طراحی مدلی جهت پیش بینی بازده شاخص کل بورس اوراق بهادار(با تاکید بر مدلهای ترکیبی شبکه یادگیری عمیق و مدلهای خانواده GARCH)
محورهای موضوعی : مهندسی مالیمهدی ذوالفقاری 1 , بهرام سحابی 2 , محمد جواد بختیاران 3
1 - گروه علوم اقتصادی، دانشکده مدیریت و اقتصاد، دانشگاه تربیت مدرس،تهران، ایران
2 - گروه علوم اقتصادی، دانشکده مدیریت و اقتصاد، دانشگاه تربیت مدرس، تهران، ایران
3 - گروه علوم اقتصادی، دانشکده مدیریت و اقتصاد، دانشگاه تربیت مدرس، تهران، ایران
کلید واژه: پیش‌بینی, خانواده GARCH, شاخص بورس اوراق بهادار, شبکه یادگیری عمیق,
چکیده مقاله :
در سال‌های اخیر، توسعه‌ی پردازنده‌های کامپیوتری موجب معرفی الگوریتم‌های جدیدی برای پیش‌بینی دادههای مالی شده است که یکی از این الگوریتم‌ها، یادگیری ماشین (Machine Learning) است. از اینرو در پژوهش حاضر به معرفی یک مدل ترکیبی از شبکه یادگیری عمیق (Deep Learning) و مدلهای منتخب خانواده GARCH جهت پیش‌بینی کوتاه‌مدت بازدهی روزانه شاخص کل بورس اوراق بهادار تهران پرداخته می‌شود. مهمترین ویژگی شبکه یادگیری عمیق در این است که بدون محدود بودن به مدلهای معین، می‌تواند خود را با نوسانات متغیرهای بازار هماهنگ و تعدیل نماید. در این پژوهش از میان مدلهای شبکه یادگیری عمیق، شبکه عصبی بازگشتی مبتنی بر حافظه کوتاه‌مدت و بلندمدت (RNN-LSTM) انتخاب و از مدلهای دارای حافظه‌کوتاه مدت GARCH و EGARCH در ساختار آن استفاده می‌شود. همچنین دو متغیر مستقل قیمت نفت و نرخ دلار در ساختار مدل ترکیبی، کمک فراوانی به آن در پیش‌بینی دقیقتر دادههای مالی میکند. نتایج تحقیق نشان می‌دهد که مدلهای ترکیبی دقت پیش‌بینی بالاتری نسبت به مدلهای تکی دارند. همچنین براساس معیارهای ارزیابی خطای پیش‌بینی RMSE و MAPE، مدل RNN-LSTM-EGARCH برپایه توزیع GED دارای خطای پیش‌بینی کمتری نسبت به 23 مدل دیگر دارد. در این راستا، معیار بررسی صحت پیش‌بینی دیبولد-ماریانو (DM) نیز یافته‌های فوق را تایید میکند.
Given the development of machine learning models in predicting financial data in recent years, this study introduces a combination of Deep Learning Network and selected GARCH family models to predict short-term daily returns of the Tehran Stock Exchange Index. The most important feature of the deep learning network is that it can adapt and adjust itself to the volatility of market variables without being limited to specific models. In this study, short-term and long-term memory based neural network (RNN-LSTM) models are used for deep learning network models and GARCH and EGARCH models are used in its structure. Also, the two independent variables of oil price and dollar rate in the structure of the hybrid model help to predict the financial data more accurately. Comparison of the results of hybrid model prediction error with individual models shows that the RNN-LSTM-EGARCH hybrid model has higher prediction accuracy than competing models. competing models.
بناکار، احمد، (1392)کتاب شبکه‌های عصبی موجک و کاربرد آنها در سامانه‌های فازی-عصبی، انتشارات دانشگاه تربیت مدرس.
جمشیدی ویسمه، مهسا.(1396)، "معاملات الگوریتمی و پربسامد"، مدیریت تحقیق و توسعه، بورس اوراق بهادار تهران.
