[R-sig-finance] GARCH(1,1)

[Alessandro Correa]

I have run a GARCH(1,1) model using Eviews Matlab and R.
The results are too diferent.
In special, take a look at the constant of the mean equation.
Only Eviews returns a contant=mean of the serie (-0.068513).
Does anyone knows the reason for this?

************** EVIEWS 3 *****************************
Dependent Variable: Y    
Method: ML - ARCH
Sample: 1 550    
Included observations: 550    
Convergence achieved after 28 iterations    
    
 Coefficient Std. Error z-Statistic Prob.  
    
C -0.065643  0.090032 -0.729105  0.4659
    
        Variance Equation   
    
C  0.967087  0.446945  2.163770  0.0305
ARCH(1)  0.083785  0.037083  2.259421  0.0239
GARCH(1) 0.684535  0.127228  5.380361  0.0000
    
R-squared -0.000002 Mean dependent var -0.068513
Log likelihood -1171.509 Durbin-Watson stat 1.9427

************** R *****************************************
Model:
GARCH(1,1)

Coefficient(s):
    Estimate  Std. Error  t value Pr(>|t|)    
a0   0.70954     0.38302    1.852   0.0640 .  
a1   0.06994     0.03180    2.199   0.0278 *  
b1   0.75968     0.11354    6.691 2.21e-11 ***
---
Signif. codes:  0 `***' 0.001 `**' 0.01 `*' 0.05 `.' 0.1 ` ' 1 


************** MATLAB *********************************
Number of Parameters Estimated: 4

                               Standard          T     
  Parameter       Value          Error       Statistic 
 -----------   -----------   ------------   -----------
          C    -0.05598       0.089745        -0.6238
          K    0.72795        0.39708          1.8333
   GARCH(1)    0.75555        0.11697          6.4594
    ARCH(1)    0.07061        0.033108         2.1327

Alessandro de Castro Corrêa

	[[alternative HTML version deleted]]