 2021; Cuervo-Cazurra et al., 2015) or escaping FDI due to high components of home country risk (Witt & Lewin, 2007). we formulate a testable hypothesis to examine the effects of country risk on outward FDI flow for the period 2003–2016. We argue that,

Hypothesis (a): Home country composite risk components is highly significant and negatively related to outward investment motive leading to escaping FDI.

Hypothesis (b): Home country economic risk components is highly significant and negatively related to outward investment motive leading to escaping FDI.

Hypothesis (c): Home country financial risk components is highly significant and negatively related to outward investment motive leading to escaping FDI.

Hypothesis (d): Home country political risk components is highly significant and negatively related to outward investment motive leading to escaping FDI.

4.   Methods and Data

4.1.   Data

This study examines the effects of home country risk on firm's internationalization motive through outward FDI for 127 countries for the period 2003–2016. The outward FDI and country risks statistics are obtained from the databases of the United Nations Conference on Trade and Development (UNCTAD) and Political Risk Services (PRS)-International Country Risk Guide (ICRG). However, data related to the components of country risk (economic, financial, and political risks) are provided in Appendix-1, Table A1–A3, and choice of country selections for analysis are based on availability of sample data. These countries are listed in Appendix-1 Table A4. All variables are briefly described in Table 1 and the descriptive statistics are presented in section 5.1 of this paper.

4.2.   Econometric model technique

Based on theoretical perspective discussed in section 2, this study examines the effects of home country risk on outward FDI using the dynamic panel model estimator proposed by Arellano & Bover (1995) and fully developed Blundell & Bond (1998). To achieve this, we specify each model in the panel dynamic framework as follows,

where,

Model 1

Model 2

Model 3

Model 4

Model 5

Model 6

For the general model

where $\beta_{i}(i = 1, \ldots, 14)$ is the coefficient of each variable, the model error term is $\varepsilon_{i, t}$, the unobserved heterogeneity country-specific effects is $u_{i}$. And the time specific effects in the model is $\xi_{i}, O F D_{i, t}$ indicates the outward FDI variables, the ${Country}\; {ris{k}_{i, t}}$ include political risks, economic risk and financial risks. Other explanatory variables in the model include $\tau_{X}, \omega_{Y}$ and $\lambda_{Z}$ which indicates the interaction terms for economic and political risk, political and financial risk, and economics and financial risk respectively. The lagged regressor is added to the regression model due to the persistence flow of OFD variables persistence over time. The experiment time horizon is 2003–2016 (T = 14) and the number of countries (N = 127) changes per year.

In addition to the presence of numerous econometric problems on FDI related analysis, previous studies have showed that FDI flow are persistence cross country (Rahman et al., 2019; Guru & Yadav, 2020). Therefore, it is necessary to employ a model that explains the inclusion of lagged dependent variables in the right-hand side of the model and account for econometric problem such as, endogeneity, heterogeneity, simultaneity bias, omitted variable bias, reverse causality, etc., we specify a dynamic model and follow Arellano & Bover (1995) and Blundell & Bond (1998) to estimate the effects of home country political risk components on outward FDI using the system GMM (SGMM) approach.

5.   Results & discussion

This section of the paper discusses the empirical results of all pre-analysis tests (descriptive statistics, correlation, multicollinearity), and the main analysis using different econometric techniques (SGMM, DGMM, OLS, and FE), as well as the robust checks.