حافظی، رضا. شهرابی، جمال.هداوندی، اسماعیل، (1392)، "توسعه مدلی ترکیبی هوشمند برای پیش‌بینی بازار سهام تهران"، مجله تحقیق در عملیات و کاربردهای آن،سال دهم،شماره 2، دوره 10، صص49-35.
حنفی‌زاده، پیام. جعفری، ابوالفضل(1389).مدل ترکیبی شبکه‌های عصبی مصنوعی پیش‌خور و خودسازمانده کوهونن برای پیش‌بینی قیمت سهام. مطالعات مدیریت صنعتی، شماره19، دوره8، صص187-165.
راعی، رضا. محمدی، شاپور. فندرسکی، حنظله(1394)، "پیش‌بینی شاخص قیمت بورس سهام با استفاده از شبکه عصبی و تبدیل موجک"، فصلنامه علمی پژوهشی مدیریت دارایی و تامین مالی، شماره اول، دوره 3، صص74-55.
رفیعی امام، علینقی،(1398) کتاب انتخاب سهام به روش تحلیل بنیادی، انتشارات نص .
سلامی بیدگلی، غلامرضا. راعی، رضا. کمال‌زاده، سحر(1392)، محاسبه ارزش در معرض خطر قیمت سبد نفتی اوپک با استفاده از مدلهای حافظه بلندمدت گارچ، فصلنامه مطالعات اقتصاد انرژی، شماره 39، دوره دهم، صص 19-1.
سید حسینی، میرمیثم. احمدی، زانیار. (1393)، "مفاهیم معاملات الگوریتمی"، مدیریت پژوهش، توسعه و مطالعات اسلامی، گزارش شماره2.
صادقی، حسین. ذوالفقاری، مهدی، (1390) کتاب مبانی مدلهای پیش‌بینی در علوم اقتصادی، انتشارات نور علم.
فلاح شمس، میرفیض. دلنواز اصغری، بیتا(1388)پیش‌بینی شاخص بورس اوراق بهادار تهران با استفاده از شبکه‌های عصبی، شماره 9، دوره 3، صص212-191.
کریمی، محمد شریف. امام وردی، قدرت اله. دباغی، نیشتمان(1392).ارزیابی و شناسایی مناسب‌ترین گزینه سرمایه‌گذاری دارایی و مالی در ایران در بازه زمانی 1389- 1380. اقتصاد مالی، شماره 25، دوره7،صص 207-177.
نریمانی،حکیمی پور.اله رضایی،اسعد(1392)کاربرد روش شبکه عصبی مصنوعی و مدلهای واریانس ناهمسانی شرطی در محاسبه ارزش در معرض خطر. اقتصاد مالی, شماره24، دوره 7، صص 137-101.
Abramovich, Yuri l(1981).Controlled method for adaptive optimization of filters using the criterion of maximum SNR. Radio Engineering and Electronic Physics, 26(3):87–95.
Akansu, Ali N., Kulkarni, Sanjeev R., and Malioutov Dmitry M., editors(2016). Financial Signal Processing and Machine Learning. Wiley-IEEE Press.
Amano Akihiro (1987). A small forecasting model of the world oil market.615-35.
Alex Graves (2013). Generating Sequences With Recurrent Neural Networks.
Almgren, Robert. and Chriss, Neil(2001). Optimal execution of portfolio transactions. Journal of Risk, 3:5–40
Anwar, S., & Mikami, Y. (2011). Comparing accuracy performance of ANN, MLR, and GARCH model in predicting time deposit return of Islamic bank. International Journal of Trade, Economics and Finance, 2(1), 44
Appel, Gerald., (2005). Technical analysis: power tools for active investors. FT Press.
Babaei, Sadra, Sepehri, Mohammad Mehdi and Babaei, Edris., (2015). Multi-objective portfolio optimization considering the dependence structure of asset returns. European Journal of Operational Research, 244(2), pp.525-539.
Bildirici, M., & Ersin, Ö. Ö. (2009). Improving forecasts of GARCH family models with the artificial neural networks: An application to the daily returns in Istanbul Stock Exchange. Expert Systems with Applications, 36(4), 7355-7362.
Bollinger, John, (2001). Bollinger on Bollinger bands. McGraw Hill Professional.
