ﺑﻬﯿﻨﻪ ﺳﺎزي اﻟﮕﻮي ﺳﺮﻣﺎﯾﻪ ﮔﺬاري در ﻧﺰول ﻫﺎي اﺳﺎﺳﯽ ﺑﻮرس اوراق ﺑﻬﺎدار ﺗﻬﺮان در ﭼﺎرﭼﻮب روﯾﮑﺮد ﻋﻮاﻣﻞ ﻧﺎﻫﻤﮕﻦ و ﻣﺪل ﺳﺎزي ﻋﺎﻣﻞ ﺑﻨﯿﺎن ﺑﺎ اﺳﺘﻔﺎده از اﻟﮕﻮرﯾﺘﻢ ژﻧﺘﯿﮏ

The optimization of Investment Beliefs in Tehran Stock Exchange Break Points Based on Heterogeneous Agent Models Framework and Agent Based Modelling with Genetic Algorithm

ﻫﺪف اﯾﻦ ﻣﻘﺎﻟﻪ ﺑﺮرﺳﯽ اﻟﮕﻮﻫﺎي رﻓﺘﺎري ﺳﺮﻣﺎﯾﻪ ﮔﺬاران در زﻣﺎن ﻧﺰولﻫﺎي اﺳﺎﺳﯽ ﺑﻮرس اوراق ﺑﻬﺎدار ﺗﻬﺮان در ﭼﺎرﭼﻮب ﻣﺪل ﻋﺎﻣﻞ ﻧﺎﻫﻤﮕﻦ و ﺑﻬﯿﻨﻪ ﺳﺎزي اﻟﮕﻮﻫﺎ ﺑﺎ اﺳﺘﻔﺎده از اﻟﮕﻮرﯾﺘﻢ ژﻧﺘﯿﮏ ﻣﯽﺑﺎﺷﺪ. در اﺑﺘﺪا ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ ﻣﺒﺎﻧﯽ ﻧﻈﺮي، ﻧﺰولﻫﺎي اﺳﺎﺳﯽ در ﺑﻮرس اوراق ﺑﻬﺎدار ﺗﻬﺮان ﺑﺮ اﺳﺎس ﺳﻪ ﻣﻌﯿﺎر روﻧﺪ ﺷﺎﺧﺺ، ارزش ﻣﻌﺎﻣﻼت روزاﻧﻪ و ارزش ﮐﻞ ﺑﺎزار ﻣﺸﺨﺺ ﮔﺮدﯾﺪ ﺑﺮ اﯾﻦ اﺳﺎس از ﺳﺎل 1380 ﺗﺎ ﺳﺎل 1394 ﺳﻪ ﻧﻘﻄﻪ ﺷﮑﺴﺖ و ﻧﺰول اﺳﺎﺳﯽ ﺗﻌﯿﯿﻦ ﺷﺪ: رﯾﺎﺳﺖ ﺟﻤﻬﻮري ﻣﺮﺑﻮط ﺑﻪ اﻧﺘﺨﺎب دوﻟﺖ ﻧﻬﻢ در ﺳﺎل 1384، ﺑﺤﺮان ﻣﺎﻟﯽ ﻏﺮب در ﺳﺎل 1387 ﺑﺎﻻﺧﺮه ﻧﺰول داﻣﻨﻪ دار ﺑﺎزار ﭘﺲ از رﯾﺎﺳﺖ ﺟﻤﻬﻮري ﻣﺮﺑﻮط ﺑﻪ دوﻟﺖ ﯾﺎزدﻫﻢ و ﺗﻮاﻓﻖ اوﻟﯿﻪ ﻫﺴﺘﻪاي. ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ ﻣﺪل ﻋﺎﻣﻞ ﻧﺎﻫﻤﮕﻦ ﺑﺮوك و ﻫﻮﻣﺰ و اﺻﻼﺣﺎت ﺑﻌﺪي آن ﻗﯿﻤﺖﻫﺎي ﺑﺎزار و ﺑﻨﯿﺎدي 40 روز ﻗﺒﻞ و ﺑﻌﺪ از ﻧﺰولﻫﺎي اﺳﺎﺳﯽ ﺗﻌﯿﯿﻦ و ﮐﺪ ﻧﻮﯾﺴﯽ و ﺷﺒﯿﻪﺳﺎزي ﺑﺎ اﺳﺘﻔﺎده از ﻧﺮم اﻓﺰار ﻣﺘﻠﺐ اﻧﺠﺎم ﮔﺮدﯾﺪ و ﭘﺲ از آن ﺑﺎ ﺗﻮﻟﯿﺪ ﭘﺎراﻣﺘﺮﻫﺎي اﺻﻠﯽ، ﻋﺎﻣﻞﻫﺎي رﻓﺘﺎري ﻓﺮا اﻋﺘﻤﺎدي و اﺣﺴﺎﺳﺎت ﺑﺎزار اﻋﻤﺎل ﮔﺮدﯾﺪه و ﻣﯿﺰان ﺗﻄﺒﯿﻖ ﺧﺮوﺟﯽ ﺷﺒﯿﻪﺳﺎزي ﺑﺎ دادهﻫﺎي ﺑﺎزار