Development of a comprehensive model to predict stock prices in the stock market with an interpretive structural modeling approach

Abstract

The aim of this study was to develop a comprehensive model for stock price forecasting in the Tehran Stock Exchange (TSE) using a mixed Delphi-fuzzy approach of Interpretive Structural Modeling (ISM). The model incorporated technical, fundamental, macroeconomic, and emotional factors. In this study, the fuzzy Delphi method was employed to identify key criteria from the investor’s perspective among the 54 stock price prediction criteria extracted from the existing literature. Subsequently, the interpretive structural modeling method was utilized to examine the relationships between these criteria and establish a hierarchical model. The findings of the ISM revealed that the price per share and the money flow index held significant positions at the bottom of the hierarchy and exerted a strong driving force. Among the criteria occupying lower positions in the hierarchy, the exchange rate and the power indicator relative and exponential moving averages were identified as the most influential indicators.Keywords: Stock Market, Prediction Model, Fuzzy-Delphi Approach, Interpretive Structural Modeling (ISM) IntroductionStock markets play a crucial role in promoting economic growth by efficiently allocating resources and generating liquidity, particularly in urbanized smart cities. Stock market analysis encompasses both technical and fundamental approaches. Technical analysis relies on historical stock price data and technical indicators, making it particularly useful for short-term forecasts and trading strategies. On the other hand, fundamental analysis is based on the information about companies and the broader economy, focusing more on long-term forecasts and investment strategies (Latif et al., 2024). Combining technical and fundamental analysis can enhance long-term predictions.Empirical analysis of stock markets presents significant challenges. Multiple factors, such as market dynamics, industry trends, company performance, economic conditions, political events, and globalization, exert influence on stock markets. These factors are intricately interconnected, rendering the prediction of stock market behavior a complex task (Yang et al., 2020). Therefore, the primary objective of this research was to develop a framework for forecasting stock prices in the Tehran Stock Exchange (TSE) market. The proposed approach integrated a mixed Delphi-fuzzy methodology with Interpretive Structural Modeling (ISM), incorporating technical, fundamental, macroeconomic, and emotional factors into the analysis. Materials and MethodsThe sample consisted of two groups: investors and academic experts. The investors, who engaged in the stock market prediction and investment by analyzing technical, fundamental, and macroeconomic indicators, possessed a deep understanding of financial indicators. Due to their expertise in the field, they were considered as expert investors.To begin, a questionnaire comprising 54 initial criteria related to stock price prediction was distributed to the experts. The experts were asked to rate the importance of each criterion on a Likert scale ranging from very low (1) to very high (5). After consolidating the responses, 15 criteria were selected as significant for predicting stock prices in the stock exchange. Subsequently, a questionnaire was designed to explore the relationships between the key criteria identified. This questionnaire was administered to 10 university experts, who were requested to determine the relationships between the criteria based on theoretical foundations. Based on the responses received and employing the ISM method, the relationships between the criteria were examined and the fundamental criteria were identified. FindingsIn the initial phase of the fuzzy-Delphi method, the literature and previous studies were thoroughly examined to identify theoretical concepts related to stock price forecasting from 4 perspectives: technical, fundamental, macroeconomic, and behavioral perspectives. A questionnaire was then formulated to determine the priority and importance of various indicators. The indicators with the highest average importance were selected within each perspective. The results revealed that among the 54 indicators considered, 15 were identified as crucial criteria for stock price forecasting. These indicators included exponential moving averages, price channel indicator, relative strength indicator, equilibrium trading volume indicator, price indicator, price-to-earnings ratio (P/E), operating profit-to-sales ratio, gross profit-to-sales ratio, company sales growth rate, share purchase-to-sales ratio, dividend per share (DPS), money flow index (MFI), earnings per share (EPS), exchange rate, and trading volume. These indicators were deemed essential for accurate price forecasting in the stock market. Discussion & ConclusionThe findings highlighted the significance of the price-to-earnings ratio and the MFI as the most important criteria for predicting stock prices. In the ISM, these criteria were positioned at the highest level (Level 8) of the hierarchy, indicating their strong driving force and considerable influence. On the other hand, the exchange rate, relative strength indicator, and exponential moving average were identified as the most influential indicators at the top levels (1 and 2) of the hierarchy. These criteria, particularly the macroeconomic components and the relative strength indicator, played a pivotal role in the stock price prediction process.

Keywords

• stock market
• prediction model
• fuzzy-delphi approach
• interpretive structural modeling (ism)