Detection and forecasting of extreme events in stock price triggered by fundamental, technical, and external factors

The sporadic large fluctuations are seen in the stock market due to changes fundamental parameters, technical setups, and external factors. These termed as Extreme Events (EE). EEs may be positive or negative depending on impact of these During such events, price time series is found nonstationary. Hence, Hilbert-Huang transformation (HHT) used identify based their high instantaneous energy ($IE$) concentration. analysis shows that $IE$ concentration very during both EE with $IE>E_{\mu}+4\sigma,$ where $E_{\mu}$ $\sigma$ mean standard deviation energy, respectively. Further, support vector regression predict an EE, close being most helpful input than open-high-low-close (OHLC) inputs. maximum prediction accuracy for one step using OHLC prices 95.98\% 95.64\% Whereas, two steps prediction, accuracies 94.09\% 93.58\% from predicted similar statistical characteristics were obtained original data. emphasizes importance monitoring factors lead a compelling entry exit strategy investors can gain lose significant amounts capital events.