Designing a NISQ reservoir with maximal memory capacity for volatility forecasting

Current noisy intermediate-scale quantum (NISQ) devices are far from being fault-tolerant, however, they permit some limited applications. In this study, we use a hybrid quantum-classical reservoir computing model to forecast the CBOE volatility index (VIX) using the S&P500 (SPX) time-series. The NISQ component of our model is executed on IBMs 53 qubit Rochester chip. We encode the SPX values in the rotation angles and linearly combine the average spin of the six-qubit register to predict the value of VIX at next time step. Our results demonstrate a potential approach towards utilizing noisy quantum devices for non-linear time-series forecasting tasks.