以ar(3)-garch(2,1)模型為例：
首先在主視窗輸入
lsrrrr（-1）（-2）（-3）
得出
variablecoefficientstd.errort-statisticprob.
rr(-1)0.0076060.0590140.1288830.8975
rr(-2)0.0580050.0585490.9907070.3227
rr(-3)0.1211100.0589852.0532450.0410
然後在點estimate在下拉選項中選擇arch
在命令視窗中再次輸入
並在arch出填入2，garch處為1，得出結果
variancebackcast:on
garch=c(4)+c(5)*resid(-1)^2+c(6)*resid(-2)^2+c(7)
*garch(-1)
coefficientstd.errorz-statisticprob.
rr(-1)0.0133920.0568630.2355140.8138
rr(-2)0.1204810.0621461.9386710.0525
rr(-3)0.0959210.0560701.7107430.0871
varianceequation
c0.0001273.59e-053.5533270.0004
resid(-1)^2-0.0439070.029463-1.4902530.1362
resid(-2)^20.2486250.0788553.1529600.0016
garch(-1)0.0797690.2119420.3763720.7066
r-squared0.003674meandependentvar0.001397
adjustedr-squared-0.017908s.d.dependentvar0.013305
s.e.ofregression0.013423akaikeinfocriterion-5.819411
sumsquaredresid0.049910schwarzcriterion-5.729472
loglikelihood833.3564durbin-watsonstat1.974819
rr是上證綜合指數的周收益，用此ar（3）-garch（2，1）是用殘差來檢驗超額收益的。
以ar(3)-garch(2,1)模型為例：
首先在主視窗輸入
lsrrrr（-1）（-2）（-3）
得出
variablecoefficientstd.errort-statisticprob.
rr(-1)0.0076060.0590140.1288830.8975
rr(-2)0.0580050.0585490.9907070.3227
rr(-3)0.1211100.0589852.0532450.0410
然後在點estimate在下拉選項中選擇arch
在命令視窗中再次輸入
lsrrrr（-1）（-2）（-3）
並在arch出填入2，garch處為1，得出結果
variancebackcast:on
garch=c(4)+c(5)*resid(-1)^2+c(6)*resid(-2)^2+c(7)
*garch(-1)
coefficientstd.errorz-statisticprob.
rr(-1)0.0133920.0568630.2355140.8138
rr(-2)0.1204810.0621461.9386710.0525
rr(-3)0.0959210.0560701.7107430.0871
varianceequation
c0.0001273.59e-053.5533270.0004
resid(-1)^2-0.0439070.029463-1.4902530.1362
resid(-2)^20.2486250.0788553.1529600.0016
garch(-1)0.0797690.2119420.3763720.7066
r-squared0.003674meandependentvar0.001397
adjustedr-squared-0.017908s.d.dependentvar0.013305
s.e.ofregression0.013423akaikeinfocriterion-5.819411
sumsquaredresid0.049910schwarzcriterion-5.729472
loglikelihood833.3564durbin-watsonstat1.974819
rr是上證綜合指數的周收益，用此ar（3）-garch（2，1）是用殘差來檢驗超額收益的。