基于R语言的迪基-福勒检验的蒙特卡罗模拟

迪基-福勒(Dickey-Fuller)检验是时间序列的平稳性检验中常用的一种方法;由于检验统计量的极限分布是由标准维纳过程关于轨道的积分来表达的,很难得到其密度函数的显式表达式,因而确定检验临界值非常困难,蒙特卡罗方法是解决这类问题的金钥匙;基于R语言对t统计量的平稳性检验的临界值的随机模拟程序进行了研究,填补了文献空白,其计算程序和方法对于金融工程或经济计量统计分析与研究具有广泛的指导意义。

Monte Carlo Simulation of Dickey-Fuller Test Based on R Language

Dickey Fuller test is a commonly used method in the stationary test of time series.Because the limit distribution of its statistics is expressed by the integral of the standard Wiener process about its trajectory. It is very difficult to obtain the explicit expression of its density function, so it is very difficult to determine the critical value of test. Monte Carlo method is the golden key to solve this kind of problem. Based on R language, we study the critical value of stochastic simulation program for statistic stationarity test, which fills in the blank of literature. Its calculation program and method have a wide range of guiding significance for financial engineering or econometric statistical analysis and research.