New technique boosts accuracy of stock market volatility forecasts exponentially!
The researchers developed a new method to improve forecasting of asset return volatility. By using a multivariate stochastic volatility model and a technique called volatility reprojection, they were able to create a more accurate representation of volatility. This method outperformed traditional approaches using squared daily asset returns. The study showed that the reprojected volatility series led to significantly better volatility forecasting results, especially when considering the correlated movements of asset return volatility.