In this section, we will focus on the preliminary analysis, result interpretation and discussion. The step-by-step procedure of the analysis and methodology used for the result findings is presented below (as shown in Fig. 3):

Fig. 3 Flowchart showing Analytical Methodology. Full size image

The summary statistics for our balanced panel data, providing an overview of the central tendency, spread, shape, and statistical significance of five variables are shown in Table 2. Our study contains a total of 342 observations with mean of CO 2 , FinTech, TI, REC and GDPPC as 5.83, 0.13, 3.98, 0.93, and 4.20 respectively. The spread or dispersion of the variable CO 2 , REC and GDPPC shows moderate variability, whereas FinTech and TI shows low and high spread around the mean, respectively. REC shows the most skewed distribution, with a high negative skew, and CO 2 has a significant positive skew. The variables have varying degrees of normality, with CO 2 , FinTech, TI, REC, and GDPPC showing significant deviations from normality (based on the Jarque–Bera test). The results of the Jarque–Bera statistics thus show the rejection of the null hypothesis, which suggests a normal distribution. This finding highlights the necessity of a more potent approach to handling the non-normal distribution. As a result, our study deploys the GMM-PVAR strategy to deal with the non-normality situation. Next, the results obtained from the correlation matrix (as shown in Table 3) depicts the negative impact of FinTech and GDPPC on CO 2 whereas this effect is positive in case of TI and REC.

Table 2 Descriptive Statistics. Full size table

Table 3 Correlation Matrix. Full size table

Next, our study employs cross-sectional dependency (CSD) test and the slope heterogeneity test in order to offer the basic framework for our analysis. Results for CSD test and slope heterogeneity test are given in Table 4. Results obtained points to the presence of cross-sectional dependency and slope heterogeneity in our analysis. Therefore, it is essential to look at the second-generation tests in the econometric assessment’s later portions in order to get over these challenges. We likewise opt for the second-generation CIPS unit root test. According to the CIPS test findings (as shown in Table 5), all indicators are stationary at the first difference at the 1% level of significance.

Table 4 Cross-Sectional Dependence Tests and Slope homogeneity. Full size table

Table 5 Second generation unit root tests (CIPS). Full size table

We next verified whether or not there is cointegration among the variables before proceeding with result estimate. According to the Pedroni and Westerlund panel cointegration test results (as presented in Table 6), cointegration exists among the variables since all statistics reject the null hypothesis that there is no cointegration at the 1% and 5% significance levels.

Table 6 Cointegration Tests (Pedroni and Westerlund test). Full size table

We used the panel VAR approach for our results in the subsequent stage of our investigation.

First, as indicated in Table 7, we give the eigenvalue stability requirement for our models. This led us to the conclusion that the calculated eigenvalues satisfy our stability criterion as their values are less than 1. Additionally, Fig. 4 displays the VAR stability graph for our model. The following figure demonstrates that all of these eigenvalues fall inside the unit circle, proving the stationarity of our estimations and allowing us to proceed with the panel VAR model forecasting of the variance decomposition and impulse response function.

Table 7 Eigenvalue stability condition. Full size table

Next, our study uses Forecast error variance decomposition (FEVD) to demonstrate the inter-relationship among the variables. To find out how sensitive one indicator is to changes in another indication in the particular model over a long period of time, FEVD is used. Here, we showed the outcomes over the second, fifth, and tenth time-periods (as shown in Table 8). From the results obtained, it was observed that the forecast error variance of each variable is mostly impacted by its own historical shocks. Throughout all times, CO 2 is heavily impacted by its own prior shocks, but after 5th period horizons, FinTech begins to have a major impact. The impact of FinTech on CO 2 increased from 0.0136 over 5th horizon to 0.025 over 10th horizon period. Although CO 2 has a significant influence, particularly in the immediate term (two periods ahead), FinTech is primarily self-explanatory. Similar to FinTech, TI is mostly described by its own shocks, with the other factors having little impact. With a small contribution from FinTech and TI, CO 2 has a major impact on REC. The impact of FinTech on REC increases over the time with 0.0058 over 2nd horizon to 0.102 over 10th horizon. With CO 2 having a significant impact at shorter timeframes, GDPPC exhibits a strong dependency on its own shocks. In addition, the impact of FinTech and TI on GDPPC shows increasing trend from 2nd horizon to 10th horizon.

