genetic algorithm

The purpose of this paper is to present a new technique to the portfolio selection using Genetic Algorithm and Fuzzy Synthetic Evaluation. Portfolio selection is a multi-objective/criteria decisionmaking problem in financial management. The proposed approach (Genetic Algorithm and Fuzzy Synthetic Evaluation) solves the problem in two stages. In the first stage، by using genetic algorithm and fuzzy synthetic evaluation، weight of criteria will be calculated. In second stage، using Fuzzy Synthetic Evaluation، Portfolios will be prioritized. A multi objective genetic algorithm is used to determine return and risk in the efficient frontier in Tehran stock market. In this research, we have used of firms’ performance between 1396-1400 in civil engineering, construction, investment and construction materials and tools manufacturers in order to determine portfolio selection. The main advantage of proposed approach is helping an investor to find a portfolio which have Best performance، portfolio selection doesn’t rely to expert knowledge

The optimization of asset portfolios, taking into account market advancements, has emerged as a pivotal subject in financial economics. Constructing asset portfolios is acknowledged as a critical decision for investors. Consequently, researchers focus on identifying factors that influence the selection of portfolios with high returns and controlled risk. Portfolio optimization, a cornerstone of financial economics, has gained prominence in the face of ever-evolving market dynamics. Investors' decision-making plays a pivotal role in portfolio construction, prompting researchers to explore factors that influence the selection of portfolios with high returns and controlled risk. Numerous models have addressed the optimization problem of stock portfolio management, each tailored to specific conditions and constraints. This research focuses on developing a multi-objective optimization model that incorporates investor memory and behavioral preferences. The literature on portfolio optimization is vast and diverse, encompassing various approaches and methodologies. Traditional optimization models, such as Markowitz's Mean-Variance model, aim to maximize expected returns while minimizing risk. However, these models often fail to capture the complexities of real-world investment decisions, which are often influenced by behavioral factors. Investor memory refers to the tendency of investors to base their current investment decisions on past experiences. This can lead to biases and suboptimal outcomes. Behavioral preferences, on the other hand, encompass a range of psychological factors that influence investor behavior, such as risk aversion, overconfidence, and herding.

We perceive stock portfolio optimization as a multi-objective challenge, considering two primary criteria. The first criterion involves investor memory, which encompasses utilizing historical price data and market trends to anticipate future performance. The second criterion pertains to behavioral preferences, which involves integrating investor risk aversion, overconfidence, and herding behavior into the model. We employ a Genetic Algorithm (GA) to optimize portfolios under both criteria. GA is a robust optimization technique that can effectively handle complex problems with multiple constraints. The study population comprises companies listed on the Tehran Stock Exchange for the year 2021. 

The achieved results suggest that investor memory serves as a more suitable criterion for optimal portfolio construction compared to behavioral preferences because investor memory incorporates market data and trends, providing a more objective basis for decision-making. Additionally, incorporating market return alongside investor memory data yielded superior results than using behavioral preferences with market return. This indicates that the combination of investor memory and market data can lead to more efficient and profitable portfolios. Pairwise comparisons of portfolios created using investor memory and behavioral preference criteria revealed that investor memory consistently outperformed behavioral preferences across different risk levels. This finding highlights the importance of considering investor memory when constructing optimal stock portfolios.

This study contributes to the literature on portfolio optimization by demonstrating the effectiveness of incorporating investor memory and behavioral preferences into the decision-making process. The findings suggest that investor memory is a more suitable criterion for portfolio optimization than behavioral preferences. Moreover, the combination of investor memory and market return data can lead to more efficient and profitable portfolios. The findings of this study have important implications for investors and portfolio managers. Investors should consider incorporating investor memory into their decision-making process when constructing stock portfolios. Additionally, portfolio managers can use the proposed multi-objective optimization model to create more efficient and profitable portfolios for their clients. This study provides a foundation for future research on portfolio optimization. Future studies can explore other factors that influence investor behavior, such as social media sentiment and news sentiment. Additionally, researchers can investigate the application of other optimization techniques, such as machine learning algorithms, to portfolio optimization.

