Unveiling How Digital Finance Enhances Sustainability-Oriented Organizational Performance: Insight into Mediating Role of Heterogeneous Technological Progress and Green Innovation

Due to the increasingly unstable global environment, businesses are being pushed to change their current practices in order to meet sustainability targets. All countries in the global world are placing more emphasis on how well businesses perform in the areas of governance, social responsibility, and the environment. Enterprises must adopt green practices as the guiding principles, technology as the enabler, and innovation as the engine to further the green innovation (GI) driven strategy and advance the greening of traditional industries. Vietnam, a nation in the Asia-Pacific area that is transitioning from a developing to an industrialized nation, must deal with a wide range of environmental challenges and consumption difficulties as a result of its accelerating economic expansion. Given that Vietnam's manufacturing sector has consistently acted as a major engine for the country's highly industrialized economy, it is urgent to keep an eye on how these operations are affecting these companies' overall performance, particularly their environmental performance. Organizations implement a variety of strategies at once to support their main processes with supporting sub-strategies in order to reach their sustainable goals [1]. In order to eliminate this pollution, it is vital to put into effect GI strategies and practices. Innovation is crucial right now since stakeholders and customers are constantly concerned about environmental protection. As a result, established business models must alter, and this includes a move toward GI in businesses [2].

Small and medium-sized businesses (SMEs) have not been given as much attention in recent studies on organizational sustainability and sustainable resource use [3]. SMEs are interdependent groups operating in a worldwide systemic environment that promotes the use of sustainable development resources, including social, natural, and financial ones. These groups also provide a sizable portion of the environmental effects. Still, businesses, especially those small ones, may find it difficult to implement them. This is a result of small businesses' inability to comprehend green initiatives and their lack of financial resources. However, small businesses must implement GI because they produce around 70% of the industrial pollutants and have little concern for the environment. Additionally, there has been ongoing push from various parties, including the government and industry, to integrate GI practices and principles into corporate operations and create cutting-edge goods and technologies.

Corporate growth is significantly influenced by technological progress [4]. Innovation activities, which are the fundamental driver of technological progress, are more vulnerable to financial constraints than physical capital investments. SMEs in particular are more likely than large businesses to experience financial constraints. Since many SMEs that engage in high-risk innovative initiatives are not compelled to disclose their financial reports, it is more expensive for financial institutions to evaluate their creditworthiness, making information asymmetry a bigger problem for SMEs. The results of these projects are typically intangible and cannot be utilized as collateral, which is another drawback.

Several research have indicated a positive correlation between improved digital finance (DF) and increased levels of GI [5]. Nevertheless, upon reviewing the available literature, it is evident that there has been a scarcity of research that directly examines the influence of DF on GI [6]. While Jiang et al. [5, 7] conducted empirical studies on the influence of the DF on GI at a micro level, they failed to consider the potential mechanisms and heterogeneity involved.

Moreover, the investigation of the prospective influence of DF on the sustainability centers on the correlation between DF and sectoral performance as well as DF and firm performance. On the facet of sectoral performance enhancement, Ma et al. [8] explored the advancement of China's manufacturing sector and the coordinated development of DF as a facet of sectoral performance enhancement, utilizing a coupling coordination approach. The impact of DF on agricultural green total factor productivity in China was investigated by Shen et al. [9]. Based on the paper's findings, it was evident that DF had a substantial positive influence by alleviating credit constraints. The relationship between energy efficiency and DF was investigated by Wu et al. [10]. It has been observed that areas with higher levels of DF development tend to experience energy efficiency gains. Regarding firm performance, Hossain et al. [11] investigated SME financing models designed to alleviate borrowing constraints. Through the implementation of a data-driven credit risk score methodology, they discovered that DF facilitates sustainable finance by decreasing transaction costs and default risks. Shinozaki [12] evaluated the function of DF within the framework of the Russia-Ukraine conflict on enterprises in Central and West Asia, as a mechanism for shock resilience and sustainability.

While numerous scholarly articles have emphasized the critical role of DF in sustainability, less attention has been paid to the intricacies of how DF can improve and enhance all facets of sustainability-oriented organizational performance (SOP), including environmental, financial, governance, and human performance.

