Dependent Variable: LOG(Y)   
Method: Least Squares   
Date: 03/18/12   Time: 14:34   
Sample (adjusted): 1995Q2 2011Q4   
Included observations: 67 after adjustments   
Convergence achieved after 9 iterations   
   
Variable Coefficient Std. Error t-Statistic Prob.  
   
C -1.514036 0.875055 -1.730218 0.0886
X1 0.003278 0.002527 1.297120 0.1994
X2 0.525178 0.142449 3.686780 0.0005
X3 -0.013303 0.017436 -0.762950 0.4484
ar(1) 0.986600 0.020078 49.13875 0.0000
   
R-squared 0.983711     Mean dependent var  -0.874449
Adjusted R-squared 0.982660     S.D. dependent var  0.177436
S.E. of regression 0.023365     Akaike info criterion  -4.603443
Sum squared resid 0.033848     Schwarz criterion  -4.438913
Log likelihood 159.2153     Hannan-Quinn criter.  -4.538338
F-statistic 936.0465     Durbin-Watson stat  1.966890
Prob(F-statistic) 0.000000   
   
Inverted AR Roots       .99   

请问Inverted AR Roots   ？这个值有什么含义。只知道当这个值等于1的时候，AR过程是不稳定的，可是趋近于1，如这儿的0.99呢？是好是坏