Artificial Intelligence Approaches for Performance Prediction of Investment Companies in the Iranian Capital Market
نویسندگان
1 Department of Accounting, Sari Branch, Islamic Azad University, Sari, Iran
2 Department of Accounting, Semnan Branch, Islamic Azad University, Semnan, Iran
3 Department of Accounting, Sari Branch, Islamic Azad University, Sari, Iran
doi
چکیده
Investment companies listed on the Iranian Stock Exchange play a crucial role within the nation's capital market. Consequently, the development of a robust model for identifying, evaluating, and ranking these companies using artificial intelligence (AI) techniques offers a valuable approach to enhancing investment quality. This study employs AI models to assess and rank investment companies listed on the Iranian Stock Exchange. Specifically, an Adaptive Neuro-Fuzzy Inference System (ANFIS) algorithm is proposed. Input variables, encompassing company management characteristics, capital market financial metrics, internal control quality, industry factors, competitive landscape, management quality, credit and economic status, accrual financial ratios, cash financial ratios, and capital market ratios, are utilized. Each input variable undergoes a fuzzification process, involving the definition of membership functions. An ANFIS model is then constructed, utilizing these ten input variables—company management characteristics (Management), capital market financial variables (Financial), internal control quality (IQC), industry factors (Industry), competitive situation (Competitive), quality management (QM), credit and economic situation (CES), accrual financial ratios (AFR), cash financial ratios (CFR), and capital market ratios (CMR)—to predict the performance of investment companies (Performance) as the output variable. The performance of the ANFIS model is subsequently compared with that of a neural network. The results indicate that the ANFIS method outperforms the neural network.