5.1.   Descriptive statistics and multicollinearity test

The disaggregate data of country risk and firms' investment outflow are analyzed. The descriptive statistics and the correlation matrix are presented in Table 2. It describes the nature and distribution of data as well as dispersion. The mean distribution of the country risk components variable shows that economic risk (C) and financial risk (F) have the lowest and highest mean respectively. Regarding the variables standard deviations, economic risk (C) has the lowest clustered data around the mean compared to political risk (P) with more spread-out data. The correlation matrices provide observatory evidence on the level of bivariate relationships between variables. We found that the interaction term $\lambda_{z}$ and economic component, $\tau_{x}$ and economic component, as well as $\lambda_{z}$ and $\tau_{x}$ among the explanatory variable(s) have correlation more than 0.50, which is suspectedly high. These high bivariate coefficients suggests that the problem is related to multicollinearity, hence variables are further tested using Variance Inflation Factor (VIF) to confirm the results of the correlation matrix. The VIF result presented in Table 3 reveals that the individual values of the explanatory variables vary between 1.050–2.560 which is considered far less than 10, a threshold suggested by Wooldridge (2010); Green (2012), and the overall mean values are between 1.000–1.346 which is not significantly greater than 1, indicating the absence of multicollinearity effects among the explanatory variables (Neter et al., 1985; Kamal et al., 2019).

5.2.   Effects of home country risk on outward FDI

The aggregate and disaggregate home country's component risks such as economics risk, financial risks, and political risks are examined and reported in Table 4, 5, 6 and 7. Six (1–6) different empirical models showed in Equations (2–7) are evaluated to investigate the effects of home country risk on outward FDI in global perspective. Whilst model 1 Equation 2, examines the overall country risks effect using the aggregate dataset of economics, financial, and political risks, model 2–4 (Equations 3–5) investigates the effects of the disaggregate or individual component risks on outbound investment and show the direct linkages between home country risk components and direct investment abroad without the influence of other risk variables in the regression model. They are simply in a bivariate regression framework and measures the comovement between domestic risk and outward FDI. Model 5 Equation (6) simultaneously examines all disaggregate risk. It examines the effects of one risk index in the presence of other risks indexes with respect to outward FDI. This study also designs the interaction variables to examines the joint effects of home country risk on outbound investment which is specifically examined in model 6. These models are estimated using the fixed effect (FE) estimator, OLS pooled regression (OLS), difference GMM (DGMM) and the two-step system GMM (SGMM) approach. These econometric techniques have been discussed in section 4.2, and the index risk rating by ICRG (2019) show to assigns low points to high-risk components and vice versa.

The study relies on SGMM estimator as the main estimation technique employs to examine home country risks on outward FDI, because of its capability to deal with several econometric problem associated with endogeneity, heterogeneity, reverse causality, simultaneous bias etc., hence it offers more efficient and consistent estimates compare to the other estimation techniques. However, the results of other techniques such as DGMM (Table 5), FE (Table 6) and OLS (Table 7) estimators, are used as robustness check to validate the consistencies of the main estimator (SGMM). Two tests were used to test the validity of SGMM and DGMM estimated coefficients in the models, and these include the Sargan/Hansen J tests and the Arellano and Bond Autoregressive (AR 2) test. Results presented in Table 4–5 indicates that the Hansen's (1982) J tests values are insignificant which suggests the validity of over identifying restrictions, confirming that the employed sets of instruments in the regression model are not endogenous, and the values of the Arellano-Bond (1991) tests which examines whether error terms have correlation do not reject the absence of second order serial correlation in all estimated models. Aggregate of the home country economics risk (E), financial risk (F) and political (P) risk data shown in Table A1–A3 computed as $0.5(E+F+P)$ is used as proxy for home country composite risk (ICRG, 2017). Table 4 model 1 reports the result of home country composite risk effects on outward FDI for panel of 127 countries using twosteps SGMM estimator. The result shows that home country composite risk effect has negative and significant impact at 5% significant level on outward.

FDI for 127 countries, and the effect is moderately significant (−0.693). This implies that an increase in home country risk decreases global outflow of investment, and this can lead to relocation of investment. This result is in line with Kottaridi et al. (2019), Kobrak et al. (2018) etc. papers.

