garch模型python步骤_GARCH模型的建模步骤？

泻药，我将建立道琼斯工业平均指数(DJIA)日交易量对数比的ARMA-GARCH模型来演示建模步骤。

原文链接：R语言： GARCH模型股票交易量的研究道琼斯股票市场指数​tecdat.cn

获取数据

load(file='DowEnvironment.RData')

日交易量

每日交易量内发生的 变化。

plot(dj_vol)

首先，我们验证具有常数均值的线性回归在统计上是显着的。

在休息时间= 6时达到最小BIC。

以下是道琼斯日均交易量与水平变化(红线) 。

summary(bp_dj_vol)##

## Optimal (m+1)-segment partition:

##

## Call:

## breakpoints.formula(formula = dj_vol ~ 1, h = 0.1)

##

## Breakpoints at observation number:

##

## m = 1 2499

## m = 2 896 2499

## m = 3 626 1254 2499

## m = 4 342 644 1254 2499

## m = 5 342 644 1219 1649 2499

## m = 6 320 622 924 1251 1649 2499

## m = 7 320 622 924 1251 1692 2172 2499

## m = 8 320 622 924 1251 1561 1863 2172 2499

##

## Corresponding to breakdates:

##

## m = 1

## m = 2 0.296688741721854

## m = 3 0.207284768211921

## m = 4 0.113245033112583 0.213245033112583

## m = 5 0.113245033112583 0.213245033112583

## m = 6 0.105960264900662 0.205960264900662 0.305960264900662

## m = 7 0.105960264900662 0.205960264900662 0.305960264900662

## m = 8 0.105960264900662 0.205960264900662 0.305960264900662

##

## m = 1

## m = 2

## m = 3 0.41523178807947

## m = 4 0.41523178807947

## m = 5 0.40364238410596 0.546026490066225

## m = 6 0.414238410596027 0.546026490066225

## m = 7 0.414238410596027 0.560264900662252

## m = 8 0.414238410596027 0.516887417218543 0.616887417218543

##

## m = 1 0.827483443708609

## m = 2 0.827483443708609

## m = 3 0.827483443708609

## m = 4 0.827483443708609

## m = 5 0.827483443708609

## m = 6 0.827483443708609

## m = 7 0.719205298013245 0.827483443708609

## m = 8 0.719205298013245 0.827483443708609

##

## Fit:

##

## m 0 1 2 3 4 5 6

## RSS 3.872e+19 2.772e+19 1.740e+19 1.547e+19 1.515e+19 1.490e+19 1.475e+19

## BIC 1.206e+05 1.196e+05 1.182e+05 1.179e+05 1.178e+05 1.178e+05 1.178e+05

##

## m 7 8

## RSS 1.472e+19 1.478e+19

## BIC 1.178e+05 1.178e+05

lwd = c(3,1), col = c("red