Forecasting the Price of the Cryptocurrency Using Linear and Nonlinear Error Correction Model

We employed linear and nonlinear error correction models (ECMs) to predict the log returns of Bitcoin (BTC). The ECM is best model for predicting BTC compared neural network autoregressive in terms RMSE, MAE, MAPE. Using a ECM, we are able understand how affected by other coins. In addition, performed Granger-causality tests on fourteen cryptocurrencies.