واﻗﻌﯽ ﻣﻮرد ﺑﺮرﺳﯽ، آزﻣﻮن و ﺗﺤﻠﯿﻞ ﻗﺮار ﮔﺮﻓﺖ و ﺑﻬﺘﺮﯾﻦ اﺳﺘﺮاﺗﮋي ﺳﺮﻣﺎﯾﻪ ﮔﺬاري اﻧﺘﺨﺎب و ﺑﺎ اﺳﺘﻔﺎده از اﻟﮕﻮرﯾﺘﻢ ژﻧﺘﯿﮏ ﺑﻬﯿﻨﻪ ﺳﺎزي ﺷﺪ ﻧﺘﺎﯾﺞ ﭘﮋوﻫﺶ ﻧﺸﺎن ﻣﯽدﻫﺪ ﮐﻪ ﻣﺪل ﻋﺎﻣﻞ ﻧﺎﻫﻤﮕﻦ، اﺳﺘﺮاﺗﮋي دﻧﺒﺎل ﮐﻨﻨﺪﮔﺎن روﻧﺪ ﻣﺘﻀﺎد را ﺑﻪ ﺻﻮرت ﻣﻨﺎﺳﺒﯽ ﭘﯿﺶ ﺑﯿﻨﯽ ﻣﯽ ﻧﻤﺎﯾﺪ و ﻫﻤﭽﻨﯿﻦ ﺑﺎ اﺳﺘﻔﺎده از اﻟﮕﻮرﯾﺘﻢ ژﻧﺘﯿﮏ ﻣﯽﺗﻮان اﻧﺤﺮاف ﻣﻌﯿﺎر و ﻣﯿﺎﻧﮕﯿﻦ ﺿﺮاﯾﺐ اﺳﺘﺮاﺗﮋيﻫﺎي ﺳﺮﻣﺎﯾﻪﮔﺬاري را ﺑﻬﺒﻮد داده و ﺗﻌﺪاد ﺗﻄﺒﯿﻖ ﺑﺎ دادهﻫﺎي ﺑﺎزار واﻗﻌﯽ را اﻓﺰاﯾﺶ داد

This paper survey beliefs of investor on Tehran stock exchange at three break point date (BPD). at first three BPD with several criterion : average of the share price indices , average value of the stock market turnover , average value of the stock market capitalization .according this three BPD are : the election of Mahmood Ahmadinejad at 2005 , financial crisis at 2008 and the election of Hassan rouhani at 2013 . In addition this paper is base of Brock and Hommes heterogeneous agent model (HAM) framework. Samples are the shares of companies that 40 days before and 40 days after was traded .then with MATLAB software code was writhed and simulation done. Finding shows that strategy of contrarian trend chaser is the best and we can with genetic algorithm optimize average and standard deviation of coefficient of investment strategy and adaption with real market at break point dates.