The results of model 2 shown in Table 4, suggests that the economic risk index has positive and significant impact on global outflow of investment in home countries. This indicates that increase in home country economic risk index increases outward flow of investment at the rate of 0.209% at 1% significant level. Whilst negative risk factors impact negatively and discourages investors, positive factor encourages investors and facilitates the production of goods and services (Mussa, 2000). These findings show that home country economic risk index supports and facilitates the outflow of investment in home country, hence escaping FDI in home country during the period under review may not be due to economic risk factors. The variables for the financial risk index are listed in Table A2 of appendix 1 and result from model 3 (Table 4) estimation indicates that home country financial risk component has negative effect (−0.721) on outward investments. However, this risk effect is low but significant at 5% level of significance. This suggests that increase in home country financial risk decreases the outflow of investment. However, Table 4 model 4 suggests that the overall estimate of political risk in home country is high (−0.696) and negatively significant. This indicates that a unit increase in home country risk decreases the outflow of global investment at the rate of 0.696 at 10% significant level.

Table 4, Model 5 presents the simultaneous estimation of all individual risk components in a regression model. This result shows that political risks can discourage prospective investors, reduce the volumes of investment flow abroad if productions chains are disrupted, and in some cases badly affected firms are "pushed out" of business. That is political risk may lead to FDI escapism (Kottaridi, et al., 2019; Kobrak, et al., 2018). Similarly, the effect of political risk on outward FDI in the presence of other risk components is also negative and significant, but the overall risk effect is very low (−1.003). This could be due to the positive effects of economic risk index which serve to cushion the negative effect of political risk home country (see Table 5, model 5). Due to the same or similar sources of risks, we examine the joint effects of home country risks. With regards to the joint interaction of financial risk with other risk factors, home country financial risk component combination with political risk impacts moderate negative risk on outward FDI but has joint positive effects when combine with the economic risks. This suggests that increase in home country financial risk decreases the significant effects of political risks. But increase in financial risk, increases the significant effects of the economic risks. This means that the positive economic risk index effect depends on the levels of other risk indexes. However, the results and theoretical framework related to political risks and FDI nexus only analyzes FDI inflow using disaggregate data (Wang & Lee, 2021; Barry & Di-Giuseppe, 2019).

5.3.   Results of other econometric techniques (Robustness checks)

We verify the model's (model 1–6) adequacies of the estimated coefficients using numerous estimation techniques such as DGMM (Table 5), FE (Table 6) and OLS (Table 7) as robustness checks. That is to further examine the consistency of the estimated coefficient, we re-estimate the panel regressions model using DGMM, FE, and OLS. The results might not be consistent owing to the presence of lagged dependent variable in the right-hand side of the model, but the estimated coefficients are informative given that the pooled OLS estimation is biased upward, and the FE estimations is downward biased (Nickel, 1981; Bond, 2002). That is, the consistent estimated coefficients should lie within the lower bound of FE estimates and upper bound of OLS (Arellano and Bond, 1991; Nickel, 1981). The estimated result for country's risk in model 1 is negative—0.678 (DGMM, Table 5), −0.714 (FE, Table 6), −0.653 (OLS, Table 7). This also suggests that increase in the overall country risks, decreases investment abroad, and these results appear to be consistent with the results of SGMM technique with estimated coefficient as −0.693. However, the coefficient of OLS is insignificant, but lies on the upper bound of SGMM estimate with the FE estimates downward biased.

