首页|半参数平滑转换分位数自回归模型及应用

半参数平滑转换分位数自回归模型及应用

扫码查看
已有针对平滑转换自回归模型(STAR)的研究多是将转换函数设定为Logistic函数或指数函数形式,并在均值回归框架下获得模型的估计、检验及预测结果.文章基于重心权有理插值和分位数回归方法,构建一类新的半参数平滑转换分位数自回归模型,其主要特点表现在:第一,基于重心权有理插值方法构造的平滑转换函数,形式更加灵活自由,有效减少了模型误设的风险.第二,在分位数回归框架下,利用遗传算法获得新模型在不同分位点处的平滑转换自回归系数估计,比单纯的均值回归得到的信息更为丰富.数值模拟结果显示,新模型的平滑转换自回归系数估计在无偏性、有效性和一致性方面均具有较好表现.最后,将新模型应用于上证综指日收益率的动态趋势及预测研究,细致揭示了收益率序列在不同阶段、不同分位点处的非线性和异质性变化特征.
Semiparametric Smooth Transition Quantile Autoregreeive Model and Its Application
Smooth transition autoregressive model(STAR)often sets the transi-tion function as logistic function or exponential function,and obtains the estimation,test and forecasting results by mean regression.We propose a new semi-parametric smooth transition quantile autoregressive model.Its main advantages are as follows:First,the smooth transition function is constructed by the barycentric rational in-terpolation function,which has more flexible form and can better reduce the risk of mispecification.Second,in the framework of quantile regression,the genetic al-gorithm is applied to obtain the coefficient estimation of new model,which is more informative than the mean regression.Numerical simulation results show that the autoregressive coefficient estimators have good performance in unbiasedness,effec-tiveness and consistency.Finally,the new model is applied to reveal and forecast the dynamic trend of stock returns of the Shanghai Composite Index,and the empirical results indicate that there exist nonlinear and heterogeneous characteristics of the return series.

Barycentric rational interpolationquantile regressionsmooth transi-tion autoregressiongenetic algorithm

康宁、莫璐瑶、荆科

展开 >

南京财经大学经济学院,南京 210023

南京财经大学应用数学学院,南京 210023

重心权有理插值 分位数回归 平滑转换自回归 遗传算法

国家自然科学基金青年基金教育部人文社会科学研究青年基金

1200126618YJC790069

2024

系统科学与数学
中国科学院数学与系统科学研究院

系统科学与数学

CSTPCD北大核心
影响因子:0.425
ISSN:1000-0577
年,卷(期):2024.44(3)
  • 30