The International Capital Flows and Domestic Savings–domestic Investment Nexus: A Comparative Evidence Between Heterogeneous Developing Regions

Abstract

Drawing inspiration from Feldstein and Horioka’s (1980) (FH) puzzle, our study elucidates the impact of remittances and Foreign Direct Investment (FDI) on domestic savings and investment in two disparate yet globalized developing regions: Latin America and the Caribbean and South Asia. Utilizing an extensive dataset spanning from 1984 to 2021 and employing diverse methodologies, including Dynamic System generalized method of moment, Driscoll–Kraay standard error, fully modified ordinary least squares, and dynamic ordinary least squares, our findings reveal that remittances exert a positive influence on both domestic investment and savings across both regions. However, South Asia predominantly directs remittance inflows towards investment, while Latin America and the Caribbean exhibit a propensity towards saving these funds. As for FDI, the primary developing region predominantly channels these funds into investment, whereas the lower region prioritizes savings. The impact of control variables manifests varied effects across both regions. Ultimately, our study underscores the pivotal role of foreign remittances in supporting investment and savings, underscoring the profound influence of economic growth on these dynamics. This accentuates the imperative for governments to proactively allocate financial resources to optimize economic growth and fortify financial frameworks. Moreover, focused strategies are indispensable for adeptly managing foreign inflows while navigating external shocks such as international repayments, external debt, and aid. Additionally, enhancements in monetary and fiscal policies are imperative to sustain competitive interest rates and foster stable macroeconomic conditions, thereby fostering conducive environments for both public and private domestic savings.

JEL Classification: F24; F3; P33; C23; O18

Keywords

Remittance foreign capital flows domestic savings domestic investment panel model developing regions

Introduction

The escalating influx of foreign capital in various forms, such as foreign direct investment (FDI), foreign portfolio investment (FPI), cross-border bank transactions, official development assistance (ODA), and, more recently, remittances, has emerged as a subject of significant contention within the ongoing developmental discourse in the developing countries. Notably, Ndikumana and Verick (2007) and Macias and Massa (2009) have underscored the mounting evidence indicating a surge in the diverse categories of capital flows since the 1990s. These authors have drawn attention to the fact that FDI inflows witnessed a notable 15% increase between 2004 and 2007. A similar trend can be observed in portfolio equity and bond flows, which experienced a remarkable upswing of over 340% during the same period. Concurrently, the activity in cross-border bank flows registered substantial growth (Macias & Massa, 2009). However, despite this upsurge in capital flows, Ndikumana and Verick (2007) have emphasized that their effects on investment, economic growth, and poverty amelioration remain subjects of considerable apprehension.

Theoretical economic principles suggest that capital tends to move from low-yielding to high-yielding countries as financial integration expands until returns equalize. Research exploring the relationship between domestic savings, investment, and foreign capital flows has been pivotal in understanding global finance dynamics. Feldstein and Horioka’s (1980) pioneering work focused on the link between domestic savings and investment, expecting a weaker connection as capital mobility increased. Contrary to predictions, they found a significant correlation, known as the ‘Feldstein-Horioka puzzle’, persisting mainly in advanced economies. Obstfeld and Rogoff (2000) termed this puzzle and identified it as a major enigma in international macroeconomics. Shibata and Shintani (1998) critiqued the FH model for its lack of a robust theoretical basis and proposed a theoretical framework based on Campbell and Mankiw’s (1989, 1991) permanent income approach. Their study provided support for capital mobility in several countries. Despite this, the FH model’s simplicity, empirical strength, and broader explanatory capacity make it a more preferred framework than the Permanent Income Approach.

The discourse surrounding the interplay of financial inflows with domestic investment and savings is a topic of substantial interest and investigation. Several empirical studies have delved into the impact of remittances and other financial flows on domestic investment and savings (Kapingura, 2018; Lautier & Moreaub, 2012; Osili, 2007). Investment is a pivotal driver of economic growth, necessitating adequate financing. The synergy between international capital flows and domestic savings is crucial to furnishing the requisite capital for investment. In the existing body of literature, we encounter two predominant perspectives. One camp contends that foreign capital flows hold the potential to influence development finance (Harrison et al., 2004), while another group is engrossed in scrutinizing the ramifications of remittances on the investment–savings relationship and overall economic growth (Baldé, 2011 for 37 and 34 Sub-Saharan African countries).

The debate surrounding the interplay of financial flows, domestic investment, and domestic savings is a prominent and relatively underexplored subject in economic research. While some empirical investigations have delved into the potential impact of remittances and other financial inflows on domestic investment and savings, the relationship remains multifaceted (Abbas, 2019; Dash, 2020; Dzansi, 2013; Hossain, 2014; Kapingura, 2018). Existing studies have critically examined the connection between domestic investment, savings, and remittances, with diverse perspectives emerging. (Baldé, 2011 for 37 and 34 countries of Sub-Saharan Africa; Boon, 2017 for 5 ASEAN countries; Das & Serieux, 2010 for 36 developing countries; Hossain, 2014 for 63 developing countries; Zhu et al., 2012 for rural China). However, these studies do not take into account foreign financial flows on the savings–investment nexus, nor do they consider the top and bottom globalized developing regions. Nonetheless, it is worth noting that certain studies have also pointed out the negative repercussions of remittances, such as increased demand for domestic currency potentially weakening the economy by affecting exports and imports (Prasad et al., 2007). Although the majority of the literature concentrates on the impact of remittances on domestic savings and investment, this study expands the existing research landscape by investigating whether the promotion of foreign financial flows, including remittances and FDI, equally or differentially influences domestic savings and investment in both top and bottom globalized developing economies.

This extension of our research represents a significant and valuable addition to the existing body of literature. Our comparative study exclusively focuses on developing countries, which confront a host of complex challenges such as poverty, inequality, sluggish economic growth, increasing unemployment, and insufficient economic development. Notably, these countries witnessed a 7% decline in remittance flows in 2020, further plummeting by 7.5% in 2021¹. This substantial drop in remittances raises substantial concerns for developing countries, prompting us to concentrate on these regions. The study by Sinha (2017) further showed the importance of remittance flows, mainly in developing countries. However, to the best of our knowledge, our study is the first comparative investigation encompassing the top and bottom globalized developing regions, examining the interplay between foreign financial flows and domestic savings and investment within a panel framework from 1984 to 2021. Our analysis takes into account various control variables, including economic growth, government expenditure, inflation rates, and trade, which have been selected based on previous research emphasizing their significance in the context of domestic savings and investment, as identified by studies such as Fischer (1993), Greene and Villanueva (1991), Haque et al. (2021), Kapingura (2018), and Wai and Wong (1982).

The World Bank’s initial projections indicated a consistent growth in remittances up to the year 2019. However, in 2020, the world witnessed an abrupt and unprecedented decline of approximately 20% in remittances, primarily due to the global pandemic. This decline had far- reaching consequences, affecting all countries, but its impact was particularly severe for those nations heavily reliant on remittances. Reduced remittance inflows were associated with increased poverty levels and higher unemployment rates. These, in turn, hindered long-term economic growth and financial development, as previously highlighted in studies by Aggarwal et al. (2011), Chowdhury (2016), and Rao and Hassan (2012). Figure 1 provides a visual representation of the significant sources of external financial inflows. It illustrates that FDI is the most significant source, followed by remittances, portfolio equity, and ODA. Given this context, our study aims to empirically examine the two most substantial financial flows: FDI and remittances².

Figure 1.

Source: World Development Indicators.

The selection of these two diverse developing countries is underpinned by the correlation between domestic savings and investment, as Feldstein and Horioka’s (1980) puzzle proposed. As illustrated in Figure 2, the correlation between domestic savings and investments is high for the South Asian developing region, categorizing it as the bottom globalized developing region, while it is low for the Latin America and Caribbean developing region, designating it as the top globalized developing region. Additionally, the study by Edwards (2007) found that countries with more open capital markets tend to have a lower correlation between domestic savings and investment. Another study by Rajan and Zingales (2003) also found that the decline in the correlation between domestic savings and investment in developing countries is closely associated with increased capital inflows to these countries over the past few decades. These existing studies also support and strengthen the above-stated classification.

Figure 2.

Source: World Development Indicators.

Another basis for choosing these two regions is the growth rate of remittances. According to the World Bank data, remittances are declining by 14% in 2021 compared to the pre-COVID-19 year, that is, 2019. According to The World Bank, the dropping regions are Europe and Central Asia (9.7%), East Asia and the Pacific (7.9%), and Sub-Saharan Africa (12.5%). But interestingly, some regions have found growth in remittance flow, like Latin America and the Caribbean (6.5%), South Asia (5.2%), and the Middle East and North Africa (2.3%). Therefore, Latin America and the Caribbean have the highest growth rate of remittance, followed by South Asia, the second-highest region (Figure 3). Finally, the following reasons enable us to choose Latin America and the Caribbean developing region as the top globalized developing region and South Asia as the bottom globalized developing region.

Figure 3.

Source: World Development Indicators.

