احسان طیبی ثانی

In this research, due to the importance of investing and especially investing in the stock market, we predicted the stock price return on the stock exchange through the Auto-Regressive Integrated Moving Average (ARIMA) and Recursive Neural Network (RNN) of long-short term memory (LSTM). Then, to reduce the risk of decision-making, we compared the predictive power of these two models to determine a better model. The research variable is the stock price of the top 20 (in market cap) companies on the stock exchange for the period of the 11th Feb 2015 to 22th Jan 2022. We considered the data of the last 10 days as experimental data and the previous data as educational data. Initially, we calculated the mean and standard deviation of the prediction error of both models; these criteria had less value for the LSTM recursive neural network model than the ARIMA model. To measure the significance of this difference in predictive power, we used Harvey, Liborne, and New Bold tests. The results showed that in predicting the stock prices of the top 20 companies of the stock exchange, the predictive power of the LSTM recursive neural network model was statistically and significantly higher than the ARIMA model which means better predition of stock prices and higher return for investors. In the end, it is believed that the LSTM model may have the best predictive ability, but it is greatly affected by the data processing.

Purpose of this research is Predicting Stock Price Crash Risk with a Deep Learning Approach from Artificial Intelligence and Comparing its Efficiency with Classical Predicting Methods. This research is post-event correlation type and practical in terms of purpose. The research data were extracted from the website of the Stock Exchange Organization and Codal website. The risk variable of crashing stock prices was introduced as a predictor. 3200 obser-vations were obtained from 10-year data of 320 companies between 2012 and 2021. In the following, 29 variables were identified as variables that can affect the risk of crashing stock prices. Statistical methods such as unit root test, composite data, Hausman test and variance heterogeneity test were used. Next, the top 10 algorithms in the field of deep learning were selected and used to model the mentioned variables with the CNN method. Python, Eviews and Excel software were used in this research. Examining the performance of different deep learning algorithms shows that the convolutional neural network method performs better compared to other algorithms and can improve the prediction accuracy. Therefore, it is suggested to use this algorithm in reviewing econometric data and especially predicting the risk of crashing stock prices.

According to the literature, macroeconomic variables have significant effects on financial markets. In addition to investors and traders in these markets, researchers have also paid special attention and sensitivity to these changes. The purpose of this study is to investigate the macro-structural determinants affecting the price index of the Tehran Stock Exchange in the period 1991-2019. To this purpose, the fully modified ordinary least squares estimator (FMOLS) and the Hudrick Prescott filter (HP) were used. Based on the estimation results of the econometric model, economic growth, government budget deficit, and exchange rate have had positive and significant effects on the total price index of the Tehran Stock Exchange, while negative effects on money supply (liquidity) and oil revenue uncertainty index (extracted by HP filter). Economic growth has had a significant effect on the total price index of the Tehran Stock Exchange resulting in negative returns. JEL Classification: G12, C50, C22, E44