Investigating the Role of Financial Development in Encouraging the Transition to Renewable Energy: A Fractional Response Model Approach

Abstract

This paper aims to investigate the relationship between financial development and renewable energy consumption using a fractional response model. The study examines a sample of 34 advanced economies and 64 emerging markets and developing economies from 2008 to 2020. The findings from the fractional response model indicate that financial development has a positive impact on renewable energy consumption in advanced economies. However, in emerging and developing economies, financial development negatively affects the consumption of renewable energy. Additionally, the findings illustrate that financial development has a more pronounced positive impact in advanced economies. This effect is especially strong in countries with higher levels of financial development. On the other hand, in emerging and developing economies, the consumption of renewable energy is more strongly affected by the negative impact of financial development on countries with lower financial development.

1. Introduction

Policymakers around the world view energy supply diversification and sustainability as crucial topics. Since energy is an essential component of the production, distribution and consumption of goods and services, it also serves as a major engine for economic growth and sustainable development [1,2]. Global energy consumption has increased significantly in recent years due to the rapid growth of the population and economies. In 1950, the total energy consumption was 28,654 TWh, but by 2022, it had risen to 178,899 TWh. Fossil fuels still dominate global energy use, accounting for 77% of total energy consumption [3]. One major drawback of relying on fossil fuels is the environmental impact caused by their extensive use, which has led to a significant increase in carbon emissions [4,5]. Another limitation is that traditional fossil fuels are non-renewable and are available in finite amounts. Therefore, increasing energy demand could result in severe energy supply shortages, affecting energy prices. This could significantly limit economic growth and impact social stability and food prices, especially for countries facing energy shortages [3,6,7,8]. In this context, it is crucial for countries to find ways to stabilize and sustain their energy supplies.

In recent years, experts focusing on energy and the environment have stressed the importance of adopting renewable energy sources as a viable solution to address environmental damage and the possibility of energy shortages caused by inadequate supply [6]. In addition to being theoretically infinite, renewable energy can provide sustainable power for economic growth. Renewable energy significantly decreases carbon emissions, in contrast to traditional fossil energy sources, leading to a notable reduction in greenhouse gas emissions and an enhancement in environmental quality.

In fact, as associated technologies advance, the cost of using renewable energy will drop dramatically. Additionally, because of geographical constraints, the reserves of traditional fossil fuels worldwide are unevenly distributed, with the majority concentrated in a few countries. This causes a significant oscillation in international energy prices and poses a potential risk to the energy supply. In contrast, renewable energy comes in various forms and is produced from multiple sources, including solar, wind and geothermal energy. The use of these types of energy sources could overcome the geographical restrictions of fossil fuels. Furthermore, most countries can generate renewable energy from resources that are available locally, which eases issues with the energy supply [2,9,10].

Financial support plays a crucial role in promoting the growth of the renewable energy industry, especially since it is still in its early stages. Significant investments in infrastructure and human capital development are necessary for the industry to thrive. It is also important to provide substantial funds to improve renewable energy technologies, which can help reduce production costs. Governments typically offer financial support to the renewable energy industry during its early stages, but this assistance is often limited and unsustainable. Therefore, additional sources of financing are required to sustain firm growth. Financial development (FD) plays a crucial role in promoting the growth of the renewable energy industry and increasing the demand for renewable energy. Because advanced financial institutions can offer low-cost debt financing for renewable energy projects, a well-functioning stock market can efficiently allocate capital to environmentally friendly sectors, thereby increasing the production and consumption of renewable energy [11,12,13].

In recent years, researchers have shown interest in the relationship between FD and the promotion of renewable energy consumption (REC). Although the theoretical literature suggests that FD positively affects REC, the limited empirical evidence in this regard has produced mixed results (e.g., Khan, Khan and Binh [11]; Raza, Shah, Qureshi, Qaiser, Ali and Ahmed [1]; Assi, et al. [14]; Wang and Dong [15]; Yu, Jin, Zhang and Chong [2]). A notable limitation of previous studies is their failure to consider the fractional nature of REC when measured as a proportion between 0 and 1. Conventional linear models may not effectively capture the impact of FD on REC across the entire range of FD values.

Specifically, if the consumption of renewable energy is influenced by FD, the relationship must be bounded. Otherwise, REC is predicted to lie outside the boundaries (0, 1). The Papke and Wooldridge [16] fractional response model (FRM) provides a robust approach for addressing issues posed by bounded dependent variables. The primary goal of this paper is to evaluate the impact of financial development on real economic activity by employing the Financial Repression Measure. To accomplish this, a panel dataset consisting of 34 advanced economies and 64 emerging markets and developing economies from 2008 to 2020 has been utilized. To our knowledge, this is the first paper to evaluate the impact of FD on REC using the FRM. Therefore, this paper aims to address the following questions: Does FD have a significant impact on REC? Is the impact of FD on REC the same in advanced economies and emerging and developing economies?

