Smart Statistical Model Reveals What Drives Green Finance Success in the Big Data Era

Green finance has moved from the margins of corporate strategy to the center of how modern economies fund their transition away from carbon. Banks, investment funds, and accounting firms that specialize in environmentally oriented financial products now manage a rapidly expanding share of global capital, and the pressure to understand why some of these firms thrive while others stagnate has never been greater. A new study published in the Journal of Big Data tackles exactly that question, proposing a novel analytical framework called the Smart Partial Least Squares model to identify the factors that drive the performance and growth of green finance and accounting companies in an era defined by enormous data flows.

The research, led by Hua Han of Baiyin Hope Vocational and Technical College in China together with Ali Ehsani, Nader Naghshbandi, Karlo Abnoosian, Jinlu Liu, and Mahdi Darbendi, was published as an open access article on 5 October 2026. Its central premise is deceptively simple: sustainability in the financial sector means adopting strategies and practices that benefit both the environment and society while still ensuring profitability and financial growth. Companies that genuinely adhere to sustainability principles, the authors argue, manage their resources more efficiently and prove more resilient to environmental and economic shocks. But proving that premise rigorously requires a statistical apparatus capable of handling the tangled web of variables that influence corporate performance.

That is where partial least squares comes in. Partial least squares is a family of statistical methods that sits somewhere between multiple regression and principal component analysis. Rather than asking which individual variables predict an outcome, it constructs new composite dimensions, called latent components, that capture the maximum shared variance between a block of predictor variables and a block of response variables. This makes it especially well suited to situations where predictors are numerous, highly correlated, and potentially outnumber the observations available, conditions that routinely arise in financial and accounting research where economic indicators, regulatory measures, and technology metrics tend to move together.

The Smart Partial Least Squares model proposed in the study extends this classical machinery with an explicitly intelligent, data-driven orientation designed for the big data environment. The authors describe it as a model that considers both technology-based changes and managerial perspectives when evaluating the key performance drivers of green finance firms. In practical terms, the framework organizes the determinants of company performance into internal and external categories. Internal factors comprise technological variables and management variables, reflecting the firm’s own capabilities in information and communication technology and the quality of its managerial practices. External factors comprise economic variables and legal and regulatory variables, capturing the market conditions and policy environment in which green finance companies operate.

Big data technologies are not merely the backdrop of the study; they are one of its substantive subjects. The authors emphasize that analyzing large data sets through big data technologies enables companies to collect extensive environmental, social, and economic information, which in turn supports more informed decisions about sustainable practices. Green finance itself is characterized in the paper as a relatively new concept that offers financing options to individuals, corporate entities, and governments willing to fund green activities or low-carbon initiatives. The use of big data in this domain facilitates a more accurate assessment of the investment opportunities and risks associated with green and low-carbon projects, a task that traditional financial analysis has historically struggled to perform because environmental performance data are heterogeneous, unstructured, and voluminous.

Machine learning and artificial intelligence enter the picture as the analytical engines that can convert this data abundance into insight. The study notes that big data technologies such as machine learning and artificial intelligence can be leveraged to gain a deeper understanding of how the identified drivers affect green finance and accounting firms. This matters because the relationship between, say, the intensity of ICT adoption and firm growth is unlikely to be linear or uniform across firms. Intelligent methods can detect interactions, threshold effects, and nonlinear patterns that conventional regression approaches may miss, and embedding them within a partial least squares structure allows researchers to retain the interpretability of latent-variable modeling while exploiting the pattern-recognition power of modern algorithms.

The results of the analysis deliver a clear and consequential message: both internal factors and external factors significantly contribute to the growth and efficiency of green finance firms. Technological variables and management variables on the inside, and economic variables together with legal and regulatory variables on the outside, all register as significant determinants of performance. Perhaps more striking is the finding about how these categories interact. The research shows that the interaction between internal and external factors is effective in creating a positive cycle between sustainability and green financial performance, a feedback loop that can lead to sustainable, long-term growth for companies operating in competitive markets. In other words, good environmental practice and good financial performance are not competing objectives to be traded off; under the right technological and regulatory conditions, they reinforce one another.

For practitioners, the most immediately actionable conclusion concerns information and communication technology. The authors state plainly that managers should note that any improvement in ICT application may contribute meaningfully to the overall performance of green finance and accounting firms. That recommendation carries weight in an industry where the ability to ingest, clean, and analyze environmental data streams increasingly determines which firms can credibly price green bonds, evaluate the climate risk of loan portfolios, or verify the sustainability claims of corporate clients. Firms that underinvest in their data infrastructure may find themselves not only operationally slower but also less trusted by investors who demand rigorous, data-backed evidence of environmental impact.

For policymakers, the significance of the external variables is equally pointed. Because legal and regulatory variables emerge as significant contributors to firm performance alongside economic conditions, the study suggests that the regulatory architecture surrounding green finance is not a neutral backdrop but an active ingredient in the sector’s success. Clear rules, credible standards, and supportive economic conditions appear to work in concert with firms’ internal technological and managerial capabilities, and the positive sustainability-performance cycle identified by the model depends on that cooperation. Policymakers seeking to accelerate the flow of capital toward low-carbon projects therefore have a statistical argument for strengthening both the economic incentives and the legal frameworks that green finance companies navigate daily.

The study also makes a methodological contribution to the academic literature. By integrating technological, managerial, economic, and legal-regulatory variables into a single smart partial least squares framework, and by explicitly incorporating the role of big data in optimizing green finance strategies, the authors offer a template that other researchers can adapt to related questions in sustainable finance, financial econometrics, and fintech innovation. The work arrives at a moment when the volume of environmental, social, and governance data is growing faster than the analytical tools available to interpret it, and it demonstrates that hybrid approaches, combining classical latent-variable statistics with the intelligence of machine learning, can extract decision-relevant signals from that flood. As green practices continue to gain traction across the financial industry, identifying the factors that explain the success of green finance will remain crucial, and this study provides both a conceptual map of those factors and a technical instrument for measuring them. The research received no external funding, and the authors declare no competing interests.

Subject of Research: A smart partial least squares model identifying factors driving green finance and accounting company performance using big data

Article Title: A novel smart partial least squares-based model for analyzing key factors driving green finance and accounting companies’ performance in big data era

Article References: Han, H., Ehsani, A., Naghshbandi, N., Abnoosian, K., Liu, J., & Darbendi, M. (2026). A novel smart partial least squares-based model for analyzing key factors driving green finance and accounting companies’ performance in big data era. Journal of Big Data. https://doi.org/10.1186/s40537-026-01575-6

Image Credits: AI Generated

DOI: 10.1186/s40537-026-01575-6

Keywords: green finance, big data, partial least squares, sustainability, machine learning, artificial intelligence, accounting firms, ICT adoption, regulatory variables, financial performance, low-carbon investment, Journal of Big Data