Our investigation spans a panel of 25 countries based on the high and low correlation between domestic savings and investment within developing regions from 1984 to 2021. Our approach employs a panel two-step Dynamic System-Generalized Method of Moment, supplemented by a robustness check using the DKSE, fully modified ordinary least squares (FMOLS), and DOLS estimator. This research offers a triple-fold contribution to the existing literature. First, it distinguishes itself by conducting a comparative analysis of the relationship between foreign capital flows (remittances and FDI) and the domestic savings–investment nexus within the top and bottom globalized developing regions. As far as our knowledge goes, this is the first study to introduce such a comparative analysis. Second, the study employs the Feldstein and Horioka (1980) hypothesis to categorize the top and bottom globalized developing regions based on the correlation between domestic savings and investment, which will assist in implementing regional-specific policies. Third, this research highlights that foreign capital inflows play a vital role in bridging the financial gaps in developing nations. They aid in funding current account and budget deficits while stimulating domestic investment. Notably, in developing countries with high capital mobility, the connection between foreign capital flows and the domestic savings–investment dynamics is more robust compared to those with lower mobility. This emphasizes the importance of attracting foreign capital, which is achievable through policies that enhance the investment climate and reduce trade barriers. Policymakers should leverage these inflows to finance productive investments, fostering economic growth and development in these countries.

This article is structured based on the following sections: the part second offers a critical evaluation of previous research, and the part third presents the theoretical foundation. Data sources and methodology are detailed in the part fourth. The parts 5 and 6 present a summary of the econometric results along with explanations and policy implications, followed by a concluding section.

Related Literature Survey

The FH hypothesis and the Twin Gap Hypothesis are two crucial concepts in international finance that help to understand the relationship between domestic savings, domestic investment, and capital flows. The Feldstein and Horioka (1980) hypothesis (FH) initially explored the interplay between domestic savings and investment in OECD countries to gauge capital mobility. It suggested that a high correlation indicated limited mobility, emphasizing the role of domestic savings in domestic investment. Conversely, low correlation implied substantial foreign capital involvement. Obstfeld and Rogoff (2000) coined it the ‘mother of all puzzles’ due to its intriguing nature. In addition, the Twin Gap Hypothesis expands on this, delving into savings–investment imbalances’ impact on capital flows and recognizing the complexities of global finance dynamics. Collectively, these theories illuminate capital mobility nuances, shaping our understanding of international finance, resource allocation, and the interplay between savings, investment, and global capital movements.

However, the literature concerning the FH puzzle can be categorized into two main strands. In one strand, a group of researchers argues that the correlation between domestic saving and investment serves as a measure of the FH puzzle, but they acknowledge that this approach introduces biases into the model, particularly from a policy perspective. Scholars such as Bajo Rubio and López Pueyo (1998), Coakley et al. (2004), Gundlach and Sinn (1992), Jansen (1996), Jansen and Schulze (1996), Mastroyiannis (2007), Özmen and Parmaksız (2003), and Sarno and Taylor (1998) fall into this category. In contrast, the second strand of researchers does not support the FH hypothesis. They contend that the measurement of capital mobility in the FH puzzle is flawed due to factors unrelated to capital mobility. To address these concerns, some studies in this category have introduced new factors into the FH puzzle framework. For instance, Raheem et al. (2016) brought the role of governance into the equation, Adeniyi and Egwaikhide (2013) included financial development, and Fouquau et al. (2008) incorporated economic growth, trade openness, and current account balance as additional variables in the FH hypothesis. Özmen (2007) extended the FH hypothesis by considering the exchange rate regime, while Ho (2003) augmented it by incorporating the country’s size as a contributing factor to capture the nuances of the FH puzzle better.

This study is partially a part of both strands. It focuses on the impacts of domestic investment and savings, which supports the first strand, and on the other hand, for the second strand, we have taken remittances and FDI as the primary factors, though we are accepting the concept. This study also added economic growth, government expenditure, inflation rate, and trade as additional factors of the domestic savings and investment function (Haque et al., 2021; Kapingura, 2018; Kokorović Jukan et al., 2020).

However, the positive impact of remittance on investment is a much-applauded topic in the macroeconomics literature, though most of the literature did not accommodate both the savings–investment nexus. As for instance, according to Gani (2016), remittance enhances domestic savings. On the other hand, the positive relationship between remittance and domestic investment is supported by the study of Dash (2020) for six South Asian countries, meaning remittances contribute to immediate consumption and foster investment activities, including the development of human and physical capital. As per the opinion of Dzansi (2013), remittance with good institutional quality and a well-developed financial sector can improve domestic investment. Additionally, remittance has a significant impact on household income (Edwards & Ureta, 2003), poverty (Gupta et al., 2009), rising savings rate (Sahoo & Dash, 2013), more globalized and financially developed economy (Pal & Kumar Mahalik, 2022). However, remittance can also reduce domestic investment if it is spent more on consumption rather than investment (Abbas, 2019 for Pakistan). Migrants generally transferred their additional income to their families in the form of remittances. Some earlier literature indicates that remittance amounts mostly spent on basic consumption like education, health, and household expenses (Airola, 2007; Gyimah-Brempong & Asiedu, 2015; Kangmennaang et al., 2018; Lipton, 1980; Osili, 2004). If migrants are temporary, they preferably save more, but permanent migrants send remittances for future backup. Their families invest their money in land, housing, the banking sector (Osili, 2007; Sinning, 2011), etc. This concept is again supported by Ahmed (2012), as he states that those migrants who are staying abroad for a long time they spend their money on savings and investments except for regular expenses. Therefore, the above discussion proved that the gap between domestic savings and investment is a vital barrier to economic growth, where foreign capital flows are expanding the gap (Chenery, 1967).

However, from the earlier literature, we found evidence that remittances have a mixed impact on domestic savings and investment. Therefore, conducting extensive research on the impact of remittances on both the top and bottom developing regions will provide valuable insights for policymaking in these regions, offering guidance on how to harness the potential benefits and mitigate potential challenges associated with remittance flows to enhance economic growth and stability. Based on the above discussion, we hypothesize that:

H₁: Remittances and domestic savings have a significant adverse effect on domestic investment.

H₃: Remittances and domestic investment have a significant adverse effect on domestic savings.

Other studies are examining the question of whether FDI flows can promote domestic investment and savings. On this note, Hossain (2014) found an effect of FDI flows on domestic savings and investment for 63 developing countries. Abu and Karim (2016) and Dhar and Roy (1996) support the relationship between FDI and domestic investment and savings. Ndikumana and Verick (2007) point out the impact of FDI on domestic investment. They also recommend implementing the measures by which investment will flourish more with higher economic growth. Another study by Amadou (2011) highlighted the channels through which foreign capital influences investment. He added that FDI and loans are the vital and primary channels for extreme utilization of domestic investment; this result is again analysed by Bosworth and Collins (1999) and finds the same insights.

However, a wide range of studies, emphasizing Western countries and Asia in particular, have examined the relationship between foreign capital flows, domestic savings, and investment via theoretical arguments and empirical investigations. Numerous studies have examined the influence of foreign capital on investment through the savings channel, including those by Zhu et al. (2009), Das and Serieux (2010), Baldé (2011), and Hossain (2014). There are conflicting results, but some indicate that foreign money replaces domestic savings, increasing consumption and decreasing investment. Using the Common Correlated Effects mean group model, Hossain (2014) discovered, for instance, that foreign aid and remittances negatively impacted domestic savings, with remittances showing a more pronounced crowding-out effect. The statistical insignificance of the effects of FDI and portfolio inflows supports the idea that large capital inflows have the potential to replace domestic savings, which might impede domestic investment. These results are consistent with those of Kumar (2007). However, research like Athukorala and Sen (1995) for India revealed that remittances had a negligible negative impact on savings. On the other hand, Osili (2007) discovered that remittances had a favourable effect on investment, savings, and economic growth. Additionally, research conducted in Sub-Saharan Africa by Ndikumana and Verick (2007) showed that FDI encouraged domestic investment, highlighting the need to enhance the domestic investment climate. Similarly, Amadou (2011) study on Togo highlighted the positive influence of FDI and loans on domestic investment, while FPI had a negligible impact, demonstrating that the type of foreign capital matters. However, the relationship between foreign capital and investment remains relatively unexplored in the Southern African Development Community (SADC) region, with most studies focusing on economic growth or savings, making this study a valuable contribution to the understanding of how different types of foreign capital affect domestic investment in the SADC region (Mugowo, 2017). From the related discussion about FDI with domestic investment and savings, we hypothesize that:

H₂: FDI and domestic savings have a significant adverse effect on domestic investment.

H₄: FDI and domestic investment have a significant adverse effect on domestic savings.

However, the combined theoretical and empirical relationship between savings–investment and foreign flows (remittances and FDI) is mixed in nature. Following this research gap, this study intends to clarify the relationship between domestic savings–investment with remittance and FDI in the top and bottom of globalized developing regions.

Theoretical Framework

Pagano (1993) introduced the concept of the foreign capital flows- savings investment nexus. Bailliu (2000) tried to bring that sort of thinking into his study by injecting domestic savings into his model. He clarified that a strong domestic financial sector could help international financial flows promote economic growth in developing countries. The study by Agosin and Machado (2005) states that foreign capital helps enhance domestic and foreign investment.