Briza, A.C. and Naval Jr, P.C., (2011). Stock trading system based on the multi-objective particle swarm optimization of technical indicators on end-of-day market data. Applied Soft Computing, 11(1), pp.1191-1201.
Cvitanic, Jaksa, Kirilenko, Andrei(2010).High frequency traders and asset prices ,Available atSSRN1569075, 2010.
Davallou, Maryam.Safari, Ali.(2018) Oil price forecasting using a hybrid model
Fan, s.,Hyndman, Rob J.(2012).Short-term load forecasting based on a semi-parametric additive model.IEEE Transaction on power systems,134-141.
Fatima, Samreen, Mudassir, Uddin.(2017) "Comparison of asymmetric garvh models with artificial neural network for stock markets prediction, a case study”,vol.36.
Güreşen, E., & Kayakutlu, G. (2008, October). Forecasting stock exchange movements using artificial neural network models and hybrid models. In International Conference on Intelligent Information Processing (pp. 129-137). Springer, Boston, MA.
Goodfellow Ian(2016). Deep Learning.
Hastie, Trevor, Tibshirani, Robert. and Friedman, Jerome. (2008). The elements of statistical learning: data mining, inference, and prediction, 2nd ed. New York: Springer.
Holthausen, Robert W., Leftwich, Richard W, Mayers, David (1987) .The effect of large block transactions on security prices: across-sectional analysis, J.Financ.Econ.19(2) pp237-267.
Hong, T.Liu, B.Wang, P(2016). Electric load forecasting with recency effect:A big data approach 585-597.
Kristjanpoller, Werner, Hernández, Esteban,(2017), Volatility of main metals forecasted by a hybrid ANN-GARCH model with regressors, Expert Systems with Applicationsl, vol 11, pp290-300.
Kristjanpoller, Werner, & Minutolo, Maracel C. (2016). Forecasting volatility of oil price using an artificial neural network-GARCH model. Expert Systems with Applications, 65, pp233-241.
Kristjanpoller, Werner, Minutolo, A, Marcel C, 2018. hybrid volatility forecasting framework integrating GARCH, Artificial Neural network, Technical Analysis and Principal Components Analysis, Expert Systems with Applications.
Lahmiri, S. (2017). Modeling and predicting historical volatility in exchange rate markets. Physica A: Statistical Mechanics and its Applications, 471, 387-395.
Lahmiri, S., & Boukadoum, M. An Ensemble System Based on Hybrid EGARCH-ANN with Different Distributional Assumptions to Predict S&P 500 Intraday Volatility. Fluctuation and Noise Letters, 2015, 14(01), 1550001.
Lu, X., Que, D., & Cao, G. Volatility forecast based on the hybrid artificial neural network and GARCH-type models. Procedia Computer Science, 2016, 91, 1044-1049.
Markowitz, Harry,(1959). Portfolio selection: efficient diversification of investments, Wiley, New York, NY ,x,344pp.
Moshiri ,Saeed.Foroutan ,Faezeh(2006). Forecasting nonlinear crude oil prices.pp81-95.
Monfared, S. A., & Enke, D. (2014). Volatility forecasting using a hybrid GJR-GARCH neural network model. Procedia Computer Science, 36, 246-253.
Rockafellar, Tyrrell, Uryasev, Stan and Zabarankin, Michael (2006). Generalized deviations in risk analysis. Finance and Stochastics, 10, pp51–74.
Rousseeuw,Petter J. and Driessen, Katrien(2006).Computing LTS regression for large datasets. Data Mining and Knowledge Discovery, 12, pp29–45.
Wang, Y. H. (2009). Nonlinear neural network forecasting model for stock index option price: Hybrid GJR–GARCH approach. Expert Systems with Applications, 36(1), 564-570.
Zolfaghari, M., & Sahabi, B. (2017). Impact of foreign exchange rate on oil companies risk in stock market: A Markov-switching approach. Journal of Computational and Applied Mathematics, 317, 274-289.
Zolfaghari, M., & Sahabi, B. (2019). A hybrid approach to model and forecast the electricity consumption by NeuroWavelet and ARIMAX-GARCH models. Energy Efficiency, 1-24.
