مدل‌سازی انواع شاخص بورس اوراق بهادار ایران با الگوریتم تقریب تابع ژنتیک

نوع مقاله: مقاله پژوهشی

نویسندگان

استادیار دانشگاه سیستان و بلوچستان

چکیده

از مهم­ترین مباحث جدی در بازار بورس، شاخص‌های بورس می‌باشد. هدف اصلی این پژوهش مدل‌سازی عوامل موثر بر شاخص قیمت سهام، شاخص قیمت و بازده نقدی بورس، شاخص مالی و شاخص صنعت در بورس اوراق بهادار ایران است. بدین منظور، از اطلاعات 112 متغیر کلان اقتصادی و بورس طی سالهای 1376 تا 1393 استفاده شده است که مدل‌سازی با روش الگوریتم تقریب تابع ژنتیک صورت گرفته است. با استفاده از نرم افزار MSmodeling مدل‌سازی برای عوامل موثر بر شاخص قیمت سهام، شاخص قیمت و بازده نقدی بورس، شاخص مالی، شاخص صنعت صورت گرفت تا مشخص گردد که از 108 متغیرمستقل چه متغیرهایی برانواع شاخص بورس موثر می‌باشد. نتایج پژوهش نشان می‌دهد تسهیلات اعطایی بانکها منجر به افزایش شاخص صنعت در بازار بورس می‌شود. پایه پولی و تسهیلات اعطایی بانکها و سپرده سرمایه گذاری کوتاه مدت نیز بر شاخص قیمت سهام موثر می‌باشد. همچنین متغیرهای تعداد سهام معامله شده، ارزش معاملات و تعداد خریداران باعث افزایش شاخص صنعت و شاخص قیمت سهام شده است. با توجه به یافته‌ها پیشنهاد می‌شود که تسهیلات اعطایی بانکها به بخشهای دولتی و غیر دولتی و موسسات اعتباری غیر بانکی باعث افزایش شاخص صنعت و شاخص قیمت سهام می‌شود که دولت، در مورد برخی از شرکتهای بورس وارد عمل شود و بنگاه های ورشکسته را با اعطای تسهیلات مدیریت کند و وضعیت آنها را در بازار بورس بهبود بخشد. همچنین با توجه به یافته‌های پژوهش رشد صنعت خودرو منجر به رشد شاخص مالی می‌شود و بدین منظور سیاست­گذاران بایستی به صنعت خودرو توجه ویژه‌ای نمایند.

کلیدواژه‌ها


عنوان مقاله [English]

Modeling a Variety of Indices of Iranian Stock Exchange Using Genetic Function Approximation Algorithm

نویسندگان [English]

  • Mahmoud Hashemi Tabar
  • Amir Dadras Moghadam
  • Seyed Mehdi Hosseini
  • Ebrahim Moradi
University of Sistan and Baluchestan, Zahedan, Iran.
چکیده [English]

