Iranian Journal of Finance

Abstract Two predominant methods for analyzing financial markets have been technical and fundamental analysis. However, the emergence of the Internet has altered the trading landscape. The availability of Internet and social media access plays a moderating role in information asymmetry, resulting in investors making informed decisions. Social media has turned into a source of information for investors. Through diverse communication channels on social media, investors articulate their perspectives on whether to buy or sell a stock. According to Surowiecki, the collective opinions gathered through social media frequently offer better predictions than individual opinions, a phenomenon referred to as the Wisdom of the Crowd. The wisdom of the crowd stands as an essential measure within social networks, with its potential to reduce errors and lessen information-gathering costs. In this study, we tried to evaluate the wisdom of the crowd's potential to improve stock price prediction accuracy. So, we developed a prediction model by Long Short-Term Memory based on the wisdom of the crowd. Users’ opinions in Persian about the Tehran Stock Exchange (TSE) stocks were collected from SAHMETO for eight months. The Support Vector Machine classified them into buy, sell, and neutral classes. During the research period, people mentioned 823 stocks, and 52 stocks with over 100 signals were chosen. The results of the study show that although the model presented has achieved an acceptable level of accuracy, correlations between the actual and predicted values exceeded 90%. The accuracy metrics of the proposed model compared to the base model were not improved.

Abstract Herd behavior, the tendency of individuals to mimic the actions of a larger group, significantly impacts capital markets by influencing stock prices, market liquidity, and overall market stability. This phenomenon has garnered significant attention in financial studies due to its implications for both institutional and individual investors, contributing to increased market volatility and potential crashes. Various methodologies have been developed to assess herd behavior, revealing its presence across diverse market conditions, including periods of high distress and volatility. This study examines macro herding in the Tehran Stock Exchange from March 2016 to February 2024, using weekly asset returns to measure herd behavior among listed companies. For the first time in Iran, we employ the TV method to calculate herding. The TV method offers two primary advantages: it is adept at identifying macro herding because it captures the collective trading direction of investors, and it operates independently of asset pricing models, minimizing biases associated with those models. Focusing on the collective trading direction, we aim to detect significant deviations in stock price movements indicative of herd behavior. Our findings indicate that herd behavior is more pronounced during extreme market conditions, both positive and negative, with a particularly notable increase during periods of negative market returns. This study provides insights into the dynamics of investor behavior in the Tehran Stock Exchange, highlighting the importance of monitoring such behavior to mitigate its potential adverse effects on market stability.

Abstract Embedded options are virtually new instruments identical to options in many aspects except their non-tradable nature. Testing the efficiency of the Variance Gamma and Black-Scholes-Merton model on these instruments would provide a vision of transitioning from the classical model with its deficiency to more intricate models. Considering the complicated nature of the Variance Gamma stochastic process to price options, the Fast Fourier Transform (FFT) method is used in conjunction with the Nelder-Mead Simplex method to calibrate models. This research uses the Fast Fourier Transform (FFT) to price four embedded options with the ticker symbols Hefars912, Heghadir912, Heksho208, and Hetrol911 under the two models. The result approves that the Variance Gamma process is more efficient than the Black-Scholes-Merton model in pricing embedded options. Consequently, the variance gamma process would generate fewer errors in pricing those options that can be used in a practical sense.

Abstract Portfolio allocation with Deep Reinforcement Learning (DRL) has been the focus of many researchers. In investing, a portfolio optimization strategy is selecting assets that maximize return on investment while minimizing the risk. Asset optimization involves balancing risk and return, where stock returns are profits over time, and risk is the standard deviation value of the asset's return. Many of the existing methods for portfolio optimization are essentially the expansion of diversification methods for assets in the investment. Signiant drawdowns and early entry into the share are still challenging in portfolio construction. The idea is that having a portfolio based on net money flow is less risky than allocating a portfolio based on historical data only and turbulence as risk aversion. This paper proposes a profitable stock recommendation framework for portfolio construction using the DRL model based on the net money flow (MNF) indicator. We develop a new risk indicator based on the intelligent net-flow behavior of smart money to help determine the optimal market timing for buying and selling. The experimental results of real-world trading scenario validation show that the model outperforms all the considered baselines and even the conventional Buy-and-Hold strategy. Moreover, in this paper, the effect of defining different environments made of various information with hyper parameter optimization on the performance of models has been investigated, and the performance of DRL-driven models in different markets and asset positions has been investigated. The empirical results show the dominance of DRL models based on MNF indicators.