The coefficient of the economic risk effect on outward FDI flow estimation presented in model 2 is positive (0.211) and significant at 5% significant level using the DGMM estimator, but the DGMM magnifies gaps in unbalanced panels, hence it may not be consistent (Roodman, 2009). However, the system GMM estimator shows to be both valid, significant and more consistent with a coefficient of 0.209. The result of the OLS estimator is positive with a coefficient of 0.221 at 10% significant level, whilst the FE technique has a coefficient of 0.148 and significant at 1% significance level. The positive result of all estimators appears to be consistent with the SGMM techniques and suggests that increase in home country economic risk increases global outward FDI. This study also used the DGMM, FE, and OLS estimators shown in model 3 to examines financial risks and outward FDI relationship in order to determine the parameter adequacy as well as the consistency of the SGMM model. Both the SGMM and DGMM estimations are negative and statistically significant at 5% significance level, but the estimate of the SGMM (−0.721) appears to be more reliable than the DGMM (−0.704). This is supported by the Blundell (1998); Roodman (2009) papers on the robustness of the system GMM. The values of OLS (−0.694) and FE (−0.746) estimates also lies on upper and lower bound respectively of the SGMM estimates. This further shows the consistency of the SGMM estimates. In model 4, the informative estimates of the political risk using DGMM, FE and OLS is also examined. The results shows that the coefficients of OLS (−0.627) and FE (−0.718) lies upward, and download biased respectively, and the DGMM is −0.641. These results are unanimously negative indicating that increase in political risk reduces outbound investments.

The consistency of SGMM and the DGMM estimators requires the absence of second order serial correlation in the error term, and the overall instruments validity for each model should be appropriate. The values of the second order Arellano-Bond (1981) specification tests (AR2) shown in Tables 4 and 5, indicates that the absence of second order serial correlations is not rejected. This implies that the error structures of the model are serially uncorrelated, suggesting that the results are valid, and the model's specifications derived in section 4.2 are appropriate. Beside the AR (2) test, the values of Hansen's (1982) J tests are insignificant, indicating that the over identifying restrictions are valid and the specification instruments are valid. The proliferation of instrument which can weaken both autocorrelation and Hansen tests specification, remain one major drawback in GMM estimator. However, to overcome this problem, we collapsed the instrument matrix and ensure that the group numbers is larger than the instruments (Roodman, 2009), hence all instrument ratios in the specified models are not less than 1 (≥1), see instrument ratios in Tables 4 and 5. The consistency of the estimated coefficients in Tables 4 and 5 are further examined using the FE and OLS estimators, and the values of R-squared are large, suggesting that the specified models explain at least 38.9% and 37.5% for FE and OLS respectively the effects of the components of country risk on outward FDI. The robustness checks conducted validates the model specifications, and the coefficient of the estimated parameters using SGMM estimator are consistent and unbiased, hence the results found validly describes the effects of country risks on outward FDI.

5.4.   Endogeneity issue

Outward investment flow from a firm can also be followed by several outward flow from other competing firm in home country. This explains the inclusion of outward FDI lagged values in section 4.2 Equations 1 and 3, to explain the behaviour of FDI. This may bring about a volatile outward FDI flow with time dependence. The cross-country intrinsic features of FDI can also create cross sectional heterogeneity of FDI across countries. Besides, there could be possible omission of important explanatory variables in the regression model which can lead to correlation between error term and the country risk components, thus violating the OLS pooled regression basic assumption (Abdallah et al., 2015). Country risk plays an important role in firms internationalization motive through outward FDI. For instance, high country risk disrupts firm's production and supply chains, delay foreign export, reduces foreign capital inflow, which ultimately leads to FDI escapism and a decline in the overall output level of the economy. Therefore, country's risks and outward FDI relationship can be influenced endogenously (Hoque et al., 2018). Endogeneity occurs when the condition of exogeneity becomes violated (Wooldridge, 2002). Endogeneity issue is a serious econometric problem such that one endogenous variable can misrepresent the entire estimates or severely distort OLS regression model. Thus, estimation of models prone to endogeneity issues remain an enormous challenge to static estimators such as OLS. Nevertheless, the FE estimator can resolve the simultaneity bias issue, but dealing with reverse causality remain a daunting task. To address potential endogeneity issues, we employ the DGMM and SGMM estimators which do not assume the condition of normality and heteroscedasticity but use set of internal instruments to solve endogeneity issue. However, due to its wilder choice of instruments, this study uses the robust two-step SGMM model proposed by Arellano & Bover (1995) and developed by Blundell & Bond (1998) as the main estimator technique in examining home country risks and outward FDI nexus.

