| Abstract [eng] |
Market volatility forecasting has been an important research topic in recent decades. This paper aims to forecast realized market volatility for SP500 stocks by using recurrence plots, RQA and deep neural networks. The experiments done in this paper were based on analysis done for daily stock market price changes. Logarithmic returns were calculated from daily closing stock prices which then were used to generate recurrence plots and RQA metrics. The resulting data was used to train GARCH, RNN, LSTM and ResNet models. These models were trained to forecast intervals in which stock market volatility should fall after 1, 2, 5, 10 and 21 days. Obtained results showed that RNN models give the best results when forecasting volatility after 1, 2, 5 and 10 days, while ResNet and GARCH models give better results for predictions after 21 days. |