The most important goal of every investor in the stock market is to increase returns and reduce investment risk. Therefore, the purpose of this research is to analyze the effectiveness of meta-heuristic algorithms in stock portfolio optimization. Considering that in this research, the past performance of Tehran Stock Exchange companies is examined in past studies from 1390-1399, therefore, in terms of the research design, this research was post-event using Delphi and meta-analysis techniques. The statistical community of this research Academic researchers in the field of finance and active in the Tehran Stock Exchange, and the sampling method in this research was targeted with a volume of 30 people. The data collection tool was a researcher-made questionnaire. The method of collecting information was structured interview of researchers and review of the results of various studies in the field of determining the optimal stock portfolio in Tehran Stock Exchange. In order to analyze the data, Spss software version 23 and Laserl version 5.7 were used. The results showed that among meta-heuristic algorithms of genetic algorithm, ant colony and bee colony are the most suitable tools with the aim of not stopping at local optimal points and not premature convergence. Finally, after evaluating the appropriate algorithms, a comparison of the average risk and returns of the stock portfolio in genetic algorithms, ant colony and bee colony was done in the study unit, they showed that in terms of the criteria of reducing the risk of genetic and bee algorithms and in terms of increasing the return of the optimal portfolio Stock bee algorithm has worked more efficiently.

Profit management has been one of the most controversial topics in recent research. Most research on earnings management has examined the linear relationship between independent variables and earnings management using statistical methods but they did not use these variables to predict earnings management. Today, with the growth of information technology and the introduction of artificial intelligence, including artificial neural networks into the field of scientific research, it has become possible to study nonlinear relationships between variables. In this study, an attempt was made to estimate optional accruals for predicting earnings management using artificial neural networks. Also in this research, the genetic algorithm optimizer model and Particle swarm has been used to optimize the weights of the artificial neural network model to enhance the predictive power. Then, the ability to predict profit management was evaluated using the modified Jones statistical model, artificial neural network and the combined technique of genetic algorithm, Particle swarm and neural network. The sample used in this study included 150 companies listed on the Tehran Stock Exchange between 2015 and 2020. Findings showed that the artificial neural network has a high ability to predict profit management, compared to the modified Jones linear model. The findings also indicate that the accuracy and ability of the combined model of genetic algorithm, particle swarm and neural network in predicting profit management is higher than the combined model of genetic algorithm-artificial neural network.

Earnings and earnings per share information are metrics that are considered important by many users; Therefore, companies try to attract investors with the most accurate forecast of earnings per share. On the other hand, despite the various methods of forecasting earnings, accurate forecasting of earnings per share in the financial field is not easy and most researchers are trying to determine the best way to forecast earnings; Therefore, the main purpose of this study is to investigate the accuracy of support vector machine based on genetic algorithm compared to common linear methods in forecasting earnings per share. For this purpose, samples consisting of 100 companies listed on the Tehran Stock Exchange during the years 2008-2019 have been studied. In order to achieve the objectives of the research, first by studying previous research in the field of earnings forecasting, 14 financial ratios affecting earnings forecasting have been selected. Then, in order to provide a model for predicting the profitability of companies, a combined model of support vector machine based on genetic algorithm, support vector machine and linear regression is compared. The results showed that the hybrid model of support vector based on genetic algorithm is much better in predicting the trend of earnings per share and has a higher accuracy compared to the model of support vector based on kernel functions and linear regression method. Thus, with the development of the support vector machine model based on the genetic algorithm, the model training error is reduced to 0.036 and the accuracy of the model is increased up to 75%.

aim of this study was to evaluate the integrated risk of the banking system through the metaphysical algorithms of gray wolf, genetics and particle swarming. This research is applied research in terms of purpose and correlational in nature and method. Data collection has been done through library studies, articles and sites in deductive form and data collection to refute and confirm hypotheses inductively. The statistical population of this research is the banking system and the sample includes banks listed on the Tehran Stock Exchange during the fiscal years 1392 to 1397. In order to collect the required data, the financial database of the Ministry of Economic Affairs and Finance, codal site, etc. have been used. After extracting the information, and setting them in the form of an integrated risk model, the objective function and constraints are entered in MATLAB software and the variables of risk and return (profit and loss on assets and Debts) were obtained using particle swarm algorithms, genetics and gray wolves and we compared their results using SPSS 16 software. After that, first the descriptive statistics were analyzed and then inferential statistics were performed. after reviewing the results of comparing the evaluation indicators of algorithms, it was determined that the gray wolf algorithm is efficient. Provides better goal function optimization. Also, by examining the research hypotheses, it was found that particle swarm algorithms and genetics have the same efficiency for assessing the integrated risk of the banking system. Provides better problem solving.