This research aims at establishing a theoretical framework that is comprehensive enough to provide a holistic picture of on the role of DF in supporting SMEs which plan to become innovative businesses to be able to get the high success in their future routes. To do so, this manuscript sets its sight at constructing a theoretical framework to analyze the specific influence of DF on heterogeneous technological progress (HTP) and GI which ultimately enhance SOP. Also, this theoretical gap inspired the intriguing research questions as follows.

RQ1. How far does DF impact SOP?

RQ2. Does HTP act as a mediator in the interlink between DF and SOP?

RQ3. Does GI act as a mediator in the interlink between DF and SOP?

By examining the purpose of the research, which has never been done previously, the study makes a theoretical contribution. In Vietnam, research into the implementation of DF by SME's is still in its infancy. This research endeavor represents a groundbreaking attempt to assess the implementation of DF within SMEs and to pinpoint areas that require further investigation by scholars and practitioners. The prior study found that the extent to which DF contributes to sustainable development in SMEs that aims to establish innovative firms in developing countries is poorly understood. By utilizing an understudied research approach, this article contributes to a comprehensive comprehension of the way in which DF can be utilized to improve SOP. Additionally, it introduces a novel concept to the field by proposing four dimensions of SOP that assist managers in conducting a more practical analysis of SOP.

This research investigates the mediating role of GI in the connection between SOP and DF. Hu et al. [13] argued that empirical research on the adoption of GIs in developing environments was scarce, despite the fact that numerous studies have found that the expansion of DF fosters corporate innovation. Remarkably, this study builds upon previous research that has focused solely on the adoption of GI practices, neglecting to examine the broader impact of such adoption. Besides, this could be the initial investigation into the manner in which HTP mediates the relationship between SOP and DF. This research elucidated the critical significance of GI and HTP in accelerating the advancement of SOP. Such are the distinctive qualities of this conceptual paradigm for research.

The study's practical contribution is to furnish the government with a reference framework that facilitates the implementation of enterprise digital transformation decisions and the formulation of differentiated policies for the development of DF. Additionally, the research provides practitioners with insights into the operational and organizational tasks that are entailed in the management and planning of GI processes and technological progress.

To achieve the above-mentioned study objectives, The next sections of the current manuscript are planned as follows. In Section 2, the conceptual underpinnings are briefly discussed, and in Section 3, the study's hypotheses are developed. Section 4 explains the study technique and materials in light of these perspectives. The research findings are then elaborated and interpreted in Part 5. The discussion and implications are illuminated in Section 6 while the conclusion and future research directions are discussed in Part 7 to wrap up the manuscript.

Organizations engaged in sustainable development may find DF beneficial in mitigating financing and investment risks. DF enhances the accessibility of external resources for businesses, thereby facilitating improved financial management, competitive advantage, and the promotion of sustainable development objectives [13]. DF may contribute to a greater reduction in environmental inequality in areas where businesses relocate due to their increased demands for industrial transformation and financing [30]. DF can aid organizations in surmounting these financial constraints through the provision of liquidity assistance, facilitation of efficient, accurate, and customized customer service, and improvement of credit allocation efficiency. It is expected that DF will expedite the lending process through the accurate and timely processing of loan and mortgage applications. This could potentially support organizations in their efforts to adopt environmentally friendly and sustainable business practices. By leveraging information and data sources obtained from third-party assessments, DF is capable of significantly reducing agency issues and information asymmetry in comparison to conventional finance [31]. Therefore, the first hypothesis addressed in this study is as follows.

Hypothesis 1 (H1). DF results in a substantial and positive effect on SOP.

According to literature [32], the implementation of DF has not only resulted in the creation of enough business space and the exploration of opportunities for corporate technology innovation, but it has also led to a substantial reduction in the expenses related to search and risk detection. The advantages of digital banking, such as low cost, quick speed, and extensive coverage, contribute to the achievement of low entry barriers for financial services. The implementation of DF has the potential to enhance a company's preparedness in initiating innovative technological initiatives. The advancement of DF has a significant influence on the technological innovation of enterprises [33]. The worldwide nature of this promotion effect is particularly significant in its influence on the technical innovation of enterprises operating in countries characterized by limited traditional financial development. DF has the potential to foster industry expansion, alleviate financial limitations, and expedite technical advancement. Therefore, the second hypothesis addressed in this study is as follows.

Hypothesis 2 (H2). DF results in a substantial and positive effect on HTP.