Abstract
The most important issues in the stock market are stock indices. The main topic of this research is modeling the factors affecting stock price index, stock price and stock return index, financial index, and industry index in Iran's Stock Exchange. For this purpose, data of 112 macroeconomic and stock variables from 1997 to 2014 were used. Modeling has been done using the genetic function approximation algorithm. Using MSmodeling software has been modeled for the factors affecting the stock price index, the price index and cash returns of the stock exchange, the financial index, the industry index were used to determining 108 independent variables are effective on the types of stock indexes. The results indicate that the granting facilities of banks lead to an increase in the industry index in the stock market. The monetary base and the bank's facilities and short-term investment deposit are also effective on the stock price index. Moreover, the variables of the number of shares traded, the value of transactions, and the number of buyers increases the industry index and the stock price index. According to the findings, it is concluded that banks' grant facilities to government and non-government sectors and non-bank credit institutions will increase the industry index and stock price index. In the case of some stock companies, the government even needs to take some actions itself and manage the bankrupt enterprises by granting facilities and improve their status on the stock market. Also, according to the research findings, the growth of automotive industry results in the growth of the financial index, and for this purpose, policymakers should pay particular attention to the automotive industry. Finally, given the results of this study, longer periods of time for future research and the use of other predictive methods as well as artificial intelligence are emphasized.
Introduction
Investors and managers of the stock market make use of stock indices in order to achieve a good picture of the process of this market and the ability to evaluate past and, in some cases, to predict the future. A more detailed analysis of the price trend in stock markets requires indices with a variety of functions. As a result, today a wide variety of indices are calculated and published in the Iranian Stock Exchange. The methods of calculating the indices have undergone several changes in the direction of more efficiency and providing a more precise representation of the stock trading process. Naturally, there are a bulk of factors involved in shaping the information and views of the parties to the market and, ultimately, the stock prices of the companies. Some part of these factors is indigenous and some other is due to the status of variables outside the scope of the domestic economy of the company. Accordingly, the factors affecting stock prices are wide. On the other hand, each country's economic development depends on the money and capital markets in each country's economy. Given the importance of capital market in equipping community savings towards economic activities, identifying variables that affect the stock price index is quintessential. In this research, we have been trying to fill this gap in the financial literature of our country. Despite the fact that most previous research voluntarily selected a number of variables and examined their effects on the stock price index, in this research, optimal and effective variables in types of stock index is derived using the genetic function approximation. This research investigates the factors affecting price index, financial index, industry index, price index and cash returns, which is, in this regard, innovative compared to other studies.
Case study
The data of the statistical population was collected from the Central Bank website from 1997 to 2014. Since there are a lot of factors affecting Tehran Stock Exchange index. These factors include exports, current account balance, capital account balance, monetary and credit variables, payments and receipts of government and stock transactions, energy sector, manufacturing and mining sector, housing and construction sector, transportation, and agricultural sector. Modeling was performed for the factors influencing the stock price index, stock price and cash return index, financial index, and industry index to determine the variables effective on all types of stock market indices.
Materials and Methods
Using the genetic function approximation algorithm and running the MSmodeling software, modeling was performed for factors influencing the stock price index, stuck price index and cash return, financial index, industry index to determine which of the 108 independent variables are effective on the types of stock indices. An independent variable is added to the model and the optimal regression model is presented until no significant change, based on R2 or LOF criteria, is observed in the final model.
Discussion and Results
In a nutshell, it can be postulated that monetary and credit variables have been effective on stock price index and stock return, industry index, and stock price index, in which an increase in liquidity in the society leads to a decrease in the price index and stock return in the stock market. Banking grants to government and non-government sectors and non-bank credit institutions have boosted the industry's index in the stock market. The monetary base and the facilities granted by banks to government and non-government sectors as well as short-term investment deposits also affected the stock price index. Furthermore, the effective stock variables are effective on a variety of indices. This means that the increase in the number of buyers in the stock market has reduced the price index and stock returns. The increase in the number of shares traded and the buyer in the stock market increases the financial index. The variables of the number of traded shares, the value of transactions and the number of buyers increased the industry index and the stock price index.
Conclusion
According to the findings, it is concluded that banks' grant facilities to government and non-government sectors and non-bank credit institutions will increase the industry index and stock price index. In the case of some stock companies, the government even needs to take some actions itself and manage the bankrupt enterprises by granting facilities and improve their status on the stock market. Also, according to the research findings, the growth of automotive industry results in the growth of the financial index, and for this purpose, policymakers should pay particular attention to the automotive industry. Finally, given the results of this study, longer periods of time for future research and the use of other predictive methods as well as artificial intelligence are emphasized.

کلیدواژه‌ها [English]

  • stock price index
  • stock price index and cash return
  • financial index
  • industry index
  • Generic Function Algorithm (GFA)

1-Alimran, R., & Alimran, S. (2015). Stock Market effectiveness as a result of irregular growth in liquidity. The Stock Exchange Quarterly, 22:5-24. (In Persian)

2-Alimran, S. & Alimran, R. (2013). Studying the process of fluctuationsTehran stock exchange. Financial Knowledge of Securities Analysis, (14), 119-132. (In Persian)

3-Alrub, A. A., Tursoy, T., & Rjoub, H. (2016). Exploring the long-run and short-run Relationship between macroeconomic variables and stock prices during the restructuring period: does it matter in Turkish market?. Journal of Financial Studies & Research. g1-11.(In Persian)

4-Baharmoghadam, M., & Quavaraeie, T. (2012). Relationship between the days and months of the year, macroeconomic variables and stock returns in Tehran stock exchange. Journal of Accounting Progress of Shiraz University, 4 (2), 1-26.(In Persian)

5-Das, A. (2017). An Association of Macroeconomic Variables and Stock Index, India: an empirical evidence. MERC Global’s International Journal of Management, 5(1), 1-7.

