New Keynesian economics posits that there are imperfections in the information available to economic agents in markets ([56]; [1]). These imperfections reflect information asymmetry, which leads to market failures. For example, in the functioning of the labor market, employers suffer from information asymmetry regarding their employees’ productivity ([52]). Employees reduce their productivity if they are not constantly monitored ([52]). These information problems can reduce worker productivity and, consequently, investment. Similarly, financial markets are imperfect and therefore subject to information asymmetry between lenders and borrowers ([56]). Under these conditions of imperfect information in these markets, moral hazard can limit financial development. Information asymmetry in these markets thus generates risks of loan default, thereby limiting the supply of bank credit ([58]) and investment.

It is difficult to obtain a complete measure of information asymmetry in the economy. Nevertheless, information asymmetry in the credit market is approximated by the resulting credit defaults ([42]; [58]; [13]). The latter used the credit default rate for this purpose. In our research, we focused our attention on information asymmetry in the credit market.

The literature highlights the decisive role of information asymmetry on investment. Indeed, studies have found that information asymmetry has a negative effect on investment ([23]; [59]; [12]; [15]). According to [15], information asymmetry can have a negative effect on investment and consequently affect economic growth. Similarly, [12] argues that asymmetric information (particularly adverse selection) can cause factors other than market fundamentals to influence investment flows to emerging markets. Similarly, [59] have also described the effects of information asymmetry on Foreign Direct Investment (FDI). They find a negative effect of information asymmetry on FDI in five Asian countries. Consequently, information asymmetry is likely to reduce investment in the WAEMU. The idea of a negative relationship between information asymmetry and investment is also present in [23]. [23] argues that a well-organized financial information system plays an important role for both the firm and its partners, and vice versa. In contrast, the work of [29] finds no significant relationship between information asymmetry and investment in Pakistan.

Our approach to governance is grounded in [41] institutional theory. According to North, institutions are constraints and rules of the game—both formal and informal—that serve to monitor and limit individual action. Consequently, governance refers to the rules and norms that frame and regulate behavior within an economy. In this context, poor governance in a country can thus affect investment. Governance encompasses several dimensions. Our study focuses on political governance. We assess this dimension using the variables described in Section 3.2 below.

According to the literature, the quality of governance in a country can affect investment. For example, in situations of political instability, it becomes less attractive for entrepreneurs to invest in long-term projects ([51]; [49]; [17]). Political stability is therefore essential to support financial development and increased investment. Similarly, corruption can hinder the efficient allocation of public funds and economic growth ([38]; [60]). It directly affects the productivity growth, technological progress, and investment ([57]; [30]). Consequently, reducing corruption can promote financial development and increased investment.

Empirical studies have documented the role of governance in investment and its key components. In some of these studies, high-quality institutions have a positive effect on investment ([19]; [11]). Indeed, better institutions have a positive effect on the location of FDI in OECD countries ([11]). Conversely, policy unpredictability, excessive regulation, insufficient respect for property rights, and a lack of government commitment contribute to deterring FDI flows. According to [19], a strong institutional framework is necessary for better investment establishment. More specifically, studies have analyzed the effects of certain governance variables on investment. Indeed, [3] found a significant negative effect of corruption on the growth of corporate investment in transition economies. Using panel data from 46 countries in Asia and the Pacific, [9] also demonstrated that corruption has a negative effect on FDI inflows. Thus, according to him, countries that have implemented reforms and reduced their levels of corruption receive greater FDI inflows. Corruption, therefore, represents an additional cost for investors and increases risks. Furthermore, [64] found a negative relationship between political instability and investment in certain countries. [48], in the case of Pakistan, assert that political instability leads to a decline in investment and slows economic development. [21] study provides evidence of the relationship between political instability and the volatility of private investment growth in Türkiye. Furthermore, [55] demonstrated that judicial strength and the rule of law improve portfolio investment in developing countries.

Generally speaking, research on the determinants of investment has attracted considerable interest in the literature. However, the effects of this information and governance variables on investment have not been sufficiently discussed in the context of the WAEMU.