The present paper is organized as follows: A brief overview of the relevant literature is provided in Section 2, followed by a description of the data and econometric methodologies in Section 3. Section 4 and Section 5 present the results and conclusions, respectively.

2. Literature Review

The correlation between FD and energy consumption has been extensively studied in academic research. Two contrasting hypotheses have been developed to delineate the impact of FD on energy consumption. On the one hand, FD can increase energy demand by improving consumption and fixed investment because FD enhances the efficiency of the financial system, allows the inflow of financial capital and foreign direct investment, promotes banking activities, reduces financial risk and loan costs and minimizes information asymmetry between lenders and borrowers [1,17,18]. According to the study of Sadorsky [19], FD could increase energy consumption through direct effects, business effects and wealth effects. First, FD directly increases energy consumption by alleviating credit constraints and enabling customers to buy durable goods, such as automobiles, refrigerators and washing machines. These durable items typically consume a large amount of energy, thereby increasing the overall energy demand of the economy. Second, for the business effect, a properly operating financial system can make loans available to businesses through various financing channels at low interest rates. These loans motivate businesses to increase their production scale and the quantity of goods they manufacture, which, in turn, indirectly promotes energy consumption. Third, with respect to the wealth effect, the stock market is often viewed as an indicator of economic conditions and plays a crucial role in FD. A dynamic stock market can increase the confidence of investors and consumers, leading to increased economic activity and energy consumption. Moreover, FD can indirectly increase energy consumption by promoting economic growth. That is, as the economy grows, energy demand and consumption also increase to support the expansion [20].

On the other hand, FD can lead to decreased energy usage. Specifically, to reduce the costs associated with manufacturing, businesses frequently make efforts to increase energy efficiency or decrease energy usage. Furthermore, financial institutions and markets have the ability to offer funding to businesses in order to alleviate their financial limitations. This, in turn, allows them to enhance their production methods, make investments in energy-efficient technology, and ultimately lead to a decrease in energy usage. Furthermore, listed firms are subject to public scrutiny because of the need for regular information disclosure. To maintain a positive public image, firms strive to implement energy-saving technologies and practices throughout their operations to assume social responsibility for environmental protection, thereby facilitating the objective of saving energy [1,2,13,21,22,23].

Empirically, the relationship between FD and energy consumption has been addressed extensively in the literature; however, the results are inconclusive. For example, the studies of Sadorsky [19], Chang [24], Ahmed [25], Mukhtarov, et al. [26], Ma and Fu [27] and Lefatsa, et al. [28] reported that FD has a positive effect on energy consumption, while the findings of Destek [29], Gómez and Rodríguez [30], Ouyang and Li [31] and Ahmed, et al. [32] indicated that FD could reduce energy consumption. On the other hand, some studies have stressed the complexity of the relationship between FD and energy consumption. For example, Baloch, et al. [33] concluded that the relationship between FD and energy consumption has an inverted U shape. Yue, et al. [34] suggested that there is no significant direct relationship between FD and energy consumption. Furthermore, Nguyen, et al. [35] reported that FD positively affects energy consumption in countries with weak institutional environments, while its influence is negative in countries with strong institutional environments. Çoban and Topcu [36] and Topcu and Payne [37] reported that the effect of FD on energy consumption was insignificant.

The consumption of renewable energy has been extensively studied because it strongly influences carbon emissions. Over the past few decades, there has been a significant increase in the demand for energy in general and renewable energy in particular. This growth in energy consumption can be attributed to an increase in population, changes in lifestyle and improvements in competitiveness. Statistics have revealed a dramatic increase in fossil fuel consumption in recent years, which has led to environmental problems, including carbon emissions and global warming. Researchers and policymakers around the world are investigating ways to reduce carbon emissions, decrease the consumption of fossil fuel and increase the efficiency of energy consumption by developing innovations in technologies, specifically for renewable energy [11].

Renewable energy is beneficial because of its cleanliness, endless supply and wide distribution. Therefore, the relationship between FD and REC is the subject of many current studies. Although green energy production helps mitigate future environmental damage, transitioning from fossil fuels to renewable energy presents certain challenges. The cost of production is one of the challenges in switching from fossil fuels to renewable energy sources. The process of producing renewable energy entails numerous financial challenges, such as infrastructure expenses, operational costs and startup costs. Hence, a crucial element that positively impacts the growth of the renewable energy sector and demand for renewable energy is financial development. An established financial system can offer affordable debt financing for projects in the renewable energy sector. Additionally, a robust stock market has the potential to channel investments towards sustainable businesses, thus aiding in the advancement of renewable energy production and utilization [1,12,24]. According to Tamazian, Chousa and Vadlamannati [21], FD encourages businesses to invest in cutting-edge and energy-efficient technologies because it promotes energy innovation, leading to the production of energy from alternative and cleaner sources, such as renewable energy.

