نازنين علي بيگي بني

In recent decades, the concept of responsible an‎d sustainable investment (ESG) has evolved into one of the fundamental paradigms of the global financial system. Concerns related to climate change, corporate social responsibility, an‎d governance requirements have led investors, regulators, an‎d policymakers to pay greater attention to the non-financial impacts of corporate activities. Within this context, sustainable investment has gained prominence not only as an ethical orientation but also as a strategic tool for risk management, transparency enhancement, an‎d long-term value creation. However, the rapid proliferation of sustainability reports an‎d the increasing number of corporate claims regarding responsible performance have posed a crucial challenge: how to eva‎luate the credibility an‎d authenticity of such reports an‎d prevent falling into the trap of “greenwashing.” This study aims to develop an analytical an‎d empirical model for distinguishing genuine corporate sustainability from declarative claims an‎d integrating it into the investment decision-making process. The research employs a combination of Natural Language Processing (NLP) techniques an‎d multi-objective financial optimization models to design a framework capable of differentiating between actual sustainability performance an‎d rhetorical disclosure. In the first stage, a collection of corporate sustainability reports was compiled, an‎d their textual content was preprocessed an‎d analyzed. Using text-mining algorithms an‎d linguistic feature extraction, raw ESG scores were computed for each firm. In the second stage, through sentiment an‎d intensity analysis, the degree of exaggeration an‎d unverifiable claims within the reports was quantified, allowing the raw sustainability index to be adjusted according to the credibility level of each company’s disclosure. In the third stage, a multi-objective optimization model was developed, simultaneously considering the three main investment dimensions expected return, portfolio risk, an‎d adjusted real sustainability index. The model seeks to determine an optimal portfolio composition that achieves a rational balance between profitability an‎d social responsibility. The experimental results revealed that incorporating the qualitative dimension of report credibility leads to the construction of portfolios with more stable long-term returns an‎d lower risk. Moreover, the proposed model effectively reduces the impact of greenwashing an‎d improves the accuracy of firm rankings based on actual sustainability performance. From a theoretical perspective, this research contributes to the literature on sustainable investment by introducing a content-based adjusted sustainability measurement framework. Practically, the model can be applied by institutional investors, ESG funds, an‎d financial policymakers to achieve more comprehensive an‎d realistic eva‎luations of corporate sustainability performance. Additionally, the findings may promote greater transparency in capital markets an‎d strengthen public trust in corporate sustainability reporting. Overall, this study demonstrates that integrating qualitative content analysis of sustainability reports with quantitative investment models offers an innovative an‎d effective approach to responsible financial decision-making—positioning investment not merely as a means of economic return but as a driver of sustainable development an‎d corporate social responsibility.

سرمايه‌گذاري پايدار , شاخص‌هاي پايداري , سبزنمايي

Sustainable Investment , ESG Indicators (Environmental, Social, Governance) , Greenwashing

Nazanin Alibeigi beni

Seyed Jaafar Sadjadi