Multi-step-ahead stock price index forecasting using long short-term memory model with multivariate empirical mode decomposition

•MEMD- LSTM model for multi-step ahead stock price forecasting was built.•Multi-step ahead forecasting was based on the multiple-input multiple-output strategy.•MEMD was employed to decompose the original time series without information loss.•The proposed model demonstrated its superiority than other machine learning models.•The methodology goes well beyond straightforward application of the stock market.