我选用的是钢铁板块2010-7-1至2018-7-1的收盘价数据，取对数收益率以后研究分布特征，呈现高峰厚尾，但没有明显自相关关系，随后进行ARCH检验，结果拒绝原假设，采用AR(0)-GARCH进行拟合。结果无论GARCH(1,1)还是高阶GARCH，都没有办法解决残差存在的ARCH效应。请问要怎么办。。
下面是代码和结果
> m2=garchFit(~arma(0,0)+garch(1,1),data = r,trace = F)
> summary(m2)

Title:
GARCH Modelling

Call:
garchFit(formula = ~arma(0, 0) + garch(1, 1), data = r, trace = F)

Mean and Variance Equation:
data ~ arma(0, 0) + garch(1, 1)
<environment: 0x0000000024ef1370>
[data = r]

Conditional Distribution:
norm

Coefficient(s):
mu omega alpha1 beta1
0.0150790.0342450.0657280.925587

Std. Errors:
based on Hessian

Error Analysis:
EstimateStd. Errort value Pr(>|t|)
mu 0.015079 0.032765 0.4600.64537
omega 0.034245 0.011596 2.9530.00315 **
alpha10.065728 0.009286 7.078 1.46e-12 ***
beta1 0.925587 0.010586 87.438< 2e-16 ***
---
Signif. codes:0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Log Likelihood:
-3646.474 normalized:-1.874794

Description:
Sat Dec 01 16:22:48 2018 by user: You


Standardised Residuals Tests:
Statistic p-Value
Jarque-Bera Test R Chi^2763.28020
Shapiro-Wilk TestR W 0.9734339 0
Ljung-Box Test R Q(10)6.7577630.7480986
Ljung-Box Test R Q(15)8.8175420.8868627
Ljung-Box Test R Q(20)12.849790.8837352
Ljung-Box Test R^2Q(10)56.130591.94046e-08
Ljung-Box Test R^2Q(15)57.631286.410778e-07
Ljung-Box Test R^2Q(20)59.183469.51973e-06
LM Arch Test R TR^2 85.406433.801404e-13

Information Criterion Statistics:
AIC BIC SIC HQIC
3.753701 3.765162 3.753693 3.757915



标准化残差平方不通过Ljung-Box检验，存在ARCH效应，LMARCH检验也拒绝原假设。