CAJ | 학술논문

利用日内高频数据，分别通过实现波动率模型和实现二次幂波动模型对资产价格的波动率和连续部分波动率建模，并据此得到资产价格跳跃部分的动态行为模型，分离出发生跳跃的天数、跳跃的大小和方向。结果显示，中国金融市场中的资产跳跃行为密集发生在市场波动性较大的时刻，其大小和间隔期均具有群聚现象，且五种代表性资产价格的跳跃密度均呈左偏分布，说明中国股权市场中向下发生的跳跃多于向上的跳跃。
이용일내고빈수거，분별통과실현파동솔모형화실현이차멱파동모형대자산개격적파동솔화련속부분파동솔건모，병거차득도자산개격도약부분적동태행위모형，분리출발생도약적천수、도약적대소화방향。결과현시，중국금융시장중적자산도약행위밀집발생재시장파동성교대적시각，기대소화간격기균구유군취현상，차오충대표성자산개격적도약밀도균정좌편분포，설명중국고권시장중향하발생적도약다우향상적도약。
Current studies on financial markets'asset prices'jump behavior usually use parameter estimation based on models w hich set various parameters for jumps in analysis of asset prices .T his paper uses intra‐day high frequency data to model volatility with realized volatility model and the continuous part of volatility with realized second power volatility model , which is the basis of the non‐parametric estimation of asset prices'jump behavior .This paper also separates the days on which prices jump ,identify the size and direction of jumps .Empirical results show that jumps happen heavily on days w hich are very volatile ,and the size and durations of jumps also present clustering phenomenon .The jump densities of five representative assets are all left‐skewed ,which means the down‐side jumps happen more often than up‐side jumps .