دانشگاه اصفهان مدیریت دارایی و تامین مالی 2383-1189 14 2 2026 06 22 Designing an Optimal Decision-Making Model for Investors: Integrating Artificial Intelligence and Financial Reporting Transparency ارائۀ الگوی بهینۀ تصمیم‌گیری سرمایه‌گذاران با به‌کارگیری هوش مصنوعی و با تأکید بر شفافیت گزارشگری مالی 1 24 29331 10.22108/amf.2025.143308.1934 FA سحر عباس حسن دانشجوی دکتری، گروه حسابداری، دانشکده اقتصاد و مدیریت، دانشگاه ارومیه، ارومیه، ایران پرویز پیری دانشیار، گروه حسابداری، دانشکده اقتصاد و مدیریت، دانشگاه ارومیه، ارومیه، ایران پری چالاکی استادیار، گروه حسابداری، دانشکده اقتصاد و مدیریت، دانشگاه ارومیه، ارومیه، ایران Journal Article 2024 11 11 Transparency and comparability of financial information constitute fundamental pillars of accountability and informed economic decision-making. With the rapid evolution of artificial intelligence (AI), this technology has significantly influenced financial and investment decision-making processes. This study seeks to design an optimal model for investor decision-making by exploring the integration of AI, with a particular focus on enhancing financial reporting transparency. A mixed-methods approach was adopted to achieve this objective. In the initial phase, qualitative data were gathered through in-depth interviews with 12 experts specializing in financial management and capital markets. Content analysis was employed to identify key themes and construct a preliminary model. To validate the model, a structured questionnaire was developed and distributed to 214 experts, yielding 200 valid responses. The questionnaire's reliability and validity were rigorously confirmed. Subsequently, structural equations modeling (SEM) was utilized to analyze the interrelationships among the variables. The findings revealed that all components of the proposed model, supported by both expert validation and statistical analysis, are robust and serve as critical factors in investor decision-making. Six key factors—financial, economic, political, psychological, individual, and artificial intelligence—were identified and validated as primary determinants influencing investment decisions. This study contributes to the literature by innovatively integrating AI with financial reporting transparency to enhance investment decision-making processes. The proposed model, which synthesizes these influential factors, offers novel perspectives for analyzing financial data and assists investors in achieving superior returns while mitigating risks. This framework can serve as a strategic tool for financial managers and stakeholders, enabling more precise and data-driven decision-making in complex financial environments. Keywords: Investor Decision-Making, Artificial Intelligence, Financial Reporting Transparency, Content Analysis JEL Classification: G11, C56, M41 Introduction Artificial intelligence (AI), a prominent branch of computer science, provides a robust framework for emulating human cognitive functions, including learning, pattern recognition, and decision-making. By harnessing advanced machine learning algorithms, neural networks, and natural language processing (NLP), AI systems are capable of analyzing extensive volumes of financial data with remarkable accuracy and efficiency. In this context, AI outperforms conventional methodologies by enabling the simultaneous processing of diverse data sources. Through the identification of hidden patterns and the prediction of market fluctuations, AI significantly enhances the quality of investor decision-making, particularly in the financial markets of developing countries. Key applications of AI in finance encompass the development of risk assessment models, the detection of fraudulent transactions, and the optimization of investment portfolios. Furthermore, by analyzing sentiment data extracted from news outlets and social media platforms, AI empowers investors to account for shifts in market sentiment, as well as economic and social dynamics. The integration of information technology systems not only improves the transparency of financial reporting but also ensures access to precise and timely information, thereby bolstering investor confidence and refining the decision-making process. This study, through a comprehensive review of prior research and the identification of existing gaps, seeks to propose an integrated model grounded in AI and financial reporting transparency. The primary objective of this model is to enhance the accuracy of financial forecasts and elevate the quality of investor decision-making. In this endeavor, critical investment factors such as risk, return, and liquidity are examined in conjunction with big data analytics to deliver innovative solutions that substantially strengthen investor trust in financial information and market trends. Method This study adopts a mixed-methods research design, comprising an initial qualitative phase followed by a quantitative phase. The qualitative phase utilizes content analysis, while the quantitative phase employs a correlative-survey methodology. The