Based on the previous literature review, the identification of explanatory factors for investment and its key components (private investment, FDI, public investment, etc.) has been the subject of modeling in the literature. However, due to differences in the contexts of these studies, not all of these models can explain investment in the WAEMU. Our model is based on the following specification of the model of investment determinants by [50]:

where Inv measures gross domestic investment, GR is the GDP growth rate; IR is the real interest rate; TR is trade openness; DS is domestic savings; AID is foreign aid; DB is debt service; and DEP is the dependency ratio ([50]).

This specification of the investment model applies to sub-Saharan Africa and thus most closely reflects the general context of developing countries. We adjust this investment model to the context of our study by incorporating variables related to banking development, information asymmetry, and political governance. Based on this relationship, we specify the following basic investment model for the WAEMU:

Inv i t = α i + β 0 GDP i t + β 1 IR i t + β 2 TR i t + β 3 DS i t + β 4 AID i t                   + β 5 DB i t + β 6 INF i t + β 7 BD i t + β 8 Asym i t + β 9 Gov i t + ε i t (Model 1)

where, DB i t is a measure of banking development; INF i t is inflation; Gov i t is a measure of political governance; and Asym i t is information asymmetry.

According to the literature, in developing countries, information asymmetry can also affect investment through the credit market channel (specifically, bank credit). Information asymmetry in the credit market leads to a decrease in the supply of bank credit ([58]; [13]) and, consequently, limits investment. We can therefore specify the following investment model:

Inv i t = α i + β 0 GDP i t + β 1 IR i t + β 2 TR i t + β 3 DS i t + β 4 AID i t + β 5 DB i t                   + β 6 INF i t + β 7 BD i t + β 10 AsymBD i t + β 9 Gov i t + ε i t (Model 2)

where, AsymBD i t is a variable capturing the interactive effect of information asymmetry and banking development on investment.

Table 1. Model variables and data sources.

Source: authors, March 2026.

Except for Benin, for which political risk data is unavailable, our sample covers all other WAEMU countries. These countries are Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, Niger, Senegal, and Togo. Benin, not being a very heterogeneous case regarding the issue of our study, these 7 countries are representative of all UEMOA countries. We defined the study period (2003-2023) based on data availability.

Before estimating the different models of the study, we first examined the stationarity of the different variables of the model. This study of stationarity prevents us from performing spurious regressions. In this context, there are various unit root tests on panel data ([32]; [37]). But within the framework of our study, we opted for the most robust and commonly used unit root tests in the literature, notably the tests of [33], [22], and [20].

We have summarized the results of the unit root tests in Table A1 and Table A2 for levels and first differences, respectively. The results obtained are consistent. They revealed that the variables are integrated of order 1.

Given these unit root test results, we conducted cointegration tests on the different models. These are the cointegration tests of [25] and [44]. Indeed, these tests are very common in studies and provide more robust results. These tests yielded similar results and indicated a long-term relationship between the variables across the different models. Given the size of our sample, we favored the results of [25] cointegration tests. Indeed, [18] showed that the smaller the sample size, the more robust [25] cointegration tests are compared to those of [44]. The results of [25] cointegration tests on the models reject the null hypothesis of non-cointegration of the variables in this model and the associated sub-models (see Table A3 and Table A4). There is cointegration between investment and the explanatory variables in the different models and sub-models.

The specified investment model uses macroeconomic variables. Some of these variables may therefore be jointly determined. Endogeneity issues may thus arise. Consequently, using the ordinary least squares (OLS) method may lead to biased results. Given this consideration, an appropriate estimation of our investment model requires the use of an appropriate estimator. We therefore used the FMOLS estimator ([44]) and the PMG estimator ([46]). Given the structure of our panel (N = 7 and T = 21), we prioritized the results of the FMOLS estimator. Indeed, for small panels where N is smaller than T, FMOLS estimators are highly efficient ([43]).