I_{t} = D I_{i t} + F I f t

(1)

where I_t is an overall investment; DI_it, and FI_ft are the domestic and foreign investment, respectively. This equation shows that investment includes both domestic and international investment. Whither domestic investment is the difference between predicted and actual capital stock. So, the model is described as

D I_{i, t} = σ (Y^{*}_{i t} - Y_{i t})

(2)

where Y^*_it and Y_it are the predicted and actual capital stock by domestic investors, respectively. Here σ is always greater than 1, and it works here as an adjusting coefficient. However, the predicted capital stock (Y^*_it) is a mixture of the expected growth rate and the difference between actual and full capacity output.

Y^{*}_{i t} = α_{0} + α_{1} G_{t} + α_{2} k_{t}

(3)

The value of α₀ and α₁ is always greater than 0. If we can assume that capital stock depreciates at a rate of (1−ρ) per period, then investment will be

D I_{i t} = Y_{i t + 1} - (1 - ρ) Y_{i t}

(4)

We are injecting the domestic savings (DS_it) into the model to know its importance in the investment process. In this perspective, Bailliu (2000) highlights the role of financial intermediaries in the saving–investment nexus and finds that savings generate less than investment. So, we can assume that the amount of savings (δ) is ready for investment, and the rest of the value (1–δ) will be taken by the financial intermediaries. Therefore, the amount of savings will equalize with investment after taking off the value of financial intermediaries.

δ D S_{i t} = D I_{i t}

(5)

The literature confirms that foreign capital can enhance the savings rate for investment as foreign investors invest through financial intermediaries. Then, the new capital market equilibrium will be

δ^{*} (D S_{i t} + F F_{t}) = D I_{i t}^{*}

(6)

where FF_t is the foreign capital flows. On this note, Bailliu (2000) points out two possibilities by which foreign capital can enhance domestic investment. First, foreign capital must spend more on domestic and foreign investment than consumption. And second, there is a need to choose the productive investment sources where more competition, profit, and export-oriented activities are present.

In this context, the FH hypothesis and twin gap hypothesis, it is essential to recognize that domestic investment, in conjunction with foreign capital flows, can also influence domestic savings. Therefore, this dynamic relationship signifies the intricate interplay between domestic savings, investment, and foreign capital flows, underscoring the need for a comprehensive understanding of these factors in shaping economic outcomes. It would be helpful for policymakers in developing countries to focus on policies that promote domestic savings and investment and attract foreign capital flows. This will help countries to close the twin gaps and achieve sustainable economic growth and development.

Data and Methodology

Data Sources

To address the outlined objectives, this study utilizes annual data spanning from 1984 to 2021, a timeframe selected primarily due to data availability. The factors that contribute to the dynamics of the domestic savings–investment relationship include personal remittances received (in current US dollars) (Abbas, 2019; Benhamou & Cassin, 2021; Dash, 2020; Hossain, 2014), net inflows of foreign direct investment (BoP, in current US dollars) (Dhar & Roy, 1996; Lautier & Moreaub, 2012; Ndikumana & Verick, 2008; Prasanna, 2010). The control variables are GDP per capita (in current US dollars), general government final consumption expenditure (expressed as a percentage of GDP), consumer price inflation (measured annually), and trade as a percentage of GDP. These variables were chosen based on both theoretical considerations and the practicality of data availability. The dependent variables of this study encompass domestic investment, quantified as gross fixed capital formation as a percentage of GDP, and domestic savings, represented as gross domestic savings in current US dollars. All data for both dependent and independent variables were sourced from the World Bank’s World Development Indicators. All the data are converted into a natural logarithm except inflation and FDI³ to avoid heterogeneous error and minimal outliers in the series. The empirical log-linear models are as follows (Model 1–Model 4):

\begin{array}{l} Model 1 : {lnDI}_{i, t} = β_{0} + β_{1} {lnDS}_{i, t} + β_{2} {lnREM}_{i, t} + β_{3} {lnEG}_{i, t} + \\ \begin{matrix} \begin{matrix} \begin{matrix}  \end{matrix} \end{matrix} \end{matrix} β_{4} {lnGE}_{i, t} + β_{5} {INF}_{i, t} + β_{6} {lnTRADE}_{i, t} + ε_{it} \end{array}

(7)

\begin{array}{l} Model 2 : {lnDI}_{i, t} = β_{0} + β_{1} {lnDS}_{i, t} + β_{2} {FDI}_{i, t} + β_{3} {lnEG}_{i, t} + \\ \begin{matrix} \begin{matrix} \begin{matrix}  \end{matrix} \end{matrix} \end{matrix} β_{4} {lnGE}_{i, t} + β_{5} {INF}_{i, t} + β_{6} {lnTRADE}_{i, t} + ε_{it} \end{array}

(8)

\begin{array}{l} Model 3 : {lnDS}_{i, t} = β_{0} + β_{1} {lnDI}_{i, t} + β_{2} {lnREM}_{i, t} + β_{3} {lnEG}_{i, t} + \\ \begin{matrix} \begin{matrix}  \end{matrix} \end{matrix} β_{4} {lnGE}_{i, t} + β_{5} {INF}_{i, t} + β_{6} {lnTRADE}_{i, t} + ε it \end{array}

(9)

\begin{array}{l} Model 4 : {lnDS}_{i, t} = β_{0} + β_{1} {lnDI}_{i, t} + β_{2} {FDI}_{i, t} + β_{3} {lnEG}_{i, t} + \\ \begin{matrix} \begin{matrix}  \end{matrix} \end{matrix} β_{4} {lnGE}_{i, t} + β_{5} {INF}_{i, t} + β_{6} {lnTRADE}_{i, t} + ε it \end{array}

(10)

where lnDI and lnDS are the log of domestic investment and savings, respectively; lnREM is the log of remittance flow; FDI is the foreign direct investment; lnEG is the log of economic growth; lnGE is the log of government expenditure; INF is the inflation; lnTRADE is the log of trade openness, and ε is the error term.

Methodology

Test of Stationarity

For measuring the order of integration of the variables, this study used Levin et al. (2002) and Im et al. (2003) unit root test. These two first- generation unit root tests are used for robustness. Because we are employing cross-country data series to analyse balanced panel data, there may be specific crucial difficulties associated with cross-sectional dependence among countries. That is, if any country experiences a shock, it will affect the entire region. To remove these difficulties from the data series, this study uses a cross-sectional dependence test and slope homogeneity test. The test like Breusch and Pagan (1980), Pesaran (2015, 2021) as well as the slope homogeneity test by Pesaran and Yamagata (2008) and Blomquist and Westerlund (2013), is used for checking the cross-sectional dependency and slope homogeneity. The result evidenced the cross-sectional dependence; therefore, in this analysis, we employ the cross-sectional augmented IPS (CIPS) test as introduced by Pesaran (2007) to examine unit root presence. The CIPS test differs by integrating cross-sectional data, expanding the IPS test statistic with added components considering cross-sectional interdependence. This method is preferred over the IPS test of Im et al. (2003), which lacks explicit handling of cross-sectional dependencies, potentially resulting in biased outcomes.

Cointegration Test

This study also uses the cointegration test by Pedroni (2004) to find out the existence of a long-run relationship between the data series. Also, we have added Westerlund (2007) cointegration test for robustness checking.

Dynamic System Generalized Method of Moments

The two-step dynamic system-generalized method of moments (GMM) developed by Blundell and Bond (1998) was employed in the present study to estimate the Model 1–Model 4. Building upon the first difference GMM, it is less effective because of its high autoregressive parameters, finite sample properties, and sensitivity to unobserved heterogeneity. Lagged dependent variables are used as instruments by Blundell and Bond (1998) to capture unobserved individual-specific traits. Rather than using within-group or traditional panel ordinary least squares (OLS) estimations, the System GMM is utilized because the latter approaches frequently produce estimates that are skewed and inconsistent due to their inability to handle dynamic panel bias and potential endogeneity. In dynamic panel data models, Within groups addresses unobserved country-specific effects within specified time intervals, whereas OLS levels ignore unobservable, time-invariant country effects (Arellano & Bover, 1995; Blundell & Bond, 1998, 2000). Approximate upper and lower limits are assigned to estimates obtained from Within Groups estimators and OLS levels, respectively (Bond et al., 2001; Hoeffler, 2002). However, when heteroscedasticity and autocorrelation within individuals are expected, or when independent variables are not strictly exogenous and exhibit correlation with previous and present error realizations, the System GMM provides accurate and efficient parameter estimates. It successfully addresses the endogeneity issue by instrumenting the lagged dependent variable and other endogenous factors with variables presumed to be uncorrelated with fixed effects (Nickell, 1981). Surprisingly, the System GMM indicates greater accuracy than the difference GMM estimator when the series closely approximates random walks.

Driscoll–Kraay Robust Standard Errors (DKSE)

To estimate long-run coefficients in pooled OLS regression, Driscoll and Kraay developed the Driscoll–Kraay robust standard error estimator (Driscoll & Kraay, 1998). The standard errors of the pooled OLS models are adjusted using this estimator to account for any cross-sectional or temporal dependencies. It combines a set of moment condition cross-sectional averages with the Newey–West adjustment. This dynamic non-parametric approach examines the linear connection between panel data without being constrained by the number of panel cross-sections, which makes it particularly helpful as the time dimension increases.

