[问答] 如何用rmgarch拟合bekk_garch模型？ [推广有奖]

2关注
2粉丝

博士生

15%

还不是VIP/贵宾

威望: 0 级
论坛币: 13 个
通用积分: 0.0010
学术水平: 0 点
热心指数: 0 点
信用等级: 0 点
经验: 550 点
帖子: 63
精华: 0
在线时间: 406 小时
注册时间: 2010-7-8
最后登录: 2019-7-24

楼主

icelaugh 发表于 2015-2-2 21:31:33 |AI写论文

是否 +2 论坛币

k人参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

以前似乎是rgarch包，请问怎么写模型，谢谢！

扫码加我拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏9 回帖

关键词：GARCH模型 ARCH模型 MGARCH GARCH ARCH 模型如何

相关帖子

沙发

DM小菜鸟 发表于 2015-2-26 17:27:54

之前有个包叫：mgarchBEKK可以直接用如果你的版本不是新的，可以到hithub找一下这个包

如果是现在用的话，rmgarch的话——

library(rugarch)
library(ccgarch)
library(rmgarch)
library(gogarch)
library(MTS)
library(mgarch)
save.image("~/Documents/sheetR/MGARCH-Examples/Simulations.RData")
#**************************************************************#
########################## M-GARCH #############################
#**************************************************************#
### EXEMPLOS DE CCC-GARCH E DCC-GARCH USANDO O PACOTE 'ccgarch'
### CCC-GARCH
### DCCt-GARCH (Tse e Tsui)
### DCCe-GARCH (Engle)
### CP-GARCH
### F-GARCH
### (G)O-GARCH
### (G)ICA-GARCH

h11.spec<-garchSpec(model = list(omega=0.1,alpha=0.05,beta=0.7),cond.dist = "norm")
h22.spec<-garchSpec(model = list(omega=0.2,alpha=0.5,beta=0.5),cond.dist = "norm")
h11<-garchSim(spec = h11.spec, n = 1000)
h22<-garchSim(spec = h22.spec, n = 1000)
R<-matrix(c(1,0.5,-0.5,1),ncol=2,byrow=T)

#**************************************************************#
######################### 3-BEKK-GARCH ##########################
#**************************************************************#

bekk.spec = garchSpec(model = list(alpha = 0.2, beta = 0.7))
bekk.A0=matrix(c(0.1,0,0,-0.1,0.2,0,0.3,-0.1,0.4),ncol=3,byrow=T)
bekk.A1=matrix(c(0.7,0,0.1,0,0.2,-0.1,0.2,-0.1,0.4),ncol=3,byrow=T)
bekk.B1=matrix(c(0.2,0,-0.5,0,0,0.2,0,-0.1,0.1),ncol=3,byrow=T)
bekk.S0=diag(3)
bekk.epsilon1=garchSim(bekk.spec, n = 1000)
bekk.epsilon2=garchSim(bekk.spec, n = 1000)
bekk.epsilon3=garchSim(bekk.spec, n = 1000)
bekk.epsilon<-cbind(bekk.epsilon1,bekk.epsilon2,bekk.epsilon3)

bekk.S<-array(0,dim=c(1000,3,3))
bekk.S[1,,]=diag(3)

for(t in 2:1000){
  bekk.S[t,,]=bekk.A0%*%t(bekk.A0) + bekk.A1 %*%  t(bekk.epsilon[t,]) %*% bekk.epsilon[t,] %*% t(bekk.A1) + bekk.B1%*%bekk.S[t-1,,]
}
MTSplot(bekk.S)

###### 3.1-Simulacao utilizando o pacote mgarch #####
bekk.sim <- mvBEKK.sim(series.count = 3, T = 2500)
names(bekk.sim)
# > names(bekk.sim)
# [1] "length"          "series.count"    "order"
# [4] "params"          "true.params"    "eigenvalues"
# [7] "uncond.cov.matrix" "white.noise"    "eps"
# [10] "cor"             "sd"
bekk.sim$length
bekk.sim$series.count
bekk.sim$order
bekk.sim$params
bekk.sim$true.params
bekk.sim$eigenvalues
bekk.sim$uncond.cov.matrix
matrix.bekk.sim<-cbind(bekk.sim$eps[[1]],bekk.sim$eps[[2]],bekk.sim$eps[[3]])
plot.ts(matrix.bekk.sim)

##### 3.1.a-Estimacao do BEKK via mgarch #####
fit.bekk.mgarch<-mvBEKK.est(matrix.bekk.sim)
##### 3.2-Estimacao do BEKK via MTS #####
system.time(fit.bekk.mts<-BEKK11(matrix.bekk.sim))
names(fit.bekk.mts)
# [1] "estimates"  "HessianMtx" "Sigma.t"

fit.bekk.mts$estimates