The coefficient of determination (R²) ranges from 0.77 to 0.80 across the various estimates, indicating that the selected explanatory variables account for 77% to 80% of the variation in the dependent variable. These relatively high values reflect a good fit of the model. As part of the robustness analysis, we adopted an approach that used alternative unit root and cointegration tests on panel data. We therefore conducted alternative unit root tests by [33], [22], and [20] to verify the robustness of the stationarity results. We have summarized the results in Table A1 and Table A2. At the end of the unit root tests, we conducted alternative cointegration tests of [25] and [44]. The results of these tests are consistent regarding the existence of a long-term relationship between the variables of the different models. We have summarized the results of [25] cointegration tests in Table A3 and Table A4. Regarding the robustness analysis, we also used alternative estimation methods for the study’s different models. These are the FMOLS estimator of [44] and the PMG estimator of [46]. The results of these two approaches are consistent regarding the stability of the estimated coefficients. The results of the FMOLS estimates are summarized in Table 2 and Table 3 below.

The previously specified Investment Model 1 was estimated by substituting the various political governance indicators GS, CO, DA, SC, IP, IC, EC, and MP, respectively. In this case, we therefore estimated 8 sub-models: Sub-model 1 is an estimation of investment model 1 with GS as the political governance variable; sub-model 2 with CO; sub-model 3 with DA; sub-model 4 with SC; sub-model 5 with IP; sub-model 6 with IC; sub-model 7 with EC; and sub-model 8 with MP. The results are summarized in Table 2 below.

Except for sub-model 5, the estimated long-run coefficients of the “Asym” variable, capturing the direct effect of information asymmetry on investment, appear significant and negative across the various sub-models. Information asymmetry has a direct negative effect on long-term investment in the WAEMU. This result is in line with our expectations. Indeed, as theory suggests, information asymmetry generates market failures in the credit market. These market failures prevent the full utilization of productive capacity in this market, thereby limiting investment.

The long-term coefficients of the “Gov” variable, which capture the direct effect of the level of governance on investment, are all significant. The coefficients for government stability, corruption, democratic accountability, external conflicts, and military involvement in politics are negative. In the WAEMU, the risk associated with these political governance variables has a negative impact on long-term investment. Indeed, in econometric models, a positive coefficient for each of these variables implies an improvement in the political climate. Conversely, a negative coefficient implies a deterioration. Similar results are found in some of the literature ([21]; [48]; [64]; [35]). According to these studies, the negative effect can be explained by the low level of governance, which fails to stimulate investment. The coefficients for socioeconomic conditions, investment profile, and internal conflicts are positive. This result reflects an improvement in the quality of governance associated with these variables and, consequently, an improvement in investment. An improvement in the quality of these political governance factors stimulates investment in the WAEMU. These results are consistent with some of the literature ([9]; [3]; [19]; [11]; [8]).

About government stability, the results indicate a limited ability on the part of the authorities to carry out their stated programs and a lack of cohesion in government action within the WAEMU. However, political stability is necessary for the establishment of stable investment patterns and, consequently, stable capital accumulation ([21]). Consequently, political instability in the WAEMU generates risks and investment instability. Regarding corruption, the results indicate that it exacerbates resource allocation problems in the WAEMU and, consequently, reduces investment in the region. This finding is consistent with the literature. Indeed, corruption has negative consequences for resource allocation, entrepreneurship, investment, and innovation ([5]). Widespread corruption and informal economies make it difficult for new firms to enter the market ([14]). These barriers to market entry are thought to discourage investment decisions ([2]). However, [63] found a positive coefficient for corruption on FDI inflows in West Africa. According to him, some foreign investors are much more sensitive to cost minimization and would therefore be tempted to invest in countries with a high corruption index to take advantage of tax rents. Regarding democratic accountability, the results indicate the government’s low responsiveness to public concerns and consequently hinder investment in the WAEMU. Regarding external conflicts, an increase in cross-border conflicts, foreign pressure, and war contributes to a deterioration of the political climate and reduces investment in the WAEMU. Regarding the military’s role in politics, greater military involvement in politics can lead to reduced democratic accountability, distortion of government policy, changes in government policy, and the emergence of an armed opposition. Consequently, it may limit investment in the WAEMU. Regarding socioeconomic conditions, improvements in consumer confidence, employment levels, and living standards encourage investment in these countries. Regarding the investment environment, improvements in contract enforceability, the repatriation of funds, and the reduction of payment delays in the countries concerned stimulate investment in these countries. Regarding internal conflicts, the associated risks are low in the WAEMU. Consequently, an improvement in the quality of this variable (a risk reduction) contributes to increased investment in the WAEMU. By estimating model 1 using the PMG method, we obtain estimated coefficients similar to those of the FMOLS approach, particularly for the variables of interest. For illustrative purposes, when considering sub-model 5 (the model with the investment profile as a variable of political governance), the long-term coefficient of the variable “Asym” appears significant and negative with the PMG method. This coefficient is equal to −0.109146. Similarly, the coefficient of the variable “IP” is significant and positive. This coefficient is equal to 0.628757.