Moreover, because renewable energy is more expensive than conventional energy sources are, more financial resources must be available to promote the renewable energy industry. Investment in renewable energy sources can thus be supported by the expansion of the financial sector [2]. In contrast to the established traditional energy sector, the renewable energy industry demands substantial and prolonged financial commitments during its initial phases. The research and development processes involved are highly unpredictable, leading to heightened investment risks. These are critical considerations that must not be overlooked. Entities within the financial sector, particularly commercial banks, are cautious about providing long-term, high-risk loans and typically implement stringent risk assessment measures. This cautious approach may hinder their willingness to lend to the renewable energy industry, thereby impeding the growth and utilization of renewable energy sources [13].

Although many empirical studies have been conducted to examine the connection between FD and energy consumption, few have investigated the relationship between FD and renewable energy use. Furthermore, the findings of previous studies have not been conclusive. For example, Brunnschweiler [9], who made one of the first attempts to explain how FD could influence renewable energy use, revealed that the banking sector’s FD in non-OECD countries had a positive and significant effect on REC. Additionally, Paramati, et al. [38] and Kutan, et al. [39] highlighted the importance of stock market development in promoting REC. The main aim of Khan, Khan and Binh [11] was to examine the heterogeneity of REC and FD in a global panel of 192 countries. To account for distributional and unobserved individual heterogeneity, the researchers utilized panel quantile regression and found that FD had a highly significant and positive impact on REC across all quantiles, except for the highest quantile (95th quantile), where the impact was found to be insignificant.

Moreover, the purpose of the Raza, Shah, Qureshi, Qaiser, Ali and Ahmed [1] study was to examine the nonlinear relationship between FD and REC in the top REC countries. They used panel smooth transition regression (PSTR) from 1997 to 2017 to show that all FD indicators negatively affected REC in countries with low FD. However, in countries with high FD, the association between FD and REC was positive. In addition, the study revealed that each FD indicator affected REC differently. Le, et al. [40] investigated the relationship between FD and REC across 55 economies from 2005 to 2014, using a two-step generalized method of moments (GMM) approach and cross-country panel data. The study revealed that FD has a positive and significant effect on renewable energy usage, but only in countries with high income levels. However, in countries with low levels of income, the effect was found to be insignificant.

Utilizing a panel dataset of G20 countries from 2005 to 2018, Wang and Dong [15] used the panel threshold model to investigate the impact of FD on renewable energy. The results revealed that there were no significant linear relationships between FD and REC. However, the relationship between FD and renewable energy was found to be nonlinear. The study revealed that FD can significantly increase REC only when the population, economic growth and technology exceed a certain threshold value; otherwise, FD can adversely affect REC. However, the findings of Assi, Isiksal and Tursoy [14] showed that FD does not play a significant role in REC. According to Saygın and İskenderoğlu [41], there is a significant positive correlation between FD and REC in developed countries. However, this correlation was evident only when measuring FD using banking variables; the coefficients were found to be negative and not statistically significant when FD was measured using stock market variables. The findings of Amin, et al. [42] confirmed that FD has a significant negative effect on REC, while Mustafa, KIVRAKLAR and Nilcan [18] asserted that both stock market and banking FD indicators positively affect REC. On the other hand, Saadaoui [43] showed that the overall FD index does not have a significant effect on the transition process in the long run, but in the short term, the impact is negative and significant.

Using a dynamic panel model and panel data, Sun, Zhang and Gao [13] analyzed the influence of FD on REC across 103 economies, showing that overall FD had a positive effect on REC. However, this effect was caused mainly by the development of financial institutions. The study also explored the impact of financial institutions and financial markets on REC in terms of depth, access and efficiency. The results showed that in the case of financial institutions, all three aspects had a positive and significant effect on REC, while for financial markets, only the efficiency aspect had a positive effect. Additionally, the study revealed that FD was more effective in promoting REC in developed economies. Yu, Jin, Zhang and Chong [2] examined the relationships between FD and REC using data from a panel of 35 countries during the 1996–2018 period. The results from the moderating effect and panel threshold models indicated that the overall FD index promotes REC and that the improvement in information and communication technology promotes the positive impact of FD on REC. Based on the previous discussion, this paper suggests two hypotheses, presented as follows:

Hypothesis 1.

FD has a significant impact on REC.

Hypothesis 2.

FD’s impact on REC differs between advanced economies and emerging and developing economies.

