Skip to content
All library documents

Options-Based Rough Volatility Estimates Improve HAR Forecasts

Article arXiv papers · Author: Zheqi Fan et al.
Read the original

Summary

This study asks whether estimates of spot volatility extracted from traded options can improve forecasts of realized volatility. It augments a Heterogeneous Autoregressive model for realized volatility with a rough stochastic volatility estimate, inferred through an iterative two-step procedure based on earlier work. A deep learning surrogate is used to speed up estimation across large options panels.

The augmented HAR-RV-RHeston specification is compared with traditional stochastic volatility models, including Heston, Bates, and SVCJ, as well as the VIX index. The reported results show improved daily realized volatility forecast accuracy, with superior performance sustained across horizons up to one month. The document presents forecasting comparisons but does not specify the data period, evaluation metrics, or statistical significance, so the strength and generality of the reported gains cannot be assessed from this summary alone.

Key ideas

  • The model tests whether options-implied spot volatility adds predictive value to a HAR realized-volatility model.
  • Spot volatility is inferred under a rough stochastic volatility model using an iterative two-step method.
  • A deep learning surrogate speeds estimation from large options panels.
  • The augmented model is benchmarked against Heston, Bates, SVCJ, and VIX.
  • Reported forecast improvements extend from daily predictions to horizons of up to one month.

Tags

OptionsVolatilityMachine learningStatisticsDerivatives pricing
Full text
# On options-driven realized volatility forecasting: Information gains via rough volatility model


# On options-driven realized volatility forecasting: Information gains via rough volatility model









We examine whether model-based spot volatility estimators extracted from traded options data enhance the predictive power of the Heterogeneous Autoregressive (HAR) model for realized volatility. Specifically, we infer spot volatility under the rough stochastic volatility model via an iterative two-step approach following Andersen et al. (2015a) and adopt a deep learning surrogate to accelerate model estimation from large-scale options panels. Benchmarked against traditional stochastic volatility models (Heston, Bates, SVCJ) and the VIX index, our results demonstrate that the augmented HAR-RV-RHeston model improves daily realized volatility forecasting accuracy and sustains superior performance across horizons up to one month.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.