There are several benefits to using the Driscoll–Kraay standard error approach. First of all, it successfully tackles heteroscedasticity and cross-sectional dependency, two crucial problems in panel data analysis. It also gets around the drawbacks of estimating large constant covariance matrices. This method can also deal with missing values, which means it may be used in panel data setups with both balanced and unbalanced data. As previously noted by Heberle and Sattarhoff (2017) and Pei et al. (2018), panel data exhibiting heteroscedasticity, spatial dependencies, and serial dependencies are subject to the use of the Driscoll–Kraay standard error approach. This emphasizes the reason why we decided to use the Driscoll–Kraay standard error method for our study. The linear model using the pooled OLS Driscoll–Kraay standard errors is formulated as follows:

y_{i t} = {x^{'}}_{i t} β + ε_{i t}

(11)

In this equation, the explanatory variables are designated as x_it, while the dependent variable is represented by y_it.

FMOLS and DOLS Estimator

Finally, this study uses the FMOLS by Phillips and Hansen (1990) to measure the long-run analysis. The main advantage is that it can employ those variables with different orders of integration. This model can also use enough lags for the data-generating process in specific modeling. Using FMOLS can effectively tackle the challenges of endogeneity and autocorrelation in the analysis, resulting in precise and reliable estimates. This method, in particular, is adopted at yielding accurate estimations.

We again use dynamic OLS (DOLS) by Stock and Watson (1993) for robustness checking the long-run analysis among the dependent and explanatory variables. This method is superior to all of the different orders of integration, like I (0) and I (1) (Bashier & Siam, 2014). There is a chance to use the semi-parametric method to analyse this model’s long-run parameters. It can work well for small sample sizes by removing the serial correlation issues and biases in the variable and heterogeneity errors (Agbola, 2013; Bashier & Siam, 2014). Another advantage is that the DOLS model also eliminates the correlation between regressors and error terms by using past and first differenced regressor values.⁴ This approach is known for its asymptotic efficiency as an estimator, and it excels in minimizing feedback issues within the cointegration framework. By incorporating these methodologies, the study ensures the effective management of endogeneity and autocorrelation concerns, ultimately contributing to the robustness of the analysis.

Empirical Results and Related Discussions

To check the stationarity properties of the variables, we examined through the Levin et al. (2002)’s, Im et al. (2003)’s and CIPS unit root test (Table A1). From the result of the unit root, we observed that some variables are stationary at level, and some are stationary at first difference. Therefore, the orders of integration of the variables are mixed, like I(0) and I(1). Log of remittance flows and domestic savings have the stationarity at first difference. Except for these two variables, others are stationary at level for both the developing regions. These three-unit root test gives a consistent result for all the variables of the two developing regions.

However, stationarity is necessary for the next cointegration test. Our next step in the analysis is the cointegration test. We use Pedroni (2004) and Westerlund’s (2007) panel cointegration tests to check whether the long-run relationship between the variables exists in these two developing regions. The results of Pedroni and Westerlund’s panel cointegration tests are reported in Table A2 and Table A3, respectively.

The results of Tables A2 and A3 confirm a long-run relationship between the variables for both developing regions. Therefore, the null hypothesis of no cointegration is not accepted, which means the variables have cointegration among each other. As we previously discussed that we are working on balanced panel data using the cross-country data series; there may be some critical issues related to cross-sectional dependency among the countries. For that, we use the cross-sectional dependency and slope homogeneity tests. The results of Table A4 indicate the presence of cross-sectional dependence and slope homogeneity among the countries for both the developing regions.

Long-run Result

After confirming the cointegration among the variables, it is necessary to check the long-run analysis between the variables for top and bottom developing regions. The results of the Dynamic system GMM estimators are reported in Table 1. The long-run results of Model 1 show that domestic savings and remittance enhance domestic investment in the top (Latin America & Caribbean) and bottom (South Asia) developing regions, whereas inflation (bottom) and trade openness (top) deteriorate it. However, government expenditure does not support domestic investment and savings in the bottom developing region, while in Latin America and Caribbean, government expenditure decreases (increases) domestic investment (domestic savings). The exciting part is that the impact of remittance on domestic investment is higher in the bottom region than in the top globalized developing regions. However, our results are identical as the computed results are supported by the study of Abbas (2019) and Dash (2020) for South Asia. These studies have found that remittance affects domestic investment in the long run. As per Baldé (2011), remittance indirectly impacts growth through investment. Therefore, foreign capital makes a borderline on domestic capital formation by imposing an investment rate; after exciding the limit, it will transfer to the domestic savings (Mohamed & Mohamed, 2003). The control variables summarize that economic growth enhances domestic investment, which is an accurate observation as growth increases the capacity to invest more because of the enhancement of money in the hands of people. Trade openness is positive in the case of the bottom region (Umer & Alam, 2015 for Pakistan), which means a more open market makes the market more competitive with more investment. Though, it is harmful in the top globalized developing region because the people may not be interested in investing much in this region. Inflation has positive impacts in Latin America and the Caribbean but negatively affects the bottom developing region. When top developing regions face a higher inflation rate due to diminishing real monetary balances, they invest more to earn more return. Government expenditure negatively impacts domestic investment in both regions. Increased expenditure leads to less investment.

Table 1.

Dynamic System Panel-data Estimation, Two-Step GMM.

Variables	Model 1: lnDI= f (lnDS, lnREM, lnEG, lnGE, INF, lnTRADE)		Model 2: lnDI= f (lnDS, FDI, lnEG, lnGE, INF, lnTRADE)		Model 3: lnDS= f (lnDI, lnREM, lnEG, lnGE, INF, lnTRADE)		Model 4: lnDS= f (lnDI, FDI, lnEG, lnGE, INF, lnTRADE)
Variables	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia
lnDI					0.498***	0.258**	0.481***	0.481***
lnDS	0.606***	0.049**	0.530***	0.420***
lnREM	0.075*	0.151***			0.074**	0.196**
FDI			0.057**	–0.754**			–0.002*	0.002*
lnEG	0.480***	0.167*	0.509***	0.238***	0.985***	1.270***	1.030***	1.030***
lnGE	–0.427***	–0.385***	–0.161**	–0.554**	0.821***	–0.765***	0.725***	–0.725***
INF	0.024**	–0.048**	0.051***	–0.247***	0.013*	0.131***	0.025**	0.025**
lnTRADE	–0.486**	0.647***	–0.789***	0.654***	0.751***	–1.281***	1.020***	–1.020***
Arellano-Bond test for AR (1) in first differences	1.457**	1.890*	1.160**	2.540**	2.260**	1.670*	2.240**	2.240**
Arellano-Bond test for AR (2) in first differences	4.600	5.900	3.800	3.846	2.700	5.200	2.000	2.843
Hansen test	61.100	62.270	65.400	66.457	38.820	10.570	39.080	39.124

Source: Authors’ estimation.

Note: Hansen test refers to the over-identification test for the restrictions in the dynamic system-GMM estimation. The AR (1) and AR (2) tests are the Arellano–Bond tests for the first and second-order autocorrelation in first differences. *p < 0.10, **p < 0.05, ***p < 0.01.

The long-run result of Model 2 shows the impact of FDI on domestic investment with the other control variables. FDI positively impacts Latin America and the Caribbean, but it negatively affects bottom developing regions. Like remittance, FDI is also a source of investment for the top developing region; the study of Mohamed and Mohamed (2003) found the similar result. Whereas domestic saving is an essential source of domestic investment, this result is supported by the study of Baldé (2011) and Kapingura (2018). This concept is also supported by the Harrod–Domar model, where they find that savings significantly impact investment, which ultimately enhances economic growth. The results of all the control variables are consistent with the result of Model 1.

Finally, the long-run result of Model 3 (domestic savings function) shows that remittance positively affects domestic savings in both regions. However, the coefficient of remittance is higher in the top developing region than in the bottom developing region, which narrates that the bottom region is not interested in saving their remittances; instead, Latin America and the Caribbean are saving more by using remittances. From the control variables, Economic growth, domestic investment, and inflation positively impact both regions. In South Asia, government expenditure and trade adversely affect domestic savings, but they are beneficial for the top developing regions. Economic growth enhances the competitive nature of the market and profitability, leading to increased capacity to spend more on savings. On the other hand, inflation is positively affecting, which means a constant enhancement of inflation makes the fear about more expenditure; that is why saving is one of the savior options for both regions. In the top developing region, the government is not using domestic savings for their consumption and investment, so it is not hampering domestic savings.

The empirical results of Model 4 show the relation between FDI and domestic savings in both regions. Other control variables are affecting domestic savings, same as Model 3. Here, FDI negatively impacts the top developing region (Benhamou & Cassin, 2021; Hossain, 2014) but positively affects the bottom region. The possible reasons behind the antagonistic relationship between domestic savings and FDI are stated here. First, most of the FDI is used for consumption; second, the demonstration effect can enhance consumption. The positive effect of FDI is evident in the South Asia, as the financial flows from FDI are mainly used for savings.