3. Materials and Methods

The dependent variable in the present paper is REC, which is measured as a proportion; thus, it ranges between 0 and 1. Estimating a model with a fractional response via standard linear methods may generate biased estimation and inference because linear models can produce predicted values of the dependent variable that fall outside the boundaries (0, 1). Moreover, linear models predict constant partial effects of unit changes in the explanatory variables, and variables with minimum and maximum bounds are likely to be affected by floor and ceiling effects. As a result, they often display nonconstant responses to changes in the predictors as they approach the bounds [16,44]. The FRM developed by Papke and Wooldridge [16] offers a solution to the aforementioned constraints and presents a robust methodology for addressing fractional dependent variables. The FRM considers the fractional nature of the dependent variable, enabling it to provide accurate predictions within the boundaries of said variable. Furthermore, it has the capability to capture the nonlinearity present in the data, a feature that is lacking in linear estimation models. This model allows for the estimation of partial effects at various percentiles of the independent variable distribution, resulting in a superior fit compared to linear estimation models. Nevertheless, the original FRM has restrictions when it comes to controlling for unobserved heterogeneity, especially in the analysis of panel data. To address this issue, Wagner [45] added firm-specific intercepts to the logit formulation of the FRM in his analysis of the factors that affect German firms’ exports; this was done to account for firm fixed effects. Therefore, the current paper will estimate the following fractional logit model to assess how FD impacts REC:

$$E\left({REC}_{it}|{\sum }_{i=1}^k{x}_{it})\right)=\Phi \left({\beta }_0+\sum _{i=1}^k{\beta }_i{x}_{it}+\sum _{i=1}^k{\alpha }_i{D}_i+\sum _{i=1}^k{\theta }_t{D}_t\right)$$

where ${REC}_{it}$ denotes the renewable energy consumption of the i-th country in year t; $X_{it}$ is a vector explanatory variable of the i-th country in year t; $D_i$ represents the country’s dummy variables; $D_t$ is the year’s dummies; and $\Phi \left(.\right)$ is the standard normal cumulative distribution function (cdf). The model is estimated via a panel dataset encompassing 34 advanced economies and 64 emerging markets and developing economies from 2008 to 2020.

Financial development refers to the degree of advancement of the financial sector in a given economy. This concept can be examined as a whole or broken down into its components, such as financial institutions or markets. In the present paper, FD is measured using three of the nine indices included in the FD index constructed by Svirydzenka [46]. The first index measures overall financial development and captures FD in terms of institutions and markets. The second and third indices measure the development of financial institutions and financial markets, respectively, in terms of their depth, access and efficiency. Following previous studies, to assess the relationship between REC and FD, a vector of control variables is included in the model: gross domestic product, trade openness and CO₂ emissions.

Table 1. Descriptions of Variables and Data Source.

Symbol	Variables	Data Source
REC	Renewable Energy Consumption (% of total final energy consumption)	World Bank
AFD	Overall financial Development	IMF
FID	Financial Institutions Development	IMF
FMD	Financial Markets Development	IMF
LPGDP	The natural logarithm of gross domestic product per capita (constant 2015 US$)	World Bank
LTRADE	The natural logarithm of trade openness (% of GDP)	World Bank
CO2	CO2 emissions (metric tons per capita)	World Bank

shows the details of the variables used.

Table 2. Descriptive Statistics.

Panel A: Advanced Economies
Variable	Mean	Median	Std. Dev.	Min	Max	Skewness	Kurtosis
REC	0.194	0.143	0.172	0.002	0.828	1.540	5.428
AFD	0.655	0.689	0.198	0.201	0.988	−0.592	2.519
FID	0.729	0.747	0.145	0.367	1.000	−0.391	2.472
FMD	0.555	0.613	0.272	0.019	0.99	−0.635	2.249
LPGDP	10.496	10.598	0.521	9.278	11.61	−0.188	2.45
LTRADE	4.616	4.589	0.631	3.153	6.093	0.295	2.803
CO2	8.183	7.424	3.766	2.97	22.558	1.299	4.371
Obs.	468
Panel B: Emerging Markets and Developing Economies
Variable	Mean	Median	Std. Dev.	Min	Max	Skewness	Kurtosis
REC	0.188	0.127	0.188	0.000	0.902	1.271	4.438
AFD	0.337	0.333	0.146	0.068	0.741	0.329	2.485
FID	0.433	0.424	0.137	0.133	0.718	0.165	2.328
FMD	0.228	0.206	0.207	0.000	0.74	0.610	2.235
LPGDP	8.882	8.748	0.799	6.627	11.205	0.244	3.373
LTRADE	4.310	4.342	0.437	3.096	5.404	−0.198	2.921
CO2	5.700	3.558	6.470	0.593	39.583	2.470	9.652
Obs.	832

displays the descriptive statistics of the variables used. The table indicates that for the advanced economies sample and the emerging markets and developing economies sample, the REC values are skewed towards the right (with values of 1.514 and 1.271, respectively) and are highly kurtotic (with values of 5.305 and 4.438, respectively). According to Gallani, Krishnan and Wooldridge [44], the FRM is likely to provide the most informative advantages in this case because the response variable has a skewed distribution, and many observations accumulate at one limit.

