Multistep forecast of the implied volatility surface using deep learning

Modeling implied volatility surface (IVS) is of paramount importance to price and hedge an option. We contribute the literature by modeling entire IVS using convolutional long-short-term memory (ConvLSTM) (LSTM) neural networks produce multivariate multistep forecasts S&P 500 IVS. Using daily SPX options data (2002–2019), we find that both LSTM ConvLSTM fit training extremely well with mean absolute percentage error (MAPE) being 3.56% 3.88%, respectively. The (8.26% MAPE) model significantly outperforms traditional time series models in predicting out sample.