-stationary properties of Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models with the nonlinear modeling capabilities … (BiLSTM) algorithms with single, double, and triple layer network architectures to forecast Bitcoin's realized price …

still sparse. In this study, both conventional econometric models and a machine learning model are used to forecast the … forecasting accuracy and risk management efficiency. The results demonstrate that the RNN outperforms GARCH and EWMA in average …

between the realization and prediction of future spot prices, and what factors affect these forecast errors. The results of … formation process, thus increasing forecast errors. Conversely, long-term speculation, proxied by the Working-T index and the … speculative pressure index, reduces forecast errors although their quantitative effect is negligible. Other relevant factors that …