# Script para el ajuste de series temporales financieras con R. Consultar la ayuda de cada una
# de las funciones para conocer las diferentes opciones.

rm(list = ls())

setwd("G:/PedroTrabajo/USC/Docencia/Master-2010-2011/Proyectos")

# Importar datos

HSI <- read.csv2("G:/PedroTrabajo/USC/Docencia/Master-2010-2011/Proyectos/HSI.csv",header=TRUE,sep=",",dec=".",fill=TRUE)

# Importar librerias necesarias para el análisis. Puede ser necesario instalar dichas librerias desde el CRAN

library(moments)

library(tseries)

library(fBasics)

library(fGarch)

# Hay que instalar la libreria rgarch que se encuentra en la página web http://rgarch.r-forge.r-project.org/
# Para ello hay que guardar la libreria en un directorio y utilizar la opción Instalar paquetes a partir
# de archivos zip locales.

library(rgarch)

# Gráfico de los rendimientos

HSI[,1]<-as.Date(HSI[,1])

plot(HSI,type="l",xlab="Year",ylab="Log-returns of HSI")

# Momentos

mean(HSI[,2])

sd(HSI[,2])

skewness(HSI[,2])

kurtosis(HSI[,2])

# Estimacion kernel

d1 = density(HSI[,2])  

plot(d1$x,d1$y,type='l',lwd = 2, col="black")

# ACF de los rendimientos

acf(HSI[,2],main="ACF of log returns of HSI index")

# Contraste Ljung-Box sobre los rendimientos

Box.test(HSI[,2],lag=12,type="Ljung-Box")

# ACF y PACF de los rendimientos al cuadrado

acf(HSI[,2]^2,main="ACF of log returns of squared HSI index")

pacf(HSI[,2]^2,main="PACF of squared log returns of Hang-Seng index")

# Ljung-Box sobre los residuos

at = HSI[,2] - mean(HSI[,2])

Box.test(at^2,lag=12,type="Ljung-Box")

# Modelo ARCH

ajusteARCH <- garchFit(HSI[