月度CPI 数据,用R语言ur.df做ADF检验,结果是:
###############################################
# Augmented Dickey-Fuller Test Unit Root Test #
###############################################
Test regression trend
Call:
lm(formula = z.diff ~ z.lag.1 + 1 + tt + z.diff.lag)
Residuals:
Min 1Q Median 3Q Max
-0.93067 -0.26553 -0.02758 0.20786 1.59219
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 50.206022 20.149122 2.492 0.0181 *
z.lag.1 -0.493135 0.198130 -2.489 0.0182 *
tt 0.001870 0.008217 0.228 0.8215
z.diff.lag -0.205406 0.181128 -1.134 0.2652
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 0.4964 on 32 degrees of freedom
Multiple R-squared: 0.3228, Adjusted R-squared: 0.2593
F-statistic: 5.084 on 3 and 32 DF, p-value: 0.005436
Value of test-statistic is: -2.4889 2.1185 3.1766
Critical values for test statistics:
1pct 5pct 10pct
tau3 -4.15 -3.50 -3.18
phi2 7.02 5.13 4.31
phi3 9.31 6.73 5.61
这一下三个类型的统计量都落在了置信区间内,无法拒绝原假设(存在单位根,存在截距项,存在线性趋势),这下怎么搞?
序列差分之后倒是很正常了
Value of test-statistic is: -4.816 7.8299 11.6497
Critical values for test statistics:
1pct 5pct 10pct
tau3 -4.15 -3.50 -3.18
phi2 7.02 5.13 4.31
phi3 9.31 6.73 5.61
都拒绝了原假设
该怎么搞?是只做一阶差分,还是联通截距趋势一起建模??
求指点