Abstract Portfolio optimization has always been the main concern of investors. What differentiates different optimization models from each other is the risk measure. The main contribution of this paper is to provide a portfolio optimization model that considers systemic risk so that it can help investors make optimal investment decisions as a general model. For this purpose, two models are presented. In the first model, systemic and systematic risk were considered simultaneously, and in the second model, only systemic risk was considered. In the two mentioned models, delta conditional value at risk (∆CoVaR) and the Markowitz model are used respectively to measure systemic risk and a benchmark model. Also, the criteria used to compare the performance of the reviewed models include the ratio of reward-to-risk, along with the Sortino ratio and the Omega ratio. The problem of optimization and examination of the results was carried out on a selected sample, 38 companies listed in the Tehran Stock Exchange (TSE) from 2013 to 2023. The results of empirical analysis of out-of-sample data (during a period of 1198 days) show that based on all three mentioned criteria, the first proposed model shows the best performance among the three models. In addition, the performance of the second model is ranked second. In short, it can be said that considering systemic risk in portfolio optimization leads to better performance than the Markowitz model.

Abstract The accurate imputation of missing values in time series data is paramount for maintaining the integrity and reliability of analyses and predictions. This article investigates the effica-cy of various missing values imputation methods, encom-passing well-known machine learning and statistical tech-niques. Moreover, for a better understanding, they imple-mented two financial data time series: S&P 500 and Bitcoin markets spanning from 2016 to 2023 on a daily frequency. Initially utilizing complete datasets, controlled missingness was introduced by randomly removing 45 data points. Then, these methods applied multiple imputation strategies for estimating and substituting these missing values. Experi-mental evaluation yielded insightful findings regarding the performance of the different methods. The examined ma-chine learning methods, including k-Nearest Neighbors (k-NN), Random Forest, Deep Learning, and Decision Trees, consistently outperformed their statistical counterparts, such as Mean Imputation, Regression Imputation, Hot-Deck Im-putation, and Expectation-Maximization Imputation. Nota-bly, Random Forest emerged as the most effective method, showcasing superior performance in terms of accuracy and robustness. Conversely, the Mean Imputation method exhibited com-paratively inferior outcomes, suggesting its limited suitabil-ity for financial time series data. This research contributes to the ongoing discourse on data integrity within finance ana-lytics and serves as a comprehensive guide for practitioners seeking optimal missing values imputation methods. The empirical evidence provided herein advances the under-standing of imputation techniques' relative performance and their application in financial data, facilitating enhanced de-cision-making processes and yielding more reliable predic-tions.

Abstract Women have been trying to gain independence throughout history. In recent years, advances in technology and business have helped women to achieve this goal. According to women's personality and psychological characteristics, there are differences in their trading styles. One of the factors influencing the choice of this type of strategy is stress. In the last few years, stress and anxiety caused by Corona have become epidemic. In order to test the hypotheses, women traders active in the financial markets of Iran were examined using a Likert questionnaire in 2022, and interesting results were obtained. In order to carry out the research of this study, an interview was conducted first to find suitable questions and validity. Then, the statistical population and sample were selected, and the final questionnaire was distributed among them. MATLAB software was used to identify the number of common descriptive characteristics of the respondents, and finally, using EViews software, statistical analysis related to hypothesis testing was performed. The result shows that women play more conservatively and are risk-averse during the period of coronavirus infection. It has no effect on the volume and capital used in the transaction. The Corona anxiety has significant effects on the three dependent variables of conservatism, trading style, and trading (volume, capital, and number of transactions).