Nonetheless, we have once more employed the DKSE, FMOLS, and DOLS estimators to assess the long-term result’s stability; these estimators provide identical outcomes to those of the Dynamic System GMM. Tables 2, 3, and 4 are representing the findings of the study. From the analysis above, we have found six main findings: first, remittances enhance domestic investment and savings in the two regions. Second, the combined result states that domestic investment is an important determinant of domestic savings and vice-versa. Third, remittance impacted domestic investment more in the bottom developing region rather than the top region. Oppositely, remittance positively affects domestic savings more in the top developing regions. Fourth, the FDI flows are used on investment, not in the savings in the top developing region, but it is the opposite in the bottom region, as here savings are more by using FDI. Fifth, economic growth positively impacts throughout all the models for both regions. Finally, other control variables are mixed in values. For better understanding, we have also added the graphical representation of the long-run results of domestic investment and domestic savings functions in Figure 4, Figure 5, Figure 6, and Figure 7, respectively.

Table 2.

Driscoll-Kraay Robust Standard Errors (DKSE).

Variables	Model 1: lnDI= f (lnDS, lnREM, lnEG, lnGE, INF, lnTRADE)		Model 2: lnDI= f (lnDS, FDI, lnEG, lnGE, INF, lnTRADE)		Model 3: lnDS= f (lnDI, lnREM, lnEG, lnGE, INF, lnTRADE)		Model 4: lnDS= f (lnDI, FDI, lnEG, lnGE, INF, lnTRADE)
	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia
	Coefficient
lnDI					3.265***	1.459***	3.662***	3.542***
lnDS	0.190***	0.122**	0.190***	0.171***
lnREM	0.063**	0.049**			0.647***	0.778***
FDI			0.063**	–0.065**			–0.051*	0.052*
lnEG	0.028**	0.139***	0.028**	0.812**	0.213**	0.174**	0.556**	0.654**
lnGE	–0.062*	–0.357*	–0.062*	–0.572*	0.456*	–1.025**	0.658**	–0.384**
INF	0.014*	–0.006*	0.014*	–0.124*	0.284***	0.153*	0.210**	0.521**
lnTRADE	–0.168**	0.391***	–0.168*	0.821*	0.227*	–0.347**	2.922***	–0.524***

Source: Authors’ estimation.

Note: *, **, and *** indicate the 1%, 5%, and 10% levels of significance, respectively.

Table 3.

FMOLS Estimators.

Variables	Model 1: lnDI= f (lnDS, lnREM, lnEG, lnGE, INF, lnTRADE)		Model 2: lnDI= f (lnDS, FDI, lnEG, lnGE, INF, lnTRADE)		Model 3: lnDS= f (lnDI, lnREM, lnEG, lnGE, INF, lnTRADE)		Model 4: lnDS= f (lnDI, FDI, lnEG, lnGE, INF, lnTRADE)
	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia
	Coefficient
lnDI					0.286*	0.095*	0.771*	2.859*
lnDS	0.138*	0.038*	0.144*	0.09*
lnREM	0.024*	0.038*			0.173*	0.089*
FDI			0.000**	–0.000*			–4.45E–12**	6.84E–12***
lnEG	0.003*	0.172*	0.008*	0.086*	0.042*	0.253*	0.699*	0.433**
lnGE	–0.119*	–0.463*	–0.244*	–0.204*	0.946*	–0.802*	0.164***	–1.593***
INF	0.000*	–0.005*	0.000**	–0.003*	0.000*	0.001*	2.65E–05***	0.025*
lnTRADE	–0.005*	0.075*	–0.018*	0.004*	0.145*	–0.031*	0.927*	–0.011***
R² (OVERALL)	0.701	0.880	0.705	0.905	0.977	0.976	0.902	0.888
Adj R²	0.676	0.862	0.682	0.890	0.971	0.973	0.891	0.883

Source: Authors’ estimation.

Note: *, **, and *** indicate the 1%, 5%, and 10% levels of significance, respectively.

Table 4.

DOLS Estimators.

Variables	Model 1: lnDI= f (lnDS, lnREM, lnEG, lnGE, INF, lnTRADE)		Model 2: lnDI= f (lnDS, FDI, lnEG, lnGE, INF, lnTRADE)		Model 3: lnDS= f (lnDI, lnREM, lnEG, lnGE, INF, lnTRADE)		Model 4: lnDS= f (lnDI, FDI, lnEG, lnGE, INF, lnTRADE)
	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia	Latin America & Caribbean	South Asia
	Coefficient
lnDI					0.213*	1.707*	1.225*	0.248*
lnDS	0.178*	0.103*	0.528*	0.106*
lnREM	0.025*	0.026*			0.413*	0.327*
FDI			0.000*	–0.000**			–4.55E–12*	4.98E–12*
lnEG	–0.000*	–0.010**	–0.206*	–0.015*	0.379*	1.398*	0.011*	0.935*
lnGE	–0.235*	–0.074**	–0.925*	0.072**	0.525*	–5.091*	0.097*	–0.779*
INF	0.000*	–0.000***	–0.000*	–0.001*	0.000*	0.019*	–7.03E–05*	0.018*
lnTRADE	–0.062*	0.013***	–0.337*	0.015**	0.482*	–0.666*	0.063*	–0.492*
R² (OVERALL)	0.900	0.923	0.843	0.922	0.976	0.987	0.992	0.994
Adj R²	0.765	0.915	0.578	0.913	0.947	0.964	0.981	0.983

Source: Authors’ estimation.

Note: *, **, and *** indicate the 1%, 5%, and 10% levels of significance, respectively.

Figure 4.

Note: Authors own calculation.

Figure 5.

Note: Authors own calculation.

Figure 6.

Note: Authors own calculation.

Figure 7.

Note: Authors own calculation.

Discussions of the Results

The observations regarding the interplay between domestic savings and domestic investment in Latin America, the Caribbean, and South Asia align with the FH hypothesis. This economic theory posits that in regions or countries with high domestic savings, there is a greater capacity for domestic investment. The positive and mutually reinforcing association among domestic savings and investment in these two regions reflects the essence of the FH hypothesis. In Latin America and the Caribbean, where there is evidence of increased domestic savings stimulating higher domestic investment, the FH hypothesis finds support. The established financial systems and economic structures in this region allow for more efficient capital mobilization, thus facilitating investment. According to the World Bank, the average domestic savings rate in Latin America and the Caribbean increased from 20.8% in 2010 to 23.4% in 2021. The average domestic investment rate increased from 21.4% to 22.7% during the same period. This suggests that there is a positive relationship between domestic savings and investment in the region. Similarly, with its traditionally high savings rates in South Asia, the availability of domestic savings serves as a robust foundation for increased domestic investment. This situation aligns with the FH hypothesis, which posits that regions or countries with a propensity for savings can channel those savings into productive investments. According to the World Bank, South Asia’s average domestic savings rate increased from 27.4% in 2010 to 30.6% in 2021. The average domestic investment rate increased from 26.5% to 28.9% during the same period, which also suggests a positive relationship between them.

Regarding remittances, the FH hypothesis is indirectly supported by the positive correlation between remittances and both domestic savings and investment in these regions. Remittances act as an additional source of income for households, leading to increased savings and potentially contributing to higher investment, aligning with the essence of the FH hypothesis. Remittances can essentially serve as a form of savings that can be later invested in various ventures. According to the World Bank, remittances to Latin America and the Caribbean reached US$135 billion in 2021, while remittances to South Asia reached US$163 billion in the same year. These significant inflows of remittances contribute to the overall savings and investment potential in these regions.

However, the effect of FDI on domestic investment and savings demonstrates that the relationship can vary based on regional dynamics. In Latin America and the Caribbean, where FDI positively influences investment but negatively impacts savings, therefore, the FH hypothesis’s direct applicability is less clear. However, this suggests that while FDI may enhance investment, it might also encourage higher consumption, reducing savings. According to the World Bank, FDI inflows to Latin America and the Caribbean reached US$163 billion in 2022, accounting for 2.9% of the region’s GDP, whereas the average domestic savings rate decreased from 20.8% in 2010 to 19.5% in 2022. This decline in domestic savings rates coincides with the increase in FDI inflows to the region. In South Asia, where FDI negatively influences investment but positively affects savings, the FH hypothesis faces a different context. Here, FDI might compete with domestic investment interests or lead to resource allocation constraints. However, it appears to promote capital retention, thus encouraging higher savings. According to the World Bank, FDI inflows to South Asia reached US$87 billion in 2021, accounting for 1.4% of the region’s GDP. While FDI has been relatively low in South Asia compared to Latin America and the Caribbean, it has been associated with increased domestic savings rates.

Moreover, the overall findings of the study have important implications for the twin gap hypothesis, which states that a country with a current account deficit must also have a budget deficit or vice versa. The study suggests that foreign capital flows, such as remittances and FDI, can help to close the twin gaps in developing countries. This is because foreign capital flows can finance current account deficits and budget deficits, and they can also boost domestic investment. However, the study also suggests that the relationship between foreign capital flows and the domestic savings–investment nexus can vary depending on the regional context. For example, in Latin America and the Caribbean, FDI may encourage higher consumption, reducing savings. This could potentially put upward pressure on the current account deficit, exacerbating the twin gaps. Therefore, in creating policies to support economic growth and development, policymakers in developing countries should give serious consideration to these results. Overall, the findings demonstrate remittances as a more reliable and direct source of foreign borrowings and transfers than FDI, which can support domestic savings and investment in developing countries. This is due to the fact that remittances are more likely to be invested in small enterprises, healthcare, and education, among other beneficial endeavours. However, FDI may be utilized for a number of things, such as capital flight, large-scale project investment, and consumption.

