r1.2=garchFit(~arma(1,1)+garch(3,3),data=r,include.mean=T,cond.dist="norm",trace=F)
r1.2
出现
警告信息：1: In arima(.series$x, order = c(u, 0, v), include.mean = include.mean) :  possible convergence problem: optim gave code=12: In sqrt(diag(fit$cvar)) : 产生了NaNs> r1.2
Title: GARCH Modelling
Call: garchFit(formula = ~arma(1, 1) + garch(3, 3), data = r, cond.dist = "norm",     include.mean = T, trace = F)
Mean and Variance Equation: data ~ arma(1, 1) + garch(3, 3) [data = r]
Conditional Distribution: norm
Coefficient(s):         mu          ar1          ma1        omega       alpha1   0.06690990  -0.77288507   0.80637339   0.06132273   0.05217649       alpha2       alpha3        beta1        beta2        beta3   0.06315568   0.00000001   0.00000001   0.26568818   0.60267032  
Std. Errors: based on Hessian
Error Analysis:         Estimate  Std. Error  t value Pr(>|t|)    mu      6.691e-02   6.341e-02    1.055  0.29132    ar1    -7.729e-01   1.023e-01   -7.553 4.26e-14 ***ma1     8.064e-01   9.657e-02    8.351  < 2e-16 ***omega   6.132e-02   1.196e-02    5.129 2.92e-07 ***alpha1  5.218e-02   2.115e-02    2.467  0.01363 *  alpha2  6.316e-02   2.333e-02    2.707  0.00678 ** alpha3  1.000e-08   1.374e-02 7.28e-07  1.00000    beta1   1.000e-08          NA       NA       NA beta2   2.657e-01   2.523e-01    1.053  0.29232    beta3   6.027e-01          NA       NA       NA    ---Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Log Likelihood: -3961.721    normalized:  -1.966115
Description: Sun Jul 28 11:07:09 2013 by user: Administrator 原因是什么啊？