China semiconductor industry policy, industry efficiency and synergic aggregation based on GARCH analysis

The article presents a description of the development trend of the semiconductor industry, on the basis of which it is concluded that it is necessary to assess the efficiency of the industry. It is shown that with the help of estimates of variance based on extreme values, it is possible to construct conditional variance models with better predictive properties than in the case of models based solely on asset returns. The theoretical properties of extreme value estimates suggest that their use in conditional variance models will provide more accurate estimates of volatility. The article presents a review of GARCH class models, which are not only the most widely used, but also the most flexible tool for modeling and forecasting conditional variance.