【摘要】 创业板汇集了大批高成长性的科技企业,面临着高度不确定的市场风险。本文考虑创业板市场的非对称性和厚尾性,在偏t分布下利用5分钟高频数据构建已实现GARCH模型衡量其波动率,同时考虑到微观结构噪声影响,在RV的基础上引入对噪声稳健的BPV已实现测度,从而提升高频Va R模型的预测准确性。对创业板指和深证成指的实证结果表明,已实现GARCH模型在参数显著性和模型预测准确性方面均优于低频下的ARMA-f GARCH-Va R模型;基于RV和BPV的已实现模型没有显著差异;相比于深市主板,创业板面临的尾部风险更大。同时,为更好地监测与管控创业板市场风险,从政府和创业企业两个层面提出对策建议。更多还原

【Abstract】 The Growth Enterprises Market( GEM) has gathered a large number of high growth technology enterprises,which are faced with highly uncertain market risks. In this paper,considering the asymmetry and thick tail of GEM market,we use5-minute high-frequency data to build the realized GARCH model to measure its volatility under skewed t distribution. At the same time,considering the influence of microstructure noise,we introduce the noise robust BPV realized measure on the basis of RV,so as to improve the prediction accuracy of high-frequency Va R model. The empirical results of GE Index and SZSE composite index show that the realized GARCH model is better than ARMA-fgarch-Va R model in terms of parameter significance and model prediction accuracy; there is no significant difference between the realized models based on RV and BPV; compared with the main board of Shenzhen stock market,GEM faces greater tail risk. At the same time,in order to better monitor and control the GEM market risk,this paper puts forward countermeasures and suggestions from the government and the entrepreneurial enterprises.更多还原