4. Empirical Results

The results from

Table 3. Regression Results Overall Financial Development Index.

	Panel A: Advanced Economies
Variables	Coef. FE	Coef. FRM	FRM Average Partial Effects (APE)
			APE	APE at Specific Percentiles
				APE (1%)	APE (5%)	APE (50%)	APE (95%)	APE (99%)
AFD	0.022 [0.028]	0.451 ** [0.203]	0.060 ** [0.027]	0.055 ** [0.023]	0.056 ** [0.023]	0.060 *** [0.028]	0.063 ** [0.030]	0.064 ** [0.031]
LPGDP	0.038 ** [0.017]	0.482 *** [0.170]	0.064 *** [0.023]	0.050 *** [0.013]	0.054 *** [0.016]	0.067 *** [0.025]	0.077 ** [0.031]	0.079 ** [0.033]
LTRADE	0.069 *** [0.013]	0.781 *** [0.110]	0.104 *** [0.015]	0.063 *** [0.005]	0.072 *** [0.007]	0.107 *** [0.016]	0.143 *** [0.023]	0.146 *** [0.023]
CO2	−0.014 *** [0.002]	−0.123 *** [0.019]	−0.017 *** [0.003]	−0.021 *** [0.004]	−0.021 *** [0.004]	−0.018 *** [0.003]	−0.010 *** [0.001]	−0.007 *** [0.001]
Constant	−0.394 ** [0.190]	−9.049 *** [1.730]
Joint Significance	2343.77 ***	78,956.23 ***
Obs.	442	442
	Panel B: Emerging Markets and Developing Economies
Variables	Coef. FE	Coef. FRM	APE	APE (1%)	APE (5%)	APE (50%)	APE (95%)	APE (99%)
AFD	−0.086 *** [0.030]	−0.570 ** [0.244]	−0.068 ** [0.029]	−0.072 ** [0.033]	−0.071 ** [0.032]	−0.068 ** [0.029]	−0.064 ** [0.026]	−0.062 ** [0.025]
LPGDP	−0.013 [0.011]	0.140 [0.122]	0.017 [0.015]	0.015 [0.012]	0.016 [0.013]	0.017 [0.015]	0.018 [0.017]	0.019 [0.018]
LTRADE	0.022 *** [0.008]	0.176 *** [0.062]	0.021 *** [0.008]	0.020 *** [0.007]	0.020 *** [0.007]	0.021 *** [0.008]	0.022 *** [0.008]	0.022 *** [0.009]
CO2	−0.004 *** [0.002]	−0.217 *** [0.033]	−0.026 *** [0.004]	−0.035 *** [0.007]	−0.034 *** [0.007]	−0.027 *** [0.005]	−0.002 *** [0.001]	−0.0001 [0.0001]
Constant	0.419 *** [0.083]	−1.897 ** [0.925]
Joint Significance	1720.31 ***	160,000 ***
Obs.	832	832

*** and ** donate significance at the 1% and 5% respectively. Robust standard errors in brackets.

indicate that, when measured using the overall financial development index (AFD), the effect of FD is significant and positive in advanced economies, suggesting that an improved financial system helps promote REC. This finding is strongly in line with previous empirical evidence (i.e., Sun, Zhang and Gao [13]; Yu, Jin, Zhang and Chong [2]; Alsagr and van Hemmen [47]; Khan, Khan and Binh [11]). However, the effect is significantly negative in emerging and developing economies. This finding is consistent with that of Raza, Shah, Qureshi, Qaiser, Ali and Ahmed [1], who reported a negative relationship between FD and REC in countries with low FD. Additionally, the results presented in Table 3 indicate that the impact of overall FD on REC in advanced economies exhibits an escalation towards the upper tail of the predictor. That is, the coefficient of AFD increases from 0.055 at the first percentile to 0.064 at the 99th percentile. These findings suggest that in countries in the upper percentile of overall FD, REC is more strongly influenced by the positive effect of FD. Moreover, the negative effect of AFD on REC in emerging economies diminishes along the range of the predictor.

However, the findings in

Table 4. Regression Results Financial Institutions Development Index.

