Susceptible Reservoir Architectures for Regime-Conditional Volatility Forecasting
2026-07-24 • Machine Learning
Machine Learning
AI summaryⓘ
The authors developed new methods called Susceptible Architectures (SUSA) to improve predictions of stock market volatility by using specially designed systems (reservoirs) that respond to different market conditions like calm or stress. They tested these methods on various U.S. stocks and exchange-traded funds, comparing their performance to established models like GARCH. Their models performed well, especially on certain assets, and worked well when combined with other forecasting approaches to give better overall predictions. The authors also explored quantum computing versions of their models for potential future use.
volatility forecastingreservoir computingGARCH modelQLIKE loss functionHARQ modelensemble learningquantum computingregime switchingtime series predictionfinancial econometrics
Authors
Aliaksei Kaliutau
Abstract
Volatility forecasting is dominated by persistence and measurement noise, leaving limited residual structure for nonlinear models to exploit. We introduce Susceptible Architectures (SUSA), a reservoir-design principle for volatility forecasting, and its two concrete implementations, based on complex-valued open-chain and periodic reservoirs and regime-conditioned experts to interpret reservoir features across calm, onset, recovery, and persistent-stress states. We also implement open-system $q$-qubit counterparts in Qiskit while retaining a common AR-Ridge anchor and a bounded residual correction trained under QLIKE. We evaluate models on 16 U.S. equity and exchange-traded-fund series using three disjoint chronological training, validation, and test folds, a 12-observation input window, and a five-observation forecast horizon. The proposed models perform competitively with GARCH, achieving statistically significant QLIKE improvements for specific assets (IWM, XLP). Also models' forecasts complement HARQ-style predictions: a stacked ensemble improves mean QLIKE by 0.0116 over its strongest constituent and wins in 75% of test scenarios.