CAJ | 학술논문

在低频数据领域内，向量GARCH模型和向量SV模型的参数难于准确估计，利用这些模型很难解决多个资产的协方差矩阵的预测问题．向量ARFIMA模型可以对利用高频金融数据计算得到的多个资产收益的协方差矩阵进行建模，但是随着变量维数的增加，向量ARFIMA模型同样也面临着参数过多而难于准确估计的问题．因此，提出了基于金融高频数据的正交ARFIMA模型．正交ARFIMA模型通过主成分分析将一组变量的协方差矩阵问题转化成了分别考虑它们的主成分的一元波动问题，这样一元的ARFIMA模型可以很直接的得到应用．正交ARFIMA模型通过主成分分析的方法有效的降低了变量的维数，使得其参数估计问题得到很好的解决，对于金融工具的定价、资产配置、风险管理等问题的解决有着深刻的意义．
재저빈수거영역내，향량GARCH모형화향량SV모형적삼수난우준학고계，이용저사모형흔난해결다개자산적협방차구진적예측문제．향량ARFIMA모형가이대이용고빈금융수거계산득도적다개자산수익적협방차구진진행건모，단시수착변량유수적증가，향량ARFIMA모형동양야면림착삼수과다이난우준학고계적문제．인차，제출료기우금융고빈수거적정교ARFIMA모형．정교ARFIMA모형통과주성분분석장일조변량적협방차구진문제전화성료분별고필타문적주성분적일원파동문제，저양일원적ARFIMA모형가이흔직접적득도응용．정교ARFIMA모형통과주성분분석적방법유효적강저료변량적유수，사득기삼수고계문제득도흔호적해결，대우금융공구적정개、자산배치、풍험관리등문제적해결유착심각적의의．
Vector GARCH model and vector SV model, which are based on the low frequency data, are difficult to be estimated. So it isn＇t easy to apply them to forecast the covariance matrix of many assets. Vector ARFIMA model can be used to model the covariance matrix of many assets computed by using high frequency financial data, but it is difficult to be estimated when the mention grows. This paper puts forward orthogonal ARFIMA model, which changes the modeling of the covadance matrix of many assets to the modeling of the variances of their principal factors by principal factor analysis. Orthogonal ARFIMA model lowers the mention of many assets, and it is easy to be estimated. It is meaningful to asset pricing, asset allocation, risk management and so on.