_||_
Abramovich, Yuri l(1981).Controlled method for adaptive optimization of filters using the criterion of maximum SNR. Radio Engineering and Electronic Physics, 26(3):87–95.
Akansu, Ali N., Kulkarni, Sanjeev R., and Malioutov Dmitry M., editors(2016). Financial Signal Processing and Machine Learning. Wiley-IEEE Press.
Alex Graves (2013). Generating Sequences With Recurrent Neural Networks.
Almgren, Robert. and Chriss, Neil(2001). Optimal execution of portfolio transactions. Journal of Risk, 3:5–40
Amano Akihiro (1987). A small forecasting model of the world oil market.615-35.
Anwar, S., & Mikami, Y. (2011). Comparing accuracy performance of ANN, MLR, and GARCH model in predicting time deposit return of Islamic bank. International Journal of Trade, Economics and Finance, 2(1), 44
Appel, Gerald., (2005). Technical analysis: power tools for active investors. FT Press.
Babaei, Sadra, Sepehri, Mohammad Mehdi and Babaei, Edris., (2015). Multi-objective portfolio optimization considering the dependence structure of asset returns. European Journal of Operational Research, 244(2), pp.525-539.
Banakar, Ahmed, (2012) The book of wavelet neural networks and their application in fuzzy-neural systems, Tarbiat Modares University Publications.
Bildirici, M., & Ersin, Ö. Ö. (2009). Improving forecasts of GARCH family models with the artificial neural networks: An application to the daily returns in Istanbul Stock Exchange. Expert Systems with Applications, 36(4), 7355-7362.
Bollinger, John, (2001). Bollinger on Bollinger bands. McGraw Hill Professional.
Briza, A.C. and Naval Jr, P.C., (2011). Stock trading system based on the multi-objective particle swarm optimization of technical indicators on end-of-day market data. Applied Soft Computing, 11(1), pp.1191-1201.
Cvitanic, Jaksa, Kirilenko, Andrei(2010).High frequency traders and asset prices ,Available atSSRN1569075, 2010.
Davallou, Maryam.Safari, Ali.(2018) Oil price forecasting using a hybrid model
Fallah Shams, Mirfaiz. Delnawaz Asghari, Bita (2018) Prediction of Tehran Stock Exchange index using neural networks, number 9, volume 3, pp. 191-212.
Fan, s.,Hyndman, Rob J.(2012).Short-term load forecasting based on a semi-parametric additive model.IEEE Transaction on power systems,134-141.
Fatima, Samreen, Mudassir, Uddin.(2017) "Comparison of asymmetric garvh models with artificial neural network for stock markets prediction, a case study”,vol.36.
Goodfellow Ian(2016). Deep Learning.
Güreşen, E., & Kayakutlu, G. (2008, October). Forecasting stock exchange movements using artificial neural network models and hybrid models. In International Conference on Intelligent Information Processing (pp. 129-137). Springer, Boston, MA.
Hafizi, Reza. Shahrabi, Jamal. Hadavandi, Esmail, (2012), "Development of an intelligent hybrid model for predicting the Tehran stock market", Journal of Research in Operations and Its Applications, Year 10, Number 2, Volume 10, pp. 35-49.
Hanafizadeh, Payam. Jafari, Abolfazl (2009). The combined model of Kohonen feed forward and self-organizing artificial neural networks for stock price prediction. Industrial Management Studies, No. 19, Volume 8, pp. 165-187.
Hastie, Trevor, Tibshirani, Robert. and Friedman, Jerome. (2008). The elements of statistical learning: data mining, inference, and prediction, 2nd ed. New York: Springer.
Holthausen, Robert W., Leftwich, Richard W, Mayers, David (1987) .The effect of large block transactions on security prices: across-sectional analysis, J.Financ.Econ.19(2) pp237-267.
Hong, T.Liu, B.Wang, P(2016). Electric load forecasting with recency effect:A big data approach 585-597.
Jamshidi Vismeh, Mehsa. (2016), "Algorithmic and high-frequency transactions", research and development management, Tehran Stock Exchange.