Expanding upon the viewpoints expressed by García-Quevedo et al. [34], it will be illustrated that financial constraints have an effect on the choices that organizations make regarding innovation as well as the results that ensue from those choices. As a consequence, the capacity of an organization to innovate in the green sector could be adversely affected by financial constraints. Fortunately, the ascent of DF has the potential to effectively mitigate the credit risk faced by banks and aid businesses in addressing this matter. Chen and Yoon [35] posit that financial technology has the potential to alleviate financial burdens and reduce costs associated with debt financing. DF revolutionized credit pricing models through the transparency of credit and information [32]. Green growth will benefit from DF's potential to reduce credit approval requirements and provide businesses with an efficient financing alternative; this will accelerate industrial upgrading and foster a more sustainable environment for the development of green projects. In addition, DF enhances the efficacy of risk management and resource allocation capabilities, thereby mitigating mismatch concerns and reducing financial constraints [35]. By potentially increasing its investments in research and development, DF may be able to influence GI. Previous research has suggested that GI could be stimulated by financial support for environmental innovation research and development. Therefore, the third hypothesis addressed in this study is as follows.

Hypothesis 3 (H3). DF results in a substantial and positive effect on GI.

In light of the perpetual evolution of technology, technological components are indispensable for the adoption of GIs. Technological progress has facilitated waste reduction and the adoption of environmentally wise practices. The advancements in the processes of product and service creation, manufacturing, and distribution have caused significant and far-reaching disruptions in every aspect of human existence [36]. Recent advancements in the production and delivery processes have introduced fresh opportunities for innovation through the integration of state-of-the-art digital manufacturing technology and precise equipment [37]. It is well-established, according to Fang et al. [38], that the Internet promotes technological innovation among SMEs through the reduction of information asymmetry, the optimization of resource allocation, and the expansion of innovative openness. Furthermore, the acquisition and implementation of external knowledge are essential components of enterprise innovation, alongside the requisite financial backing. The utilization of big data technology can facilitate the acquisition of vast quantities of detailed information from numerous sources. These emerging technologies have not only enhanced quality and accelerated time-to-market, but have also contributed to more environmentally sustainable production processes [39]. Therefore, the fourth hypothesis addressed in this study is as follows.

Hypothesis 4 (H4). HTP results in a substantial and positive effect on GI.

SMEs have the imperative to embrace novel technology and enhance their operations in order to operate in a sustainable manner. Previous scholarly literature has indicated that advancements in technology have played a significant role in mitigating carbon emissions [40]. Drawing upon the viewpoints put forth by Zhao et al. [41], the concept of technical progress can be defined as the proficient application of energy resources in order to mitigate carbon emissions. Technological progress is considered the primary catalyst for increasing total factor production while adhering to carbon intensity limits. To achieve the goal of reducing carbon emissions, it is feasible to enhance the technological capabilities of industrial firms, thereby increasing their rate of resource utilization and decreasing their unit energy consumption. Furthermore, it has been shown that allocating resources towards research and development endeavors, specifically targeting energy conservation and emission reduction objectives, can effectively enhance environmental conditions and mitigate haze pollution. Hence, it can be argued that over time, advancements in technology have the potential to significantly improve energy efficiency and contribute to the mitigation of haze pollution. Therefore, the fifth hypothesis addressed in this study is as follows.

Hypothesis 5 (H5). HTP results in a substantial and positive effect on SOP.

As both an exogenous and endogenous force, GI has the capacity to foster sustained economic growth [42]. It is imperative for businesses to give precedence to GI due to its capacity to enhance customer satisfaction and mitigate environmental damage. GI is expected to increase output while decreasing resource consumption. Zhao et al. [43] conclude, based on their research into the effectiveness of GI in China, that GI is a crucial element in enhancing resource efficiency. GI is crucial for the development of a decarbonized economy and the sustainability of the industrial sector. It has the potential to transform the structure of the industrial sector from one dependent on fossil fuels to one that fosters sustainable economic expansion. Consequently, enterprises bolster their competitiveness and sustain their advantage through brand enhancement, market expansion, and adherence to environmental protection regulations. Environmental protection can be supported by businesses through the production of superior goods and services, as well as the implementation of innovative procedures designed to boost productivity and efficiency. In other words, GI is essential for businesses to increase their future competitiveness and maintain legitimacy. Therefore, the sixth hypothesis addressed in this study is as follows (Figure 1).