	Panel A: Advanced Economies
Variables	Coef. FE	Coef. FRM	FRM Average Partial Effects (APE)
			APE	APE at Specific Percentiles
				APE (1%)	APE (5%)	APE (50%)	APE (95%)	APE (99%)
FID	−0.116 *** [0.030]	−0.392 [0.250]	−0.052 [0.033]	−0.055 [0.040]	−0.055 [0.036]	−0.052 [0.033]	−0.050 [0.031]	−0.050 * [0.030]
LPGDP	0.028 * [0.016]	0.427 ** [0.173]	0.057 ** [0.023]	0.046 *** [0.015]	0.049 *** [0.017]	0.059 ** [0.025]	0.067 ** [0.031]	0.069 ** [0.032]
LTRADE	0.044 *** [0.013]	0.66 *** [0.110]	0.087 *** [0.015]	0.058 *** [0.006]	0.064 *** [0.008]	0.090 *** [0.016]	0.117 *** [0.023]	0.112 *** [0.024]
CO2	−0.014 *** [0.002]	−0.123 *** [0.019]	−0.017 *** [0.003]	−0.021 *** [0.004]	−0.021 *** [0.004]	−0.018 *** [0.003]	−0.010 *** [0.001]	−0.007 *** [0.0001]
Constant	−0.063 [0.187]	−7.193 *** [1.872]
Joint Significance	2864.95 ***	88,649.82 ***
Obs.	442	442
	Panel B: Emerging Markets and Developing Economies
Variables	Coef. FE	Coef. FRM	APE	APE (1%)	APE (5%)	APE (50%)	APE (95%)	APE (99%)
FID	−0.129 *** [0.034]	−0.719 *** [0.207]	−0.085 *** [0.025]	−0.092 *** [0.029]	−0.091 *** [0.028]	−0.086 *** [0.025]	−0.080 *** [0.022]	−0.079 *** [0.021]
LPGDP	−0.002 [0.010]	0.213 * [0.117]	0.025 * [0.014]	0.022 ** [0.010]	0.024 ** [0.012]	0.026 * [0.014]	0.029 * [0.017]	0.030 [0.018]
LTRADE	0.020 *** [0.008]	0.159 *** [0.060]	0.019 *** [0.008]	0.018 *** [0.007]	0.018 *** [0.007]	0.019 *** [0.008]	0.020 ** [0.008]	0.020 ** [0.008]
CO2	−0.005 *** [0.001]	−0.218 *** [0.033]	−0.026 *** [0.004]	−0.035 *** [0.007]	−0.035 *** [0.007]	−0.028 *** [0.005]	−0.002 *** [0.001]	−0.0001 [0.0001]
Constant	0.372 *** [0.080]	−2.270 ** [0.895]
Joint Significance	1651.93 ***	170,000 ***
Obs.	832	832

***, ** and * donate significance at the 1%, 5% and 10% respectively. Robust standard errors in brackets.

show that financial institution development (FID) has an insignificant effect on REC in advanced economies; however, its effect is significantly negative in emerging economies. Furthermore, the results indicate that the negative effect of FID on REC in emerging economies decreases as the value approaches the upper tail of the predictor. The coefficient of FID decreases from 0.092 at the first percentile to 0.079 at the 99th percentile. These findings suggest that the negative effect of FID on REC is more pronounced in countries in the lower percentile of FID. In contrast, the findings presented in

Table 5. Regression Results Financial Markets Development Index.

	Panel A: Advanced Economies
Variables	Coef. FE	Coef. FRM	FRM Average Partial Effects (APE)
			APE	APE at Specific Percentiles
				APE (1%)	APE (5%)	APE (50%)	APE (95%)	APE (99%)
FMD	0.061 *** [0.019]	0.436 *** [0.124]	0.058 *** [0.017]	0.052 *** [0.014]	0.0518 *** [0.014]	0.059 *** [0.017]	0.062 *** [0.019]	0.062 *** [0.019]
LPGDP	0.031 * [0.017]	0.427 ** [0.170]	0.057 ** [0.023]	0.046 *** [0.014]	0.049 *** [0.016]	0.059 ** [0.025]	0.067 ** [0.030]	0.067 ** [0.032]
LTRADE	0.063 *** [0.013]	0.736 *** [0.106]	0.098 *** [0.015]	0.061 *** [0.005]	0.069 *** [0.007]	0.100 *** [0.015]	0.133 *** [0.023]	0.136 *** [0.023]
CO2	−0.014 *** [0.002]	−0.122 *** [0.019]	−0.017 *** [0.003]	−0.021 *** [0.004]	−0.021 *** [0.004]	−0.018 *** [0.003]	−0.010 *** [0.001]	−0.007 *** [0.0001]
Constant	−0.329 [0.200]	−8.251 *** [1.677]
Joint Significance	2163.68 ***	79,038.58 ***
Obs.	442	442
	Panel B: Emerging Markets and Developing Economies
Variables	Coef. FE	Coef. FRM	APE	APE (1%)	APE (5%)	APE (50%)	APE (95%)	APE (99%)
FMD	−0.001 [0.015]	0.046 [0.149]	0.006 [0.018]	0.006 [0.018]	0.006 [0.018]	0.006 [0.018]	0.006 [0.018]	0.006 [0.018]
LPGDP	−0.018 * [0.011]	0.097 [0.119]	0.012 [0.014]	0.011 [0.013]	0.011 [0.013]	0.012 [0.014]	0.013 [0.016]	0.013 [0.017]
LTRADE	0.022 *** [0.008]	0.172 *** [0.062]	0.021 *** [0.008]	0.019 *** [0.007]	0.020 *** [0.007]	0.021 *** [0.008]	0.022 *** [0.008]	0.022 *** [0.009]
CO2	−0.004 *** [0.002]	−0.222 *** [0.033]	−0.027 *** [0.004]	−0.036 *** [0.007]	−0.035 *** [0.007]	−0.028 *** [0.005]	−0.002 *** [0.001]	−0.0001 [0.0001]
Constant	0.450 *** [0.084]	−1.625 * [0.912]
Joint Significance	1759.15 ***	170,000 ***
Obs.	832	832