6.   Conclusions and policy implications

Firms escape (escapism FDI) from home country via direct investment abroad remains a crucial but unexplored phenomenon in international economics (IE) and business (IB) literatures (Michael & Lewin, 2007), that has only been viewed through the prism of home country's tax rates, institutional constraints, regulatory context, macro-organizational and legislative policies, etc. However, this study examines and contributes to literature by examining both the effects of home country composite risks as well as the component risks on outward investment flow in the global perspectives. Specifically, the study examined the composite risk (a combination of financial, Political and economic risks) and also identify the components of home country's risks which may lead to the FDI escapism phenomenon (Cuervo-Cuzurra et al., 2015; Cuervo-Cazurra & Ramamurti, 2015; Kobrak et al., 2018).

This study contributes to the literature in several ways, Firstly, home country composite risk moderately impacts global investment outflow negatively and significantly. This implies that as the joint effect of political risk, financial risk, and economic risk continue to increase globally, direct investment abroad flow decreases. The combination of these risk components and other unobserved factors maybe the rationale behind the low outward FDI in many countries in developing countries. This finding corroborates with David (2021) assertion that the decrease in both global outward and inward FDI reflects the rise in geopolitical tensions which has led to diversification of investments by multinational enterprises. This implies that the global risk in the form of joint effect of home country's risk components may promotes FDI escapism. Michael & Lewin (2007) paper that outward FDI may be undertaken as a response measure due to misalignment between home country institutional condition and multinational enterprises supports our finding.

Secondly, we found that global political and financial risks have negative impact on direct investment abroad flow, but the negative impacts are high and low respectively. Therefore, firms are likely to diversify investment abroad from risky home market occasioned by political and financial risks but are more likely to response to the global political risk than the financial risk. This investments diversification appears as escape in response to both political and financial risks in home country. These findings are supported by country-specific evidence of China (Wang et al., 2022); Brazil (Aguiar et al., 2012); and U.S. (Tallman, 1988) papers that political risks negatively influence outward FDI. Nevertheless, the finding is contrary to Quer et al. (2012) empirical results which suggests that high political risk in host country does not discourage Chinese multinationals. Regarding whether economic risk facilitates FDI escapism, this study finds that the economics risk component is low and positively sensitive to outflow of investment which supports and facilitates firm's internationalization motive through outward FDI which bring about global economic integration and development.

In essence, this study contribution shows that global financial and political risk may "sand the wheels" of the home country's economy which give rise to escaping FDI, but a low global economic risk greases the home country economy which supports, promotes and facilitates firm's internalization motive through investment outflow across the borders with reverse technology spillovers. This suggests that most countries economics policies implementation such as payment of debts, reduction in interest rate, low inflation rate, increase wage rate for laborer, tax reduction and stability of the local currency, may have paid off. Findings also show that the interaction terms among the components of home country composite risk plays a significant role in explaining investments outflow in home country.

These findings have several substantive implications that are potentially relevant. Firm's risk strategy and management team should not only pay attention to host country risk, but they should also regularly examine and identify home country components risk that can reduce or disrupt production supply chain, by designing and implementing proactive strategy that counters negative risk components on out flow of investment. However, government must engage in political dialogue and reforms aimed at not only making the country stable politically, but also create an enabling environment where business thrive and facilitate investment abroad. More so, firms need to enter an insurance contract to cover business losses due to risks and uncertainty. Despite the high-quality data employed and the use of econometric framework that is robust to endogeneity problems, the results are only within the period under review, 2003–2016. In order word, our inability to get recent country risk data was our major challenge in this study. Risk-investment relationship with mediating economic factor(s) may be examined as future research.

Acknowledgment

The study was implemented in the framework of the Basic Research Program at the National Research University, Higher School of Economics (HSE University, Moscow) in 2022.

Our sincere thanks go to the editor and the three anonymous referees for their very helpful comments and suggestions. For any other error(s), the usual disclaimer applies.

Conflict of interest

The authors declare no conflict of interest.