Hypothesis 6 (H6). GI results in a substantial and positive effect on SOP.

1.png

Figure 1. Hypothesized model

5.1 Assessment on measurement model

Content validity. Cross-loading, used to assess content validity, showed that questionnaire item meanings matched [54] ideas. This required a construct's value to be higher than the others in the same rows and columns [55]. The evaluation scale used in this study had perfect content validity, as shown in Table 1 for all assessed constructs.

Table 1. Results summary for the content validity

Convergent validity. Initially, the factor loading values need to be determined. According to Hair et al. [49], every factor in this study had a loading value greater than 0.70. Building on the perspectives of previous study [56], the value of outer loading between 0.4 and 0.7 can be retained when it is possible to enhance the AVE value.

In the second phase, the internal consistency reliability calculation is completed. The two most used methods in this regard are composite reliability (CR) and Cronbach's alpha. Exploratory research is to be carried out between 0.70 and 0.95 on the significant credibility scale, according to Hair et al. [49]. In this investigation, Cronbach's alpha (α) values fluctuate from 0.740 to 0.868, which is an appropriate range for reliability [49]. Alternatively, all CRs fluctuate from 0.822 to 0.919 are higher than 0.70, in appropriate with the suggestions of previous study [49]. Similarly, the rho_A (ρA) values are gauged, the obtained values of the hypothesized constructs fluctuate between 0.741 and 0.868, which is considerably above the cutoff point of 0.70 [57]. The following phase is to compute the AVE values. Accordingly, AVE values for the constructs in this investigation fluctuate from 0.436 to 0.813. Although AVE is less than 0.5, but CR is higher than 0.6, convergent validity of the construct is acceptable. Taken together, the statistical output demonstrated in Table 2 provides proof that the constructs used in this research are distinct in their composition.

Discriminant validity. In this research, the Fornell and Larcker criterion and the Heterotrait-Monotrait Ratio (HTMT) are used to examine the discriminant validity. Regarding this, it's important to note that the Fornell and Larcker criterion refers to how much variance a construct collects from its indicators, which must be more than the variance the construct shares with the rest [58]. Table 3 displays the formation of a diagonal with the highest values found in the data set.

The Heterotrait-Monotrait Ratio (HTMT) is the second measurement tool utilized to evaluate the discriminant variance. On each sample, all Heterotrait-Monotrait (HTMT) ratios were less than a more stringent threshold of 0.85 [59], indicating that discriminant validity was met. Accordingly, Table 4 demonstrates that all values are suitable and acceptable because they are less than 0.85.

Table 2. Results summary for convergent validity

Table 3. Results summary for discriminant validity on Fornell-Larker criterion

Table 4. Results summary for discriminant validity on Heterotrait-Monotrait ratio

5.2 Assessment on structural model

Based on the statistical results illustrated in Table 5 and Figure 2, all inner variance inflation factor (VIF) values are considerably below the critical threshold value of 3, indicating no serious collinearity problems [49]. The SRMR, also known as the standardized root mean square residual, quantifies the quality of model fit. According to literature [60], a satisfactory fit is characterized by a value below 0.08. The fit of the structural model for theory testing was found to be satisfactory, as evidenced by the SRMR value of 0.072, which falls below the established threshold of 0.08. Building on the bootstrap outcomes (percentile bootstrapping, two-tailed test, 0.05 significance level, with 10,000 resamples), DF substantially and positively affects SOP (β = 0.127; t-value = 4.926; p-value = 0.000); HTP (β = 0.195; t-value = 7.198; p-value = 0.000); GI (β = 0.291; t-value = 11.784; p-value = 0.000). In the same vein, HTP substantially and positively affects SOP (β = 0.218; t-value = 8.365; p-value = 0.000) and GI (β = 0.224; t-value = 8.700; p-value = 0.000). As anticipated by H3, the results reveal that GI demonstrates a direct positive impact on SOP (β = 0.207; t-value = 8.118; p-value = 0.000) In light of this, H1-H6 are supported. Alternatively, the mediating impact of HTP was evaluated. First, the significance of DF's indirect impact on SOP through HTP was evaluated. Given that the direct effects of DF on HTP are also supported and that the indirect effects are significant (t-value = 5.360; p-value = 0.000), it was determined that HTP partially mediates the association between DF and SOP [49]. Likewise, the indirect effect of DF on SOP through GI is significant (t-value = 6.506; p-value = 0.000), it is determined that GI partially mediates the association between DF and SOP [49]. As a result, the findings indicate partial mediation.