target sample for the qualitative phase consisted of professors and experts in financial management, as well as capital market specialists, all of whom possessed a foundational understanding of artificial intelligence (AI). Participants were selected through purposive sampling, and in-depth, semi-structured interviews were conducted. This approach, owing to its inherent flexibility, facilitated a thorough exploration of expert perspectives and the extraction of rich, nuanced data. The use of semi-structured interviews enabled the study not only to gather insights into the primary variables but also to uncover novel perspectives, previously unidentified factors, and latent relationships among variables. This contributed to the development of a robust theoretical framework and yielded findings with significant practical implications. The sampling process continued until theoretical saturation was achieved. Theoretical saturation refers to the point at which, following initial discoveries, the researcher continues data collection until the relationships between the main categories and subcategories become clear and meaningful. This process persisted until the researcher determined that additional interviews with experts (totaling 12 participants) no longer yielded new information. Table 1 provides a detailed overview of the demographic characteristics of the interviewees. Table (1) Demographic Characteristics of Interviewees No. Organization Position Education Gender 1 Capital Market Senior Manager Master of Accounting Male 2 University Professor Ph.D. in Accounting Male 3 University Professor Ph.D. in Financial Management Female 4 University Professor Ph.D. in Accounting Female 5 Capital Market Specialist Master of Finance Male 6 Capital Market Specialist Master of Accounting Male 7 Capital Market Middle Manager Ph.D. in Finance Male 8 University Professor Ph.D. in Accounting Male 9 Capital Market Accountant Master of Accounting Female 10 Capital Market Accountant Ph.D. in Finance Male 11 University Professor Ph.D. in Accounting Male 12 Capital Market Middle Manager Master of Financial Sciences Male The sample for the quantitative phase comprised all investors, buyers and sellers of stocks and other securities, and active participants in the stock market. A convenience sampling method was employed, guided by the structural equations modeling (SEM) formula, which recommends a minimum of 20 samples per latent variable and an overall minimum sample size of 200. Quantitative data were collected using a questionnaire consisting of 67 items measured on a 5-point Likert scale. Face validity was evaluated through consultations with 10 experts, and necessary revisions were implemented based on their feedback. In the qualitative phase, semi-structured interviews were conducted through in-person meetings, phone calls, and email correspondence. All interviews were recorded (with the explicit consent of participants) and subsequently transcribed and analyzed in detail. The credibility and validity of the findings were rigorously assessed using Patton’s evaluation criteria, which include credibility, transferability, and confirmability. Additional techniques such as triangulation through multiple sources, negative case analysis, and methodological flexibility were employed to enhance the robustness of the findings. The reliability of the model was evaluated using Kappa statistics, which involved comparing independent coding results to ensure consistency. For the quantitative data analysis, partial least squares (PLS) analysis was conducted using SmartPLS software. The analysis was divided into two components: the measurement model, which assessed the relationships between individual items and their corresponding dimensions, and the structural model, which examined the relationships between latent variables. Convergent validity was evaluated using the Average Variance Extracted (AVE), while discriminant validity was assessed through the Fornell-Larcker criterion and cross-loading tests. The reliability of the questionnaire was measured using Cronbach’s alpha, which yielded a value exceeding 0.7, indicating strong internal consistency. Findings In the initial phase of the study, data were gathered through semi-structured interviews with 12 experts to identify the key factors influencing investor decision-making in the context of artificial intelligence (AI) utilization. Content analysis was employed to analyze the interview data. Following an in-depth familiarization with the data, initial coding was conducted, yielding 67 preliminary codes. These codes were either explicitly or implicitly reflected in existing conceptual models. Subsequently, axial coding was performed to group related concepts and elucidate the relationships between categories. The outcomes of the initial and axial coding processes were systematically organized into a table of concepts pertinent to investor decision-making. The Kappa coefficient of 0.81 demonstrated excellent reliability of the model. Additionally, a