Karimi, Mohammad Sharif. Imam Vardi, the power of God. Dabaghi, Nishtman (2012). Evaluation and identification of the most appropriate property and financial investment option in Iran in the period of 2010-2018. Financial Economy, No. 25, Volume 7, pp. 177-207.
Kristjanpoller, Werner, & Minutolo, Maracel C. (2016). Forecasting volatility of oil price using an artificial neural network-GARCH model. Expert Systems with Applications, 65, pp233-241.
Kristjanpoller, Werner, Hernández, Esteban,(2017), Volatility of main metals forecasted by a hybrid ANN-GARCH model with regressors, Expert Systems with Applicationsl, vol 11, pp290-300.
Kristjanpoller, Werner, Minutolo, A, Marcel C, 2018. hybrid volatility forecasting framework integrating GARCH, Artificial Neural network, Technical Analysis and Principal Components Analysis, Expert Systems with Applications.
Lahmiri, S. (2017). Modeling and predicting historical volatility in exchange rate markets. Physica A: Statistical Mechanics and its Applications, 471, 387-395.
Lahmiri, S., & Boukadoum, M. An Ensemble System Based on Hybrid EGARCH-ANN with Different Distributional Assumptions to Predict S&P 500 Intraday Volatility. Fluctuation and Noise Letters, 2015, 14(01), 1550001.
Lu, X., Que, D., & Cao, G. Volatility forecast based on the hybrid artificial neural network and GARCH-type models. Procedia Computer Science, 2016, 91, 1044-1049.
Markowitz, Harry,(1959). Portfolio selection: efficient diversification of investments, Wiley, New York, NY ,x,344pp.
Monfared, S. A., & Enke, D. (2014). Volatility forecasting using a hybrid GJR-GARCH neural network model. Procedia Computer Science, 36, 246-253.
Moshiri ,Saeed.Foroutan ,Faezeh(2006). Forecasting nonlinear crude oil prices.pp81-95.
Narimani, Hakimipour. Ele Rezaei, Asad (2012) Application of artificial neural network method and conditional heterogeneity variance models in calculating value at risk. Financial Economy, No. 24, Volume 7, pp. 101-137.
Rafiei Imam, Alineqi, (2018) Stock selection by fundamental analysis method, Nas publications.
Rai, Reza. Mohammadi, Shapour. Fendersky, Hanzaleh (2014), "Prediction of stock market price index using neural network and wavelet transformation", scientific research quarterly of asset management and financing, number 1, volume 3, pp. 55-74.
Rockafellar, Tyrrell, Uryasev, Stan and Zabarankin, Michael (2006). Generalized deviations in risk analysis. Finance and Stochastics, 10, pp51–74.
Rousseeuw,Petter J. and Driessen, Katrien(2006).Computing LTS regression for large datasets. Data Mining and Knowledge Discovery, 12, pp29–45.
Sadeghi, Hossein. Zulfiqari, Mehdi, (2018) Basics of Forecasting Models in Economic Sciences, Noor Alam Publications.
Salami Bidgoli, Gholamreza. Rai, Reza. Kamalzadeh, Sahar (2012), calculating the value at risk of the OPEC oil basket price using GARCH long-term memory models, Energy Economics Quarterly, No. 39, Volume 10, pp. 1-19.
Seyed Hosseini, Mirmitham. Ahmadi, Zaniar. (2013), "Concepts of Algorithmic Transactions", Research Management, Development and Islamic Studies, Report No. 2.
Wang, Y. H. (2009). Nonlinear neural network forecasting model for stock index option price: Hybrid GJR–GARCH approach. Expert Systems with Applications, 36(1), 564-570.
Zolfaghari, M., & Sahabi, B. (2017). Impact of foreign exchange rate on oil companies risk in stock market: A Markov-switching approach. Journal of Computational and Applied Mathematics, 317, 274-289.
Zolfaghari, M., & Sahabi, B. (2019). A hybrid approach to model and forecast the electricity consumption by NeuroWavelet and ARIMAX-GARCH models. Energy Efficiency, 1-24.