Forecasting Tehran Stock Exchange Trend with Time Series Analysis, Fundamental Data, and Sentiment Analysis in News

Abstract

Forecasting financial markets has always been one of the most popular and challenging topics in finance and academic research. Numerous factors affect the trend of financial markets, and careful analysis of each of these factors can play an important role in detecting market trends. Among the most important factors affecting the Tehran Stock Exchange market is the price of the dollar, the price of gold, the news published on the Tehran Stock Exchange, the monthly sales of listed companies in the Tehran stock exchange, and so on. Using the machine learning algorithms and the Fourier series, we have predicted the trend of the total stock index of Tehran and calculated the possible future trend of the total stock index. Also, the relationship and dependence of the total index of Tehran Stock Exchange with the trend of gold and dollar prices have been obtained, the results show a certain dependence between the total index and the price of the dollar and gold. News related to the Tehran Stock Exchange has been collected from several news sources and the positive and negative polarity of this news has been obtained by using natural language processing algorithms; And we have created a total polarity index for all published news about the stock exchange, and the relationship between the polarity index derived from the news and the trend of the total index of the Tehran Stock Exchange is significant and clear. Then, using the monthly sales of companies active in the stock exchange and the market value of each company, the total monthly sales index of the Tehran Stock Exchange is defined, which has a direct and specific relationship with the total index of the Tehran Stock Exchange.