请问大神，为什么我做的这个误差修正模型是错的呢？应该怎么改正

1. library(vars)
   
2. GDP<-c(443.93,520.10,576.20,695.65,850.49,970.70,1086.17,1184.80,1305.70,
   
3.        1422.80,1607.70,1801.70)
   
4. Loan<-c(268.6,283.1,389.9,475.2777,585.2114,715.3584,901.1426,
   
5.         1045.3982,1208.9791,1432.0018,1703.3976,2049.9918)
   
6. gdp<-log(GDP);loan<-log(Loan);sg<-loan-gdp
   
7. ur.gdp<-ur.df(diff(gdp),type="drift",lag=3);summary(ur.gdp)
   
8. ur.sg<-ur.df(diff(sg),type="trend",lag=1);summary(ur.sg)
   
9. new.lm<-lm(gdp~sg)
   
10. error<-residuals(new.lm)
    
11. ur.er<-ur.df(error,lag=0,type="none");summary(ur.er)
    
12. error.new<-head(error,-1)
    
13. dy<-diff(gdp);dx<-diff(sg)
    
14. model.ecm<-lm(dy~dx+error.new)
    
15. summary(model.ecm)

复制代码

Call:
lm(formula = dy ~ dx + error.new)

Residuals:
     Min       1Q   Median       3Q      Max
-0.03496 -0.02404 -0.01050  0.01674  0.06305

Coefficients:
            Estimate Std. Error t value Pr(>|t|)   
(Intercept)  0.13687    0.01662   8.236 3.54e-05 ***
dx          -0.16007    0.23002  -0.696    0.506   
error.new   -0.09496    0.13841  -0.686    0.512   
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.03507 on 8 degrees of freedom
Multiple R-squared:  0.363,

gdp和loan如果是趋势的有滞后阶数是不是不能这么直接做呢