Pair trading strategy is one of the oldest and most common statistical arbitrage strategies. Pair formation is an important step in pair trading that examined manually and this method fails in the multivariate mode and does not consider conflicting goals in the problem structure. The main problem in this study is to present a method that creates multivariate pair combinations with multiple contradictory goals and focusing on the integration approach. Therefore, a combination of stocks optimized for two opposite objectives risk (mean-reversion) and return (spread variance) to form a set of profitable multivariate pair trading opportunities. The statistical population is companies listed on the Tehran Stock Exchange. The statistical sample limited by the need for high-frequency transactions from the top 50 companies. The problem developed in the form of a mixed integer-programming model (MIP), and due to non-convex constraints and exponential space, a multi-objective genetic algorithm used to obtain pair combinations. To achieve multiple goals, an advanced type of genetic algorithm; The Chaotic Non-dominated Sorting Genetic Algorithm (CNSGA-II) was used. The Chaos theory used to create the initial population of the genetic algorithm in order to obtain appropriate and high-precision solutions. Research has shown that the use of chaos theory can increase the degree of convergence in evolutionary algorithms. The results of the experiments of this study show that multi-objective pair trading strategies focusing on the integration approach have a significant advantage over the traditional single-objective model.

Determining the appropriate strategy for buying or selling in the foreign exchange market is very important for companies to cover exchange rate fluctuations against the national currency. This study proposes a new approach based on genetic algorithms and support vector machines for trading in the foreign exchange market.In this research, a new algorithm with the ability to generate technical rules for investment based on forecast certainty is presented. For prediction, a combination of the Combined Support Vector Machine (HSVM) algorithm for classifying the market into three different classes (uptrend, downtrend, sideway) and a dynamic genetic algorithm for optimizing trading rules based on several technical indicators Different has been used. Rials-dollar pair data is used as training and test data for the period between 1392 and 1398. The proposed architecture for machine learning, as well as the implementation and study of the proposed trading system are fully described. The research shows promising results during the test period in which the return on investment was 129%.

Choosing the optimal stock portfolio is an important issue in the economy. There are several techniques and tools to solve problem the optimal portfolio. In this research, using data of 15 stocks which randomly selected from the Tehran Stock Exchange including; PKOD, ZMYD, BPAS, FOLD, MKBT, GOLG, MSMI, PTAP, SSEP, AZAB, FKAS, NBEH, PFAN, GMRO and GSBE, the first return of these stocks are calculated daily in the period of 31/3/1394 -31/3/1399 for 5 years for 1183 days. Then and their portfolio risk is calculated using the models of mean–semi variance risk and conditional value at risk, and these two criteria are compared by the classical solution method. The portfolio optimization output with each of these risks represents a different weight per share. In the following, the mean–semi variance risk model and conditional value at risk model of metaheuristic method using MATLAB (R2019) software are compared. The results showed that the metaheuristic method of genetic algorithm compared to the classical method of mean–semi variance and conditional risk value of deviation in solving the problem of portfolio optimization had higher returns and lower risk under the risk criterion of mean–semi variance and therefore a better method for solve optimization problems.

Liquidity of stocks is an important challenge in the capital market. Identifying the factors affecting liquidity helps to predict the stock liquidity situation and thus stock risk management. The purpose of this study is to find the factors affecting the liquidity of stocks. For this purpose, in the first stage, using the research literature and experts, the influencing factors are identified and using the methods of minimum redundancy and maximum correlation (MRMR) and genetic algorithm, the effective variables are selected. In this research, using Excel software and existing raw data, the required data was created and then using support software and neural network toolbox and support vector machine was created. . Finally, the extracted variables using MRMR include stock market value, intensity of product market competition, GDP growth, equity returns, stock returns, inflation rate and family ownership, and using the financial model of financial leverage, government ownership, Equity returns, GDP growth, share buoyancy percentage, market type and board (on the stock exchange and OTC), the intensity of competition in the product market were selected.

Active Banks in the recent economic environment are obliged to encounter in a massive gamut of risks which are closely following them. If cash is regarded as cash at hand, then Liquidity risk is a kind of loss which arises of lack of fund or more specifically endured loss originating from inability of funding required capital in a reasonable way or selling off assets or being forced to have them pledged in order to cover solicited or unsolicited commitments. Hence Liquidity risk is comprised of economic loss incurred due to of providing cash and is deemed vital for operational activities of enterprises. Liquidity Mismatch in banks or maturity mismatch of sensitive assets to cash or debt may culminate in divergent of cash inflow or outflow during elapse of time which is actually stressed as Liquidity risk. Quarterly performance of 23 quoted banks in either Tehran Stock Exchange or Iran Farabourse are executed to model forecasted Banks’ Liquidity risk by means of implementing Artificial Neural Network algorithms. Applying genetic algorithm and Levenberg algorithm helped utilizing the best Training method and subsequently by facilitating Principal Component Analysis (PCA) method, we managed to optimize independent variables. Finally having hidden layers been determined and exercising calculations by Bayesian network model, the Artificial neural network is modeled and tested. All the mentioned process is performed by MATLAB software. Eventually fulfilling the asserted stages, a robust model for anticipating listed banks’ Liquidity risk is developed and findings of models for forcasted data is elaborated.