The control variables summarize that economic growth enhances domestic investment and savings, which is an accurate observation as growth increases the capacity to invest more because of the enhancement of money in the hands of people. In 2022, South Asia experienced an average real GDP growth rate of 5.8%, while Latin America and the Caribbean recorded a growth rate of 3.7%, both witnessing concurrent rises in domestic investment and savings. The Inter-American Development Bank’s (IDB) 2023 study on Latin America and the Caribbean confirmed the positive association among economic growth and domestic investment and savings. The research highlighted that a 1% rise in real GDP growth corresponded to a 0.25% increase in domestic investment and a 0.18% increase in domestic savings in the region. Similarly, the Asian Development Bank’s (ADB) 2023 study focusing on South Asia revealed a comparable positive effect of economic growth on domestic investment and savings (Dutta et al., 2017). For this region, a 1% rise in real GDP growth translated to a 0.32% increase in domestic investment and a 0.21% increase in domestic savings.

However, government expenditure does not support domestic investment and savings in South Asia, while in Latin America and Caribbean, government expenditure decreases (increases) domestic investment (domestic savings). The World Bank claims that South Asia’s government spending as a share of GDP has recently decreased, falling from 25.4% in 2010 to 22.7% in 2021. Domestic investment as a proportion of GDP fell from 26.5% to 28.9% during the same time period. This implies that domestic investment in South Asia can be negatively impacted by government spending. On the other hand, in Latin America and the Caribbean, government spending as a proportion of GDP has risen recently, rising from 20.8% in 2010 to 23.4% in 2021. In the same time frame, domestic investment fell from 21.4% to 22.7% as a proportion of GDP, while domestic savings rose from 20.8% to 23.4%. This indicates that government spending in Latin America and the Caribbean may have a detrimental impact on domestic investment but an advantageous impact on domestic savings. The disparity in the impact of government expenditure on domestic investment and savings between South Asia and Latin America and the Caribbean likely stems from varying fiscal policies and economic structures. In South Asia, government spending patterns might not be efficiently directed toward initiatives that directly foster investment or incentivize savings, potentially due to challenges in prioritizing productive sectors or inadequacies in public investment management. Conversely, in Latin America and the Caribbean, the fluctuation in government expenditure might be more strategically targeted, either through infrastructure projects or incentivizing savings programmes, thus exerting a measurable influence on both domestic investment and savings. The divergent outcomes highlight the varying effectiveness of fiscal policies and their alignment with economic development goals across these regions.

The results also suggest that inflation may be playing a role in increasing domestic investment and savings in the South Asia and Latin America and the Caribbean. According to the World Bank, inflation rates in South Asia and Latin America and the Caribbean have risen in recent years. South Asia’s average inflation rate increased from 3.6% in 2010 to 5.8% in 2022, whereas in Latin America and the Caribbean, it increased from 4.7% in 2010 to 7.2% in 2022. Domestic savings rates increased in both regions at the exact same time. The average domestic savings rate increased from 27.4% in 2010 to 30.6% in 2022. In Latin America and the Caribbean, the average domestic investment rate increased from 21.4% in 2010 to 22.7% in 2022, while the average domestic savings rate increased from 20.8% in 2010 to 23.4% in 2022. Inflation’s influence on domestic savings in South Asia and, Latin America and the Caribbean could be attributed to different factors. In South Asia, higher inflation rates might spur increased savings as individuals and businesses seek to hedge against rising prices. This could lead to heightened savings to preserve purchasing power. In contrast, in Latin America and the Caribbean, inflation might prompt a similar response, driving individuals and entities to invest more domestically or save to counter the erosion of their wealth caused by rising prices. Additionally, high inflation can also stimulate demand for financial instruments that provide returns surpassing inflation rates, thus contributing to increased savings and investment within these regions.

In Latin America and the Caribbean, trade contributes to a decline in domestic investment while bolstering domestic savings. Conversely, in South Asia, trade appears to diminish domestic savings while amplifying domestic investment. According to the World Bank, in Latin America and the Caribbean, the average trade openness ratio surged from 78.1% in 2010 to 82.4% in 2022. Concurrently, the average domestic investment rate slightly declined from 21.4% in 2010 to 22.7% in 2022, while the average domestic savings rate increased from 20.8% in 2010 to 23.4% in 2022. Projections for 2023 suggest a sustained high trade openness ratio at 82.2%, with an anticipated domestic investment rate of 22.5% and a domestic savings rate of 23.3%. In South Asia, the World Bank notes a rise in the average trade openness ratio from 40.9% in 2010 to 47.2% in 2022. Simultaneously, the average domestic investment rate increased from 26.5% in 2010 to 28.9% in 2022, while the average domestic savings rate declined from 27.4% in 2010 to 30.6% in 2022. Projections for 2023 indicate a further increase in the trade openness ratio to 48.1%, alongside an estimated domestic investment rate of 28.4% and a domestic savings rate of 30.2%. However, trade seems to affect Latin America and the Caribbean and South Asia differently. In Latin America and the Caribbean, trade could hinder domestic investment due to heavy reliance on imports, limiting funds available for investment. Yet, it might encourage savings through income generation opportunities. Conversely, in South Asia, trade might draw resources from domestic savings due to import reliance, possibly reducing overall savings (Gnangnon, 2020). However, increased trade could stimulate more domestic investment by offering access to new markets and resources. These impacts depend on trade policies, import-export dynamics, and regional economic structures.

Conclusion and Policy Implications

This study examines the impact of remittance and FDI on the domestic investment–savings nexus from 1984 to 2021. It is a comparative study between the top (Latin America and Caribbean developing region) and bottom (South Asia developing region) globalized developing regions based on Feldstein and Horioka (1980) hypothesis. It employed dynamic systems GMM, DKSE, FMOLS, and DOLS for the long-run result. The empirical result revealed that remittances enhance domestic investment and savings in both regions. However, the South Asia region is more interested in investment by using remittance, and oppositely, Latin America the Caribbean prefer to save their remittance amount more than investment. In the case of FDI, money is mainly used on investment, not in the savings in the top developing region, but it is entirely opposite in the bottom region, as here savings are more by using FDI. The other control variables are mixed in sign in both regions. From the combined result, we can confirm that the foreign remittance flows finance investment and savings. Economic growth is also an essential factor of investment and savings.

The research underscores the significance of macroeconomic stability in attracting foreign capital. Remittances and FDI complement each other, urging effective utilization of these flows. Policy focusses on driving growth, structured economy, targeted investment, and savings policies is crucial. Domestic support to manage external shocks and enhance monetary and fiscal policies can optimize benefits, promoting public and private savings. Specifically, policy recommendations for Latin America, the Caribbean, and South Asia involve incentivizing remittances and FDI, investing in education and healthcare for higher labor productivity, and trimming government expenditure to boost resources for investment. Promoting domestic savings and monitoring inflation is vital. Encouraging remittance use for investments via tax incentives can fuel economic growth. Creating a stable investment climate and easing bureaucratic hurdles can attract FDI, fostering job creation and growth. Prioritizing growth-oriented policies, where a 1% investment increase relates to 0.4% growth, underscores the importance of fostering an environment supporting investment, innovation, and human capital development for economic progress.

Careful consideration of trade liberalization’s impact on domestic investment is crucial, balancing increased investment with heightened foreign competition. Implementing trade policies aligning with domestic investment goals is vital. Effective inflation management, a 1% rise leading to 0.5% decreased growth, safeguards investment and economic development. Efficient government spending, not crowding out private investment, is crucial. Ensuring government expenditure enhances, rather than hinders, private investment fosters a conducive environment for economic growth in developing countries.

In terms of addressing the twin deficit problem demands effective fiscal consolidation and trade strategies, attracting foreign capital inflows and creating a stable investment climate also play crucial roles. By encouraging FDI and remittances, implementing growth-oriented policies, and fostering a conducive environment for investment, developing countries can ultimately promote economic growth and generate the resources necessary to tackle the twin deficit challenges. However, the ‘twin deficit’ problem (simultaneous budget and current account deficits) persists despite progress in some regions. In Latin America and the Caribbean (LAC), while Brazil (current account surplus in 2022), Panama (surplus since 2019), and Costa Rica (reduced deficit from 8.8% in 2009 to 2.2% in 2021) showcase success, Argentina (high debt and inflation), Ecuador (deficit exceeding 5% in 2023), and Honduras (weak tax collection and high spending) face challenges. Similarly, South Asia witnessed mixed progress, with India (reduced fiscal deficit from 6.7% in 2018 to 6.1% in 2023) and Sri Lanka (surplus in 2018) achieving reductions, while Pakistan (widening deficit to 4.6% in 2023) and Bangladesh (grappling with deficit amidst growth) struggle. Global shocks, fragile recoveries, and policy choices like attracting foreign investment (needing regulatory reforms and transparency) contribute to the complexities. Smaller economies like Maldives, Bhutan, and Nepal exhibit unique deficit scenarios. Global shocks, fragile recoveries post-pandemic, structural issues like inequality, corruption, and policy choices such as fiscal discipline and attracting foreign investment, all contribute to the complexities.