6-Heidari, H. & Bashiri, S. (2012). Investigating the relationship between the uncertainty of the real exchange rate and the stock price index in Tehran Stock Exchange: Observations based on the VAR-GARCH model. Journal of Economic Modeling Research. 3 (9): 71-93. (In Persian)

7-Emenike Kalu, O., & Okwuchukwu, O. (2014). Stock market return volatility and macroeconomic variables in Nigeria. International journal of empirical finance, 2(2), 75-82.

8-Farazmand., S. Kordaich, A. & Moshbaki, A. (2013). Prediction of Tehran stock exchange index using ANFIS. Asset Management and Financing, 1(1),27-44.(In Persian)

9-Fu, G., luo, Ch., & Wang, J. (2013), Accounting Information and Stock Price Reaction on Listed Companies- Empirical Evidence from 60 Listed Companies in Shanghai Stock Exchange, Journal of Business of Management, 2(2), pp. 11-21.

10-Golestani., S. Daldar, M. Seyyed, S. & Jafari, S. H. (2014). The relation between effective tax rate and dividend profit and stock returns in companies listed in Tehran stock exchange. Economic Research and Policy,22 (70), 181 – 204. (In Persian)

11-Kamroupher, M. & Hashemi, S. Z. (2012). Investigating and recognizing the main variables affecting Tehran stock exchange index and its modeling using artificial neural networks and comparing the results with technical analysis and Elliott waves. Financial Engineering and Portfolio Management (Portfolio Management). 8(30):169-184 .(In Persian)

12-Kayyei, R. (2011). An analytical view at Stock Indicators, Stock Exchange and Securities Brokers. (In Persian)

13-Khajeh, A., & Modarress, H. (2010). QSPR prediction of flash point of esters by means of GFA and ANFIS. Journal of hazardous materials, 179(1), 715-720.

14-Khaled, K. F. (2011). Modeling corrosion inhibition of iron in acid medium by genetic function approximation method: A QSAR model. Corrosion Science, 53(11), 3457-3465.

15-Li, L., Narayan, P. K., & Zheng, X. (2010). An analysis of inflation and stock returns for the UK. Journal of international financial markets, institutions and money, 20(5), 519-532.

16-Moradi, A. (2006). Relationship between Financial Ratios and Stock Returns in Tehran Stock Exchange. Master's Degree in Accounting, Tarbiat Modarres University. (In Persian)

17-Niece, A., & Paymani, M. (2014). Modeling of Tehran Stock Exchange index using randomized differential equation. Economic Research, 14 (53), 143-166.(In Persian)..

18-Prosenjit B, J. Thomas, L. & Kunal, R. (2005). Exploring QSAR of thiazole and thiadiazole derivatives as potent and selective human adenosine A3 receptor antagonists using FA and GFA techniques Chemistry, Journal of molecular modeling, 11 (6), 516-24.

19-Rogers, D., & Hopfinger, A. J. (1994). Application of genetic function approximation to quantitative structure-activity relationships and quantitative structure-property relationships. Journal of Chemical Information and Computer Sciences, 34(4), 854-866.

20-Samuel, H. Uzairu, A. Mamza1, P. & Oluwole Joshua, O. (2015). Quantitative structure-toxicity relationship study of some polychlorinated aromatic compounds using molecular descriptors. Journal of Computational Methods in Molecular Design, 5 (3):106-119.

21-Sandvik, A. A. & Følgesvold, L. R. (2016). Causal relations between stock market returns and macroeconomic variables: cointegration evidence from the Norwegian stock market (Master's thesis).

22-Zare, R., Azali, M., & Habibullah, M. S. (2013). Monetary policy and stock market volatility in the ASEAN5: Asymmetries over Bull and Bear markets. Procedia Economics and Finance, 7, 18-27.