GARCH-model

Hvordan afgør man, om en GARCH-model man har estimeret beskriver data (variablen 100timesLogReturn) tilfredsstillende? Har vedhæftet data i en Excel-fil.

Output:

Dependent variable : 100timesLogReturn
Mean Equation : ARMA (0, 0) model.
No regressor in the conditional mean
Variance Equation : GARCH (1, 1) model.
No regressor in the conditional variance
Normal distribution.

Strong convergence using numerical derivatives
Log-likelihood = -2008.98
Please wait : Computing the Std Errors ...

Robust Standard Errors (Sandwich formula)
                  Coefficient  Std.Error  t-value  t-prob
Cst(V)               0.010869  0.0047526    2.287  0.0223
ARCH(Alpha1)         0.107741   0.019644    5.485  0.0000
GARCH(Beta1)         0.886280   0.019819    44.72  0.0000

No. Observations :      1516  No. Parameters  :         3
Mean (Y)         :   0.00664  Variance (Y)    :   1.60670
Skewness (Y)     :  -0.07797  Kurtosis (Y)    :  14.15044
Log Likelihood   : -2008.982  Alpha[1]+Beta[1]:   0.99402

The sample mean of squared residuals was used to start recursion.
The positivity constraint for the GARCH (1,1) is observed.
This constraint is alpha[L]/[1 - beta(L)] >= 0.
The unconditional variance is 1.81809
The conditions are alpha[0] > 0, alpha[L] + beta[L] < 1 and alpha[i] + beta[i] >= 0.
  => See Doornik & Ooms (2001) for more details.
The condition for existence of the fourth moment of the GARCH is not observed.
The constraint equals 1.0113 and should be < 1.
  => See Ling & McAleer (2001) for details.

TESTS :
=======

Series #1/1: Standardized Residuals
---------
Normality Test

                   Statistic       t-Test      P-Value
Skewness            -0.34030       5.4145  6.1445e-008
Excess Kurtosis      0.83624       6.6571  2.7920e-011
Jarque-Bera           73.431         .NaN  1.1339e-016
---------------
ARCH 1-2 test:    F(2,1511) =  0.83706 [0.4332] 
ARCH 1-5 test:    F(5,1505) =   1.5701 [0.1654] 
ARCH 1-10 test:   F(10,1495)=  0.89224 [0.5397]

Som jeg har forstået det er Alpha[1]+Beta[1] = 0.99402 tegn på misspecifikation (=~ 1). På den positive side er der ingen ARCH-effekter i residualerne, som dog ikke er normale, hvilket ellers er en af modelantagelserne. Men sidstnævnte er vist aldrig tilfældet for data af denne type. Nogle der ka sige om den estimerede GARCH(1, 1)-model er en go model til at beskrive data (variablen 100timesLogReturn)?