Table 2. Summary of the results of the estimates for Investment Model 1.

The values in parentheses are p-values; *** significant at 1%, ** significant at 5%, *significant at 10%. Source: authors’ estimates.

The estimates of Investment Model 2 aim to analyze the indirect effects of information asymmetry on investment via the banking development channel. We estimate this model by cross-linking banking development and information asymmetry. The results are summarized in Table 3 below.

Except for sub-model 5, the long-run coefficients of the variables measuring the cross-effect of information asymmetry and banking development on investment are significant and negative. Banking development variables, however, appear positive and significant. The negative coefficients of the cross-variables stem from information asymmetry. For example, considering sub-model 1 (see Table 3 below), the estimated coefficient of “LogBD” β 7 is equal to 0.123244, and the estimated coefficient of “LogAsym” β 8 is equal to −0.025037. However, the estimated coefficient of “LogAsym*BD” β 10 is equal to −0.025037. The marginal effect of “LogAsym” ∆Asym is therefore, equal to ( β 8 − β 10 * β 7 ) . We fund ∆Asym = −0.021951. Information asymmetry thus has a negative indirect effect on investment in the WAEMU. It can therefore affect investment through the banking development channel. It would limit bank financing in the WAEMU and, consequently, reduce the investment.

Table 3. Summary of the results of the estimates for Investment Model 2.

Source: authors’ estimates.

The estimated coefficients for the level of economic activity, domestic savings, foreign aid, inflation, and the interest rate appear significant and positive in the various estimates. This result is consistent with the literature. Indeed, [40] found a positive effect of economic growth on investment in 68 low- and middle-income developing countries. Furthermore, [50] found a positive effect of economic growth, domestic savings, trade openness, and changes in aid on gross investment behavior in a panel of 97 developing countries.

The estimated coefficients for the banking development variables are significant and positive. This result is consistent with the literature. Indeed, financial development appears to be a factor in boosting investment in G7 countries ([61]). Similarly, bank credit to the private sector has a positive effect on investment in five developing countries ([62]). The results regarding interest rates and banking development are particularly consistent with the neoliberal approach to investment proposed by [39] and [53]. Unlike neoclassical theory, this investment approach emphasizes the importance of financial deepening and high interest rates in stimulating economic growth. Regarding the interest rate, Keynesian economic theory reveals a negative relationship between the interest rate and investment ([26]). When the interest rate rises, the cost of credit increases, and consequently, firms invest less. In the case of our study, due to the lack of long or continuous series on the interest rate, we approximated it by the interest ratio. In principle, a positive coefficient of this ratio implies a positive effect of the interest rate on investment and is therefore contrary to theory. But in a context of information asymmetry, such as the case of WAEMU, high interest rates do not systematically limit investment. Investors are willing to take out bank loans even at high costs. [31] also found a positive effect of the interest rate on FDI across 20 developing countries in South, East, and Southeast Asia. The debt service coefficients are negative and are consistent with the literature ([50]).