Portfolio optimization is a process in which the investor seeks to maximize return on investment or minimize risk. One of the main issues is to determine the optimization method, which is to form an optimal investment portfolio, which means minimizing investment risk and maximizing investment profit. The aim of this study was to investigate the capability of adaptive fuzzy neural inference system (ANFIS) and grid matrix (GA) strategies in selecting and optimizing the investment portfolio from among selected Tehran Stock Exchange and OTC companies. The grouping of stocks by the network matrix and the classification of companies based on their market value and the use of the law of quarters and finally their weighting is considered in proportion to the forecast return for the next month of that share. Also, a stock portfolio optimization model has been designed and presented using an adaptive fuzzy neural inference system and its combination with a genetic algorithm in which three different categories of time, technical and fundamental series variables are used as model inputs. It becomes. Research outputs show that these systems have the ability to optimize the stock portfolio.

The volume and speed of transactions in financial markets has increased significantly and has undergone extensive changes nowadays. Facing with increasing, decreasing or fluctuating trends in the stock market, determining the right trading strategy is very important. Therefore, complex meta-heuristic models are used for choosing a suitable strategy. In this research, an attempt is made to develop a new method of selecting and optimizing the stock portfolio based on the ensemble learning algorithm and genetics in order to select the best trading strategy to achieve greater returns and less risk. A combination of a six-class support vector machine (SVM) algorithm is used to predict returns and receive a buying signal; besides, a dynamic genetic algorithm is used to optimize trading rules. In this study, collective learning methods including Bagging, one of the algorithms based on Ensemble Learning, have been used to improve the accuracy of classification of returns. Data related to each share and fundamental variables in a daily time interval between years 1390 to 1399 is used as training and test data. The obtained results, comparing to traditional methods, are promising.

Predicting the financial position of companies based on financial indicators is one of the most important issues of interest to investors, creditors and other stakeholders of the company such as suppliers or retailers. Because, evaluating a company's financial position before making any investment or lending decisions seems necessary to prevent losses. The purpose of this study was to predict the financial indices of companies using the conditional average estimator method (CAE) and genetic algorithm (GA). The research method was DM-CRISP and the financial data of 130 stock companies over 10 years from 2009 to 2018 were analyzed. The results showed that the conditional average estimator method has high accuracy and ability in modeling. Also, the use of genetic algorithm in combination increases the accuracy of prediction. Capital Market Operators Can Use Research Results to Better Predict Corporate Financial and Performance Indicators.

Fluctuations in the financial markets are accompanied by signals and noise. In this paper, in addition to Singular Spectrum Analysis (SSA), a Genetic Algorithm (GA) is used to find the optimal window length and cut-off point, the objective of which is to find the minimum value for the correlation function between signal and noise components. Therefore, first, ten-year data of the overall index of Tehran Stock Exchange during 2009 to 2018 were implemented in three using the SSA method. Then it was solved in the form of an optimization problem by a genetic algorithm. The results of the first hypothesis showed that signal and noise resolution is possible in the SSA method. Also, according to the results of the research, Singular spectrum analysis based on genetic algorithm with an absolute value of less than the average value showed an improvement in prediction accuracy. Finally, considering the lowest weight correlation between time series components for signal and noise separation (finding the cut-off point) and then obtaining the optimal window length in the SSA based on GA, indicates the fact that the amount of parameters can be changed. Improve the performance of the SSA method to be useful.

Devaluation of assets in the future is one of the most important investment concerns that has led investors to choose the set of assets that have the lowest risk and highest return. The present study deals with the problem of stock portfolio optimization according to the Conditional Value at Risk based on the new and intelligent fireworks algorithm and compares it with genetic algorithm with the historical simulation method using MATLAB software. The parameters of meta-heuristic algorithms were adjusted by Taguchi method using MINITAB software. Not suspended, used. For reliability of the study, generalized Dickey-Fuller test and Phillips-Prone test were used. To evaluate the accuracy of the Conditional Value at Risk model, the kupiec proportion of failure test, Christoffersen independence test and Conditional coverage test are used.  A comparison was also made between the models by Lopez test. Findings showed that at %95 and %99  confidence levels, the conditional risk value model using the fireworks algorithm has a suitable and reliable validity for measuring market risk and optimizing the stock portfolio.