Addressing the twin deficit problem in LAC and South Asian countries demands tailored policies, acknowledging the diversity of economic contexts. Fiscal consolidation strategies entail prudent spending cuts, tax reforms, and debt management. Export diversification and infrastructure investments bolster trade resilience and competitiveness. Attracting foreign investment necessitates regulatory reforms, targeted promotion, and transparent administrative processes. Macroeconomic stability hinges on exchange rate stability, effective monetary policies, and prudent debt management. Specific examples include Argentina’s fiscal rebalancing, Brazil’s infrastructure investment, and Mexico’s security enhancements. In South Asia, India focuses on fiscal reforms, Bangladesh enhances garment industry efficiency, and Pakistan prioritizes fiscal discipline and infrastructure development.

Considering current trends and challenges offers insights into potential directions for LAC and South Asian countries. Trends suggest increased investment in technology, emphasis on environmental sustainability, regional cooperation, and bolstering social safety nets. Addressing rising debt, demographic shifts, geopolitical uncertainties, and technological disruptions will be paramount. In LAC, tackling political instability, corruption, and inequality alongside infrastructure development and export diversification are key. South Asia may focus on poverty reduction, education, healthcare, and urbanization management while also addressing political instability and resource constraints. However, these are speculative directions, subject to political, economic, and social dynamics, emphasizing the need for adaptability and effective implementation.

Footnotes

Availability of Data and Materials

Available upon reasonable request from the corresponding author.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.

Appendix

Latin America and Caribbean developing region (14): Argentina, Belize, Bolivia, Brazil, Colombia, Costa Rica, Dominican Republic, Ecuador, Honduras, Mexico, Nicaragua, Panama, Peru, Paraguay.

South Asia developing region (7): Bangladesh, Bhutan, India, Sri Lanka, Maldives, Nepal, and Pakistan.

Notes

References

Abbas

(2019). Workers’ remittances and domestic investment in South Asia: A comparative econometric inquiry. International Migration, 57(5), 89–104.

Abu

, & Karim

M. Z. A.

(2016). The relationships between foreign direct investment, domestic savings, domestic investment, and economic growth: The case of Sub-Saharan Africa. Society and Economy, 38(2), 193–217.

Adeniyi

, & Egwaikhide

F. O.

(2013). Saving-investment nexus in developing countries: does financial development matter? Journal of Economic Development, 38(2), 119.

Agbola

F. W.

(2013). Does human capital constrain the impact of foreign direct investment and remittances on economic growth in Ghana? Applied Economics, 45(19), 2853–2862.

Aggarwal

, Demirgüç-Kunt

, & Pería

M. S. M.

(2011). Do remittances promote financial development? Journal of Development Economics, 96(2), 255–264.

Agosin

M. R.

, & Machado

(2005). Foreign investment in developing countries: Does it crowd in domestic investment? Oxford Development Studies, 33(2), 149–162.

Ahmed

F. Z.

(2012). The perils of unearned foreign income: Aid, remittances, and government survival. American Political Science Review, 106(1), 146–165.

Airola

(2007). The use of remittance income in Mexico. International Migration Review, 41(4), 850–859.

Amadou

(2011). The effect of foreign capital on domestic investment in Togo. International Journal of Economics and Finance, 3(5), 223–226.

10.

Arellano

, & Bover

(1995). Another look at the instrumental variable estimation of error-components models. Journal of Econometrics, 68(1), 29–51.

11.

Athukorala

, & Sen

(1995). Economic reforms and rate of saving in India. Economic & Political Weekly, 2184–2190.

12.

Bailliu

(2000). Private capital flows, financial development, and economic growth in developing countries (No. 2000–15). Bank of Canada.

13.

Bajo

Rubio, Ó.

, & López Pueyo

(1998). An industry analysis of foreign direct investment in Spanish manufacturing, 1986–1992.

14.

Baldé

(2011). The impact of remittances and foreign aid on savings/investment in Sub-Saharan Africa. African Development Review, 23(2), 247–262.

15.

Bashier

A. A.

, & Siam

A. J.

(2014). Immigration and economic growth in Jordan: FMOLS approach. International Journal of Humanities Social Sciences and Education (IJHSSE), 1(9), 85–92.

16.

Benhamou

Z. A.

, & Cassin

(2021). The impact of remittances on savings, capital and economic growth in small emerging countries. Economic Modelling, 94, 789–803.

17.

Blomquist

, & Westerlund

(2013). Testing slope homogeneity in large panels with serial correlation. Economics Letters, 121(3), 374–378.

18.

Blundell

, & Bond

(1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115–143.

19.

Blundell

, & Bond

(2000). GMM estimation with persistent panel data: An application to production functions. Econometric Reviews, 19(3), 321–340.

20.

Bond

S. R.

, Hoeffler

, & Temple

J. R.

(2001). GMM estimation of empirical growth models. Available at SSRN 290522.

21.

Boon

T. H.

(2017). Saving, investment and capital flows: An empirical study on the ASEAN economies. In ASEAN in an Interdependent World (pp. 45–57). Routledge.

22.

Bosworth

, & Collins

S. M.

(1999). Capital inflows, investment and growth. Tokyo Club Papers, 12(11), 55–74.

23.

Breusch

T. S.

, & Pagan

A. R.

(1980). The Lagrange multiplier test and its applications to model specification in econometrics. The Review of Economic Studies, 47(1), 239–253.

24.

Campbell

J. Y.

, & Mankiw

N. G.

(1989). Consumption, income, and interest rates: Reinterpreting the time series evidence. NBER Macroeconomics Annual, 4, 185–216.

25.

Campbell

J. Y.

, & Mankiw

N. G.

(1991). The response of consumption to income: A cross-country investigation. European Economic Review, 35(4), 723–756.

26.

Chenery

H. B.

(1967). Foreign assistance and economic development. In Capital movements and economic development (pp. 268–292). Palgrave Macmillan UK.

27.

Chowdhury

(2016). Financial development, remittances and economic growth: Evidence using a dynamic panel estimation. Margin: The Journal of Applied Economic Research, 10(1), 35–54.

28.

Coakley

, Fuertes

A. M.

, & Spagnolo

(2004). Is the Feldstein–Horioka puzzle history? The Manchester School, 72(5), 569–590.

29.

Das

, & Serieux

(2010). Remittances and reverse flows in developing countries. IDEAs Working Paper Series.

30.

Dash

R. K.

(2020). Impact of remittances on domestic investment: A panel study of six south Asian countries. South Asia Economic Journal, 21(1), 7–30.

31.

Dhar

, & Roy

S. S.

(1996). Foreign direct investment and domestic savings-investment behaviour: Developing countries’ experience. Economic & Political Weekly, 2547–2551.

32.

Driscoll

J. C.

, & Kraay

A. C.

(1998). Consistent covariance matrix estimation with spatially dependent panel data. Review of Economics and Statistics, 80(4), 549–560.

33.

Dutta

C. B.

, Haider

M. Z.

, & Das

D. K.

(2017). Dynamics of economic growth, investment and trade openness: Evidence from Bangladesh. South Asian Journal of Macroeconomics and Public Finance, 6(1), 82–104.

34.

Dzansi

(2013). Impact of remittances on domestic investment: the role of institutional and financial development. Global Review of Accounting and Finance, 4(2), 65–84.

35.

Edwards

A. C.

, & Ureta

(2003). International migration, remittances, and schooling: evidence from El Salvador. Journal of Development Economics, 72(2), 429–461.

36.

Edwards

(2007). Capital controls, sudden stops, and current account reversals. In capital controls and capital flows in Emerging Economies: policies, practices, and consequences (pp. 73–120). University of Chicago Press.

37.

Feldstein

, & Horioka

(1980). Domestic saving and international capital flows. The Economic Journal, 90(358), 314–329.

38.

Fischer

(1993). The role of macroeconomic factors in growth. Journal of Monetary Economics, 32(3), 485–512.

39.

Fouquau

, Hurlin

, & Rabaud

(2008). The Feldstein–Horioka puzzle: A panel smooth transition regression approach. Economic Modelling, 25(2), 284–299.

40.

Gani

(2016). Remittances and savings in Asia: Some empirical evidence based on the life-cycle model. Journal of Finance and Economics, 4(1), 24–38.

41.

Gnangnon

S. K.

(2020). Trade openness and diversification of external financial flows for development: An empirical analysis. South Asian Journal of Macroeconomics and Public Finance, 9(1), 22–57.

42.

Greene

, & Villanueva

(1991). Private investment in developing countries: An empirical analysis. Staff Papers, 38(1), 33–58.

43.

Gundlach

, & Sinn

(1992). Unit root tests of the current account balance: Implications for international capital mobility. Applied Economics, 24(6), 617–625.

44.

Gupta

, Pattillo

C. A.

, & Wagh

(2009). Effect of remittances on poverty and financial development in Sub-Saharan Africa. World Development, 37(1), 104–115.

