Blending ARCH Models to Improve Volatility Forecasts

Summary

The document addresses volatility forecasts that miss peaks and troughs by overestimating or underestimating realized volatility. It notes that SVR-GARCH can produce forecasts that largely deviate from the previous volatility, a behavior described as backward-looking. The authors propose blending ARCH (BARCH) and augmented BARCH (aBARCH) models to improve forecasts, particularly their representation of turning points in financial time series.

The approach is illustrated with real data from SH300 and the S&P 500, and the reported empirical results suggest that both blending models improve volatility forecasting ability. The short description does not specify the model equations, forecast horizons, evaluation measures, or the size and consistency of the improvements. It also does not establish that the method generalizes beyond the cited datasets. The document frames the models as forecasting improvements; it does not report a trading strategy or demonstrate resulting portfolio performance.

Key ideas

• SVR-GARCH forecasts may overestimate or underestimate volatility around peaks and troughs.
• BARCH and augmented BARCH blend ARCH-based components to address forecast errors.
• The models aim to better capture turning-point behavior in volatility series.
• The described empirical examples use SH300 and S&P 500 data.
• The reported results indicate improved forecast ability, without specifying effect sizes or generalization tests.

Tags

# Reducing overestimating and underestimating volatility via the augmented blending-ARCH model


# Reducing overestimating and underestimating volatility via the augmented blending-ARCH model









SVR-GARCH model tends to "backward eavesdrop" when forecasting the financial time series volatility in which case it tends to simply produce the prediction by deviating the previous volatility. Though the SVR-GARCH model has achieved good performance in terms of various performance measurements, trading opportunities, peak or trough behaviors in the time series are all hampered by underestimating or overestimating the volatility. We propose a blending ARCH (BARCH) and an augmented BARCH (aBARCH) model to overcome this kind of problem and make the prediction towards better peak or trough behaviors. The method is illustrated using real data sets including SH300 and S&P500. The empirical results obtained suggest that the augmented and blending models improve the volatility forecasting ability.

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.