The R² was 0.160 for GI and for SOP while R² for HTP is 0.038. The f² effect size [49] was also employed to evaluate the strength of the structural model relationships. The analysis reveals that DF had a small effect size on HTP, SOP and GI (0.039, 0.017 and 0.097, respectively). In the same vein, HTP had a small effect size on GI and SOP (0.058 and 0.051, respectively). Additionally, GI had a small effect size on SOP (0.043).

When compared to a linear benchmark model, the model does not demonstrate good out-of-sample prediction ability for all construct indicators. The PLS_predict [61] results in Table 6 show that the RMSE obtained by the linear model (LM) is smaller than the RMSE obtained by PLS-SEM for almost all items (i.e., this results in favor of the LM prediction benchmark; the same finding holds for the mean average error). Furthermore, when employing the cross-validated predictive ability test (CVPAT) [62], the result analysis reveals that the LM has a reduced average loss than the PLS-SEM. On the other hand, PLS-SEM can outperform the naive indicator average (IA) prediction benchmark for both PLS_predict (as evidenced by the positive Q²_predict values) and CVPAT (as evidenced by the significantly negative average loss difference). As a result, the model obtains some predictive capability that permits it to pass the IA test but not the more conservative LM benchmark.

Table 5. Results of hypotheses testing

1.png

Figure 2. Structural model

Table 6. Predictive model assessment

As a consequence of the progressively volatile global landscape, SMEs are compelled to modify their existing methodologies in order to align with sustainability objectives. The implementation of a DF strategy for SMEs has been a longstanding goal of governments, especially in developing nations. This study aims to develop and validate a model that highlights the connection between DF and SOP. Significantly, the suggested model takes numerous efforts to enhance understanding of the mediation pathways between HTP and GI within its assumed structures. The analysis of the acquired outcomes highlights significant and positive relationships between the proffered constructs related to significance and effect size. Concretely, the outcome confirmed the significantly favorable correlation between DF and SOP. This association was partially mediated by both HTP and GI concurrently. The study's findings and insights would be valuable to practitioners seeking sustainable strategies to foster innovation in their enterprises. Conversely, the aforementioned observations would offer novel perspectives to professionals and policymakers, enabling them to formulate targeted approaches pertaining to HTP and GI, as well as establish legislative measures and regulatory frameworks for DF.

It is paramount to be aware of the limits because they will serve as the foundation for upcoming academic work. Utilizing cross-sectional data constitutes the principal obstacle that impedes the ability to draw definitive conclusions when analyzing results. Conversely, one could argue that conclusions regarding the research can only be drawn from the interrelationships between the variables under consideration since the study offers no definitive perspectives on compatibility. Therefore, it is advisable that forthcoming investigation integrate robust longitudinal analysis, secondary data sources, and modern statistical methodologies. Second, caution should be taken when generalizing because the particulars of the study environment may limit the conclusions drawn from the observations. The fact that all of the samples were sourced from Vietnam makes it necessary for more cross-regional research to confirm the findings before they can be generalized. Further analysis could be conducted by gathering additional data from developed and emerging markets so as to obtain more insightful information. It was also suggested that researchers conduct comparative analyses of developed and emergent nations in order to obtain more useful results. Third, the statistical information used in this study mainly relies on a self-report design. As a result, each SME receives the survey forms from just one person. Because individuals who have more favorable opinions of DF, HTP, and GI might be more inclined to fill out and submit their survey forms, the responses from single participants may be skewed. Therefore, future studies might also think about gathering the opinions of other pertinent stakeholders in this context. The convenience and snowball sampling procedures represent the fourth bottleneck and may have an impact on the study's generalizability. To ensure that the results are both representative and scientific, it is advised that future studies use the quota sampling method to collect sample data. This study's relatively small sample size is its fifth limitation, which calls for additional research. It would therefore be beneficial to have a wider geographic scope. Additionally, it is advised that more components be added to the model in upcoming research to produce a more accurate depiction of the problem.