45.

Gyimah-Brempong

, & Asiedu

(2015). Remittances and investment in education: Evidence from Ghana. The Journal of Iinternational Trade and Economic Ddevelopment, 24(2), 173–200.

46.

Haque

M. A. U.

, Kibria

M. G.

, & Selim

M. M. I.

(2021). Effects of foreign aid and remittances flows on saving and investment in developing South Asia: Panel data study. International Journal of Economics and Financial Research, 7(2), 21–27.

47.

Harrison

A. E.

, Love

, & McMillan

M. S.

(2004). Global capital flows and financing constraints. Journal of Development Economics, 75(1), 269–301.

48.

Heberle

, & Sattarhoff

(2017). A fast algorithm for the computation of HAC covariance matrix estimators. Econometrics, 5(1), 9.

49.

T. W.

(2003). The saving-retention coefficient and country-size: The Feldstein–Horioka puzzle reconsidered. Journal of Macroeconomics, 25(3), 387–396.

50.

Hoeffler

(2002). The augmented Solow model and the African growth debate. Available at SSRN 313705.

51.

Hossain

(2014). Differential impacts of foreign capital and remittance inflows on domestic savings in developing countries: A dynamic heterogeneous panel analysis. Economic Record, 90, 102–126.

52.

K. S.

, Pesaran

M. H.

, & Shin

(2003). Testing for unit roots in heterogeneous panels. Journal of Econometrics, 115(1), 53–74.

53.

Jansen

W. J.

(1996). The Feldstein-Horioka test of international capital mobility: Is it feasible? International Monetary Fund.

54.

Jansen

W. J.

, & Schulze

G. G.

(1996). Theory-based measurement of the saving-investment correlation with an application to Norway. Economic Inquiry, 34(1), 116–132.

55.

Kangmennaang

, Bezner-Kerr

, & Luginaah

(2018). Impact of migration and remittances on household welfare among rural households in Northern and Central Malawi. Migration and Development, 7(1), 55–71.

56.

Kapingura

F. M.

(2018). Relationship between foreign capital flows, domestic investment and savings in the SADC region. Development Southern Africa, 35(4), 554–568.

57.

Kokorović Jukan

, Okičić

, & Hopić

(2020). Remittances as an opportunity to increase savings and financial inclusion of youth in South East Europe. Economic research-Ekonomska istraživanja, 33(1), 2606–2619.

58.

Kumar

(2007). Does foreign direct investment help emerging economies? Economic Letter, 2, 11.

59.

Lautier

, & Moreaub

(2012). Domestic investment and FDI in developing countries: the missing link. Journal of Economic Development, 37(3), 1.

60.

Levin

, Lin

C. F.

, & Chu

C. S. J.

(2002). Unit root tests in panel data: asymptotic and finite-sample properties. Journal of Econometrics, 108(1), 1–24.

61.

Lipton

(1980). Migration from rural areas of poor countries: the impact on rural productivity and income distribution. World Development, 8(1), 1–24.

62.

Macias

J. B.

, & Massa

(2009). The global financial crisis and sub-Saharan Africa (Working Paper, No. 304, pp. 1–24).

63.

Mastroyiannis

(2007). Current account dynamics and the Feldstein and Horioka puzzle: The case of Greece. The European Journal of Comparative Economics, 4(1), 91–99.

64.

Mohamed

M. A. W.

(2003). The impact of foreign capital inflow on savings, investment and economic growth rate in Egypt: An econometric analysis. Scientific Journal of King Faisal University, 4(1), 279–308.

65.

Mugowo

(2017). Foreign direct investment and economic growth in SADC countries: A panel data analysis [Doctoral dissertation].

66.

Ndikumana

, & Verick

(2007). The linkages between FDI and domestic investment: Unravelling the developmental impact of foreign investment (Working Paper, No. 2007–13).

67.

Ndikumana

, & Verick

(2008). The linkages between FDI and domestic investment: Unravelling the developmental impact of foreign investment in Sub-Saharan Africa. Development Policy Review, 26(6), 713–726.

68.

Nickell

(1981). Biases in dynamic models with fixed effects. Econometrica, 49(6), 1417–1426.

69.

Obstfeld

, & Rogoff

(2000). The six major puzzles in international macroeconomics: is there a common cause? NBER Macroeconomics Annual, 15, 339–390.

70.

Osili

U. O.

(2004). Migrants and housing investments: Theory and evidence from Nigeria. Economic Development and Cultural Change, 52(4), 821–849.

71.

Osili

U. O.

(2007). Remittances and savings from international migration: Theory and evidence using a matched sample. Journal of Development Economics, 83(2), 446–465.

72.

Özmen

(2007). Financial development, exchange rate regimes and the Feldstein–Horioka puzzle: Evidence from the MENA region. Applied Economics, 39(9), 1133–1138.

73.

Özmen

, & Parmaksız

(2003). Policy regime change and the Feldstein–Horioka puzzle: the UK evidence. Journal of Policy Modeling, 25(2), 137–149.

74.

Pagano

(1993). Financial markets and growth: An overview. European Economic Review, 37(2–3), 613–622.

75.

Pal

, & Kumar Mahalik

(2022). Factors driving financial development in top and bottom globalized developing economies: Does economic globalization matter? Journal of Public Affairs, 22(1), e2292.

76.

Pedroni

(2004). Panel cointegration: asymptotic and finite sample properties of pooled time series tests with an application to the PPP hypothesis. Econometric Theory, 20(3), 597–625.

77.

Pei

, Huang

, & You

(2018). Nonparametric fixed effects model for panel data with locally stationary regressors. Journal of Econometrics, 202(2), 286–305.

78.

Pesaran

M. H.

(2007). A simple panel unit root test in the presence of cross-section dependence. Journal of Applied Econometrics, 22(2), 265–312.

79.

Pesaran

M. H.

(2015). Testing weak cross-sectional dependence in large panels. Econometric Reviews, 34(6–10), 1089–1117.

80.

Pesaran

M. H.

(2021). General diagnostic tests for cross-sectional dependence in panels. Empirical Economics, 60(1), 13–50.

81.

Pesaran

M. H.

, & Yamagata

(2008). Testing slope homogeneity in large panels. Journal of Econometrics, 142(1), 50–93.

82.

Phillips

P. C.

, & Hansen

B. E.

(1990). Statistical inference in instrumental variables regression with I (1) processes. The Review of Economic Studies, 57(1), 99–125.

83.

Prasad

, Rajan

, & Subramanaian

(2007). The paradox of capital, finance and development. IMF, March, 44.

84.

Prasanna

(2010). Direct and indirect impact of foreign direct investment (FDI) on domestic investment (DI) in India. Journal of Economics, 1(2), 77–83.

85.

Raheem

I. D.

, Ajide

K. B.

, & Adeniyi

(2016). The role of institutions in output growth volatility-financial development nexus: a worldwide study. Journal of Economic Studies, 43(6), 910–927.

86.

Rajan

R. G.

, & Zingales

(2003). The great reversals: the politics of financial development in the twentieth century. Journal of Financial Economics, 69(1), 5–50.

87.

Rao

B. B.

, & Hassan

G. M.

(2012). Are the direct and indirect growth effects of remittances significant? The World Economy, 35(3), 351–372.

88.

Sahoo

, & Dash

R. K.

(2013). Financial sector development and domestic savings in South Asia. Economic Modelling, 33, 388–397.

89.

Sarno

, & Taylor

M. P.

(1998). Exchange controls, international capital flows and saving-investment correlations in the UK: An empirical investigation. Weltwirtschaftliches Archiv, 134(1), 69–98.

90.

Shibata

, & Shintani

(1998). Capital mobility in the world economy: An alternative test. Journal of International Money and Finance, 17(5), 741–756.

91.

Sinha

(2017). International migration and welfare implications. South Asian Journal of Macroeconomics and Public Finance, 6(2), 209–229.

92.

Sinning

M. G.

(2011). Determinants of savings and remittances: Empirical evidence from immigrants to Germany. Review of Economics of the Household, 9, 45–67.

93.

Stock

J. H.

, & Watson

M. W.

(1993). A simple estimator of cointegrating vectors in higher order integrated systems. Econometrica, 61(4), 783–820.

94.

Umer

, & Alam

(2015). Trade oneness, size of economy and the saving– investment relationship: A dynamic analysis for Pakistan. South Asian Journal of Macroeconomics and Public Finance, 4(2), 233–257.

95.

Wai

U. T.

, & Wong

C. H.

(1982). Determinants of private investment in developing countries. The Journal of Development Studies, 19(1), 19–36.

96.

Westerlund

(2007). Testing for error correction in panel data. Oxford Bulletin of Economics and Statistics, 69(6), 709–748.

97.

Zhu

, Wu

, Wang

, Du

, & Cai

(2012). Do migrants really save more? Understanding the impact of remittances on savings in rural China. Journal of Development Studies, 48(5), 654–672.

98.

Zhu

, Wu

, Peng

, & Sheng

(2014). Where did all the remittances go? Understanding the impact of remittances on consumption patterns in rural China. Applied Economics, 46(12), 1312–1322.