Abstract:Research has substantiated that the presence of outliers in data usually introduces additional errors and biases, which typically leads to a degradation in the precision of volatility forecasts. However, correcting outliers can mitigate these adverse effects. This study corrects the additive outliers through a weighting method and let these corrected values to replace the original outliers. Then, the model parameters are re-estimated based on this new return series. This approach reduces the extent to which outliers distort volatility estimates, allowing the model to better adapt to market conditions and improving the accuracy of volatility forecasts. This study introduces this approach for the first time to generalized autoregressive conditional heteroskedasticity mixed data sampling (GARCH-MIDAS) models, so as to establish an additional outliers corrected GARCH-MIDAS model (AO-GARCH-MIDAS). This pioneering approach marks a unique innovation. The research employs a diverse array of evaluation methods to validate the model's robustness and consistently demonstrates its dependable performance. Findings unequivocally reveal the substantial influence of outliers on the model's predictive capacity, with the AO-GARCH-MIDAS model exhibiting consistent superiority across all evaluation criteria. Additionally, while the GARCH model showcases stronger estimation capabilities compared to the GARCH-MIDAS model, the latter demonstrates heightened predictive prowess. Notably, regarding variable selection, the results underscore the greater predictive informational value inherent in realized volatility over other low-frequency factors.

Predicting the long-term stock market volatility: A GARCH-MIDAS model with variable selection

Forecasting Stock Volatility Using Time-Distance Weighting Fundamental’s Shocks

Prediction of Volatility Based on Realized-Garch-kernel-type Models: Evidence from China and the U.S.

Analysis of Factors Influencing Stock Market Volatility Based on GARCH-MIDAS Model

Heterogeneous macroeconomic factors’ effects on stocks across sizes, styles, and sectors in the South Korean market

Enhancing stock volatility prediction with the AO-GARCH-MIDAS model

Forecasting stock volatility using time-distance weighting fundamental's shocks

Asymmetry effect and jump effect on stock volatility -- Modeling and prediction based on GARCH-MIDAS

Applying a combined fuzzy systems and GARCH model to adaptively forecast stock market volatility

Forecasting conditional volatility based on hybrid GARCH-type models with long memory, regime switching, leverage effect and heavy-tail: Further evidence from equity market

Volatility Prediction via Hybrid LSTM Models with GARCH Type Parameters

Volatility analysis based on GARCH-type models: Evidence from the Chinese stock market

A long short-term memory enhanced realized conditional heteroskedasticity model

Forecasting the volatility of European Union allowance futures with macroeconomic variables using the GJR-GARCH-MIDAS model

Multivariate tempered stable model with long-range dependence and time-varying volatility

A time-varying GARCH mixed-effects model for isolating high- and low- frequency volatility and co-volatility

Forecasting volatility of stock indices: Improved GARCH-type models through combined weighted volatility measure and weighted volatility indicators

Forecasting with VAR Models: Fat Tails and Stochastic Volatility

A Time-Varying Coefficient Double Threshold GARCH Model with Explanatory Variables

Forecasting stock market volatility with a large number of predictors: New evidence from the MS-MIDAS-LASSO model

Forecasting renewable energy stock volatility using short and long-term Markov switching GARCH-MIDAS models: Either, neither or both?