two-round fuzzy Delphi approach was utilized to screen the identified indicators, with all indicators achieving scores above 7, confirming their relevance and validity. In the quantitative phase, the results of Cronbach’s alpha, composite reliability, and Average Variance Extracted (AVE) confirmed the model’s acceptable convergent validity and reliability. Discriminant validity was assessed using the Fornell-Larcker criterion, which revealed that the square root of the AVE for each construct exceeded its inter-construct correlations, further validating the model. Within the structural model, standardized path coefficients, t-statistics, and effect sizes (f²) were calculated for various paths, all of which indicated statistically significant relationships at the 99% confidence level. The predictive power of the structural model for investor decision-making was established using the Q² and R² indices, as well as the Goodness of Fit (GOF = 0.60 > 0.35). The analysis of the relationships between variables demonstrated that, in addition to the direct effects of economic, market psychology, political, individual, financial, and AI factors on decision-making, financial reporting transparency—as a moderating variable—also exerted a significant influence. Consequently, the evaluated model exhibited robust explanatory and predictive capabilities for investor behavior regarding AI adoption, supported by both theoretical and empirical evidence. Discussion and Conclusion Corporate investment, as a primary source of cash flow and a catalyst for economic development, plays a pivotal role in fostering long-term value creation. Access to accurate, transparent, and timely financial information has become increasingly critical in the investment decision-making process. Emerging technologies such as artificial intelligence (AI), with their exceptional processing capabilities and advanced analytical tools, offer more precise insights and mitigate uncertainties. This study seeks to propose an optimal model for investor decision-making by leveraging AI, with a particular emphasis on financial reporting transparency, as there is currently a lack of comprehensive models in the country that analyze the interplay of these factors. During the research process, 67 initial codes were extracted from in-depth interviews with experts and subsequently categorized into six primary themes: financial, economic, political, market psychology, individual, and AI factors. The findings, corroborated by prior studies, demonstrate that the use of AI significantly enhances the transparency and accuracy of financial reporting, streamlines investment decisions, and improves risk management. Numerous studies have confirmed the advantages of AI in increasing operational efficiency, predicting market trends, and optimizing portfolio selection. While AI holds transformative potential for financial decision-making, realizing this potential requires robust technical infrastructure, standardized data frameworks, and access to up-to-date information. Establishing regulatory frameworks, fostering international collaborations, and investing in workforce training are also essential for the successful implementation of this technology. In addition to proposing a comprehensive model, this study highlights existing challenges and limitations, suggesting that a gradual, context-sensitive implementation of AI can enhance investor confidence and contribute to the development of more transparent financial markets. شفافیت و قابلیت مقایسۀ اطلاعات مالی از ارکان اصلی پاسخ‌گویی و تصمیم‌گیری‌های اقتصادی آگاهانه محسوب می‌شود. با پیشرفت سریع هوش مصنوعی، این فناوری تأثیرات چشمگیری بر تصمیمات مالی و سرمایه‌گذاری داشته است. پژوهش حاضر با هدف طراحی مدلی بهینه برای تصمیم‌گیری سرمایه‌گذاران، استفاده از هوش مصنوعی را با تأکید بر شفافیت گزارشگری مالی بررسی کرده است؛ بنابراین، این پژوهش ابتدا با استفاده از مصاحبه‌های عمیق با ۱۲ متخصص در حوزۀ مدیریت مالی و بازار سرمایه، داده‌ها را جمع‌آوری و سپس ازطریق تحلیل محتوای کیفی، الگوی اولیه استخراج کرد. به‌منظور ارزیابی مقبولیت الگو، پرسش‌نامه‌ای تهیه و به ۲۱۴ صاحب‌نظر ارسال شد که ۲۰۰ نفر به آن پاسخ دادند. اعتبار و قابلیت اعتماد پرسش‌نامه به‌طور موفقیت‌آمیزی تأیید شد. سپس با استفاده از روش مدل‌سازی معادلات ساختاری، روابط بین متغیرها تحلیل شد. نتایج نشان داد که تمامی اجزای الگوی پیشنهادی با تأیید خبرگان و نتایج آماری به‌عنوان عوامل کلیدی در تصمیم‌گیری سرمایه‌گذاران قابل‌قبول است. شش عامل مالی، اقتصادی، سیاسی، روانی، فردی و هوش مصنوعی به‌عنوان عوامل اصلی تأثیرگذار بر تصمیمات سرمایه‌گذاری شناسایی و تأیید شدند؛ بدین‌تریتب، این پژوهش به‌طور نوآورانه هوش مصنوعی و شفافیت گزارشگری مالی را برای بهبود تصمیم‌گیری‌های سرمایه‌گذاری ترکیب می‌کند. مدل ارائه‌شده با ترکیب عوامل کلیدی اثرگذار، دیدگاه‌های جدیدی برای تحلیل داده‌های مالی فراهم می‌کند و به سرمایه‌گذاران کمک می‌کند تا با کاهش ریسک، بازدهی بهتری کسب کنند. این چارچوب می‌تواند به‌عنوان ابزاری راهبردی توسط مدیران مالی و ذی‌نفعان برای تصمیم‌گیری دقیق‌تر و مبتنی‌بر داده‌های معتبر استفاده شود.

https://amf.ui.ac.ir/article_29331_7d3e9c102dbcef8c583c412005713819.pdf