Table 8 Forecast error variance decomposition (FEVD). Full size table

Figure 5 then displays the impulse response function (IRF) analysis. IRFs highlight the interdependence of all variables, provide insight into how shocks propagate across a system, and investigate if modifications to one variable have a cascade effect on other variables. From the IRFs, it was observed that FinTech has positive impact on GDPPC and REC, which indicates that rising adoption of financial technologies will not only increase the economic growth but also supports the green energy consumption. Previous literature has also observed the positive impact of FinTech on economic growth24,27,28. FinTech primarily appears to enhance economic growth by enhancing financial inclusion, increasing productivity, and enabling more seamless financial transactions. For renewable energy consumption, FinTech can support energy technology innovation, crowdsourcing renewable projects, or enabling green energy investments through digital platforms, thereby, increasing the funding and market access for renewable energy sources. FinTech may be a major factor in advancing environmental sustainability and indicating a move toward environmentally friendly technologies59,60. Similarly, findings depict decrease in CO 2 with the rise of FinTech, indicating that environmental sustainability is improved with the technological advancements in FinTech sector. This is consistent with previous studies by12,39,61,62. Fintech’s ability to lower emissions is reinforced by the fact that decentralized finance, blockchain, and more digital payments have reduced the need for physical infrastructure. Furthermore, FinTech may encourage green financing, green investments, and environmental greening by assisting businesses and consumers in making eco-friendly choices.

Fig. 5 Impulse Response for 10-year horizons generated using VAR. Full size image

Similar are the findings of technological innovation. With the rise in TI, there is positive impact on GDPPC and REC, however, this impact is negative in case of CO 2 . This indicates that investment in technological innovation exerts positive impact on economic growth, shift to cleaner energy sources, and strengthens the move towards the environmental sustainability. Firms and industry sectors benefit from increased production, improved efficiency, and new opportunities brought about by investments in TI. Technological advancements that expand markets, boost already-existing sectors, and boost overall financial performance may be the source of this growth44,45. Previous study suggests that one of the main forces behind the transition to sustainable energy is technological advancements63,64. Technological innovation helps with the developments in energy-efficient technology that lower overall energy consumption, or breakthroughs in renewable energy technologies like solar, wind, and energy storage systems, and improves the affordability, effectiveness, and scalability of green technology. Further, by making it possible for more effective industrial methods, cleaner energy consumption, and improved carbon emission control, technological advancement is supporting environmental sustainability53,54,65. For instance, developments in low-carbon inventions, clean technology, and energy efficiency help to lower the total carbon footprint of people and businesses.

Further, our results reflect the positive impact of GDPPC on REC depicting that the rising economic growth increases the renewable energy consumption. This can be the result of improved access to green technologies or increased incomes, which allow economies to spend more in sustainable energy to fulfil their expanding energy needs66,67. Further, the impact of REC on GDPPC is positive whereas the impact of REC on CO 2 is negative. These findings show that increase in renewable energy consumption can increase economic growth without increasing carbon emissions. This suggests that consuming renewable energy can help achieve economic growth without sacrificing environmental sustainability68,69. This positive effect on renewable energy on environment is also highlighted by14,19. To add further, the analysis also sheds light on the impact of CO 2 on GDPPC and REC. The findings point to the negative impact on economic growth due to rising carbon emissions and more or less stable impact on REC of rising carbon emissions. Growth may eventually be slowed down as emissions increase because the long-term financial costs of climate change and environmental degradation may exceed the short-term gains40,41. In addition, the consistent effect of CO 2 emissions on REC indicates that long-term strategic objectives, technology advancements, and regulatory support ultimately push the adoption of renewable energy. It also illustrates how energy transitions might not be immediately responsive to short-term variations in CO 2 levels and could take some time. The Graphical presentation of result findings is provided in Fig. 6.