***, ** and * donate significance at the 1%, 5% and 10% respectively. Robust standard errors in brackets.

suggest that the growth of financial markets does not significantly affect REC in emerging economies but has a significant positive effect in advanced economies. Moreover, the positive impact of financial market development (FMD) on REC in advanced economies increases moving towards the upper tail of the predictor.

According to the above findings, the positive impact of FD on REC in advanced economies is driven mainly by the development of financial markets. This finding is in line with the findings of Mustafa, KIVRAKLAR and Nilcan [18]. However, in emerging and developing economies, FD has a negative effect on REC, with financial institutions being the main driver of this impact. This finding is consistent with that of Raza, Shah, Qureshi, Qaiser, Ali and Ahmed [1], who reported a negative relationship between FD and REC in countries with low FD. These results could be explained by the significant and prolonged financial commitments the renewable energy industry requires during its early stages. The research and development processes involved are highly unpredictable, increasing investment risks. These are crucial factors that should be considered. Entities in the financial sector, especially commercial banks, in developing economies are cautious about offering long-term, high-risk loans and typically implement strict risk assessment measures. This cautious approach may hinder their willingness to lend to the renewable energy industry, thus impeding the growth and adoption of renewable energy sources [10].

In light of our results, policymakers in emerging and developing economies can establish policies that connect financial incentives, such as subsidies and tax exemptions, with renewable energy goals to encourage investment in renewable energy projects. Moreover, policymakers may find it optimal to promote the development of a green finance system through developing policies and regulations that support the financial system and establishing green funds to finance renewable energy projects and initiatives. Furthermore, in both advanced economies and emerging and developing economies, policymakers should work to encourage financial institutions to create financial products, such as green bonds, to promote the renewable energy industry and lower the risks associated with investments in renewable energy. Moreover, policymakers could introduce laws that mandate that publicly traded companies disclose their environmental, social and governance (ESG) factors. These laws motivate firms to embrace sustainable practices, such as investing in renewable energy projects and conducting their operations with renewable energy resources.

Moreover, the results presented in Table 3, Table 4 and Table 5 indicate that an increase in GDP and trade openness leads to a significant increase in REC, while REC decreases as CO₂ emissions increase. These results are consistent with the findings of Sun, Zhang and Gao [13], Yu, Jin, Zhang and Chong [2] and Khan, Khan and Binh [11]. The findings also reveal that the GDP and trade openness coefficients increase significantly as the GDP and trade openness ranges increase. Moreover, the effect of CO₂ decelerates significantly towards the upper tail of the predictor.

As a robustness check, we re-estimate our model via a linear method (fixed effect model). The results in Table 3, Table 4 and Table 5 indicate that the linear model produces results similar to those of the nonlinear model. In other words, the results from the FE model indicate that financial development has a positive effect on REC in advanced economies and a negative effect in emerging and developing economies.

Since the dependent variable in the present paper is a proportion that varies between 0 and 1, the FRM outperforms the FE model. Linear methods may not accurately represent the effects of FD on REC across the entire range of FD values, since they estimate the constant partial effects of unit changes in the explanatory variable. In contrast, the FRM can capture the nonlinearity of the data because it allows the estimation of partial impacts at various percentiles of the distribution of independent variables. Linear methods may also produce predicted values that fall outside the range of zero to one.

Table 6. Predicted Values.

Panel A. Advanced Economies
Variable	Mean	Std. Dev.	Min	Max
Observed REC	0.194	0.172	0.002	0.828
Predicted REC Estimated from FRM Regression Presented in Table 3 Panel A	0.203	0.172	0.006	0.809
Predicted REC Estimated from FE Regression Presented in Table 3 Panel A	0.203	0.172	−0.031	0.792
Predicted REC Estimated from FRM Regression Presented in Table 4 Panel A	0.203	0.172	0.006	0.810
Predicted REC Estimated from FE Regression Presented in Table 4 Panel A	0.203	0.172	−0.025	0.807
Predicted REC Estimated from FRM Regression Presented in Table 5 Panel A	0.203	0.172	0.006	0.808
Predicted REC Estimated from FE Regression Presented in Table 5 Panel A	0.203	0.172	−0.029	0.794
Panel B. Emerging Markets and Developing Economies
Variable	Mean	Std. Dev.	Min	Max
Observed REC	0.188	0.188	0.000	0.902
Predicted REC Estimated from FRM Regression Presented in Table 3 Panel B	0.188	0.187	0.0001	0.868
Predicted REC Estimated from FE Regression Presented in Table 3 Panel B	0.188	0.188	−0.047	0.942
Predicted REC Estimated from FRM Regression Presented in Table 4 Panel B	0.188	0.187	0.0001	0.868
Predicted REC Estimated from FE Regression Presented in Table 4 Panel B	0.188	0.188	−0.037	0.940
Predicted REC Estimated from FRM Regression Presented in Table 5 Panel B	0.188	0.187	0.0001	0.869
Predicted REC Estimated from FE Regression Presented in Table 5 Panel B	0.188	0.188	−0.046	0.943

shows the predicted values of REC estimated via the FE model and the FRM. The predicted values of REC estimated via the FRM clearly fall within the interval of 0–1, while those estimated via the FE model do not. Therefore, the FRM provides a more accurate fit than the FE model does.

In

Table 7. Regression Results with FDDC.

Variables	Advanced Economies			Emerging Markets and Developing Economies
	APE			APE
FDDC	0.022 *** [0.007]			−0.045 *** [0.017]
LPGDP	0.086 *** [0.024]			0.008 [0.018]
LTRADE	0.113 *** [0.018]			0.024 *** [0.009]
CO2	−0.016 *** [0.003]			−0.028 *** [0.005]
Joint Significance	63,568.54 ***			150,000 ***
Obs.	386			660
APE at Specific Percentiles
	APE (1%)	APE (5%)	APE (50%)		APE (95%)	APE (99%)
FD/Advanced Economies	0.021 *** [0.006]	0.021 *** [0.006]	0.022 *** [0.007]		0.023 *** [0.008]	0.024 *** [0.008]
FD/Emerging Markets and Developing Economies	−0.047 ** [0.018]	−0.047 ** [0.018]	−0.045 *** [0.017]		−0.041 *** [0.014]	−0.040 *** [0.013]

*** and ** donate significance at the 1%and 5% respectively. Robust standard errors in brackets.

, we apply an alternative measure of FD—domestic credit to private sector %GDP (FDDC)—to check whether the results will change. FDDC is commonly used in the literature as a proxy for overall financial development. The results in Table 7 align with the findings of the regression in which AFD is used. In advanced economies, the coefficient of FDDC is significant and positive and increases towards the upper tail of FDDC. However, in emerging and developing economies, the coefficient of FDDC is negative, and the negative effect diminishes along the range of FD.

5. Conclusions

In recent years, energy and environmental experts have emphasized the importance of using renewable energy sources to address the environmental damage and energy shortages caused by insufficient energy supply. FD is considered a key factor in promoting the growth of the renewable energy industry and increasing the demand for renewable energy [10,41]. Advanced financial institutions can offer low-cost debt financing for renewable energy projects. Additionally, a well-functioning stock market can efficiently allocate capital to environmentally friendly sectors, thereby encouraging the production and consumption of renewable energy. The present paper aimed to assess how FD affects REC in 34 advanced economies and 64 emerging markets and developing economies from 2008 to 2020. Few studies have explored the connection between FD and REC. This paper differs from past research by employing the FRM to analyze the association between FD and REC. Because the dependent variable in the present paper is a proportion that varies between 0 and 1, standard linear models may not accurately represent the effects of FD on REC across the entire range of FD values and may result in predicted values that lie outside the range of zero to one. Moreover, linear models estimate the constant partial effects of unit changes in the explanatory variable, while FRM has the ability to capture the nonlinearity of the data because it allows the estimation of partial impacts at various percentiles of the distribution of independent variables, resulting in a more accurate fit than linear estimation models provide.

The findings highlight the positive impact of FD on REC in advanced economies. Therefore, policymakers should take into account the role of FD when formulating renewable energy policies. The financial sector has the potential to offer significant support to the renewable energy industry, which can foster long-term growth and sustainability. Governments can also implement specific policies, such as offering specialized loans and favorable interest rates, to help alleviate the financial challenges of renewable energy firms. However, FD negatively affects REC in emerging and developing economies. Therefore, to encourage investment in renewable energy projects, policymakers in emerging and developing economies can implement policies that align financial incentives with renewable energy goals.

One main limitation of this paper is the omission of varying levels of environmental policy across different countries while analyzing the impact of FD on REC within advanced economies as well as emerging and developing economies. Therefore, the study suggests that future research investigate how the level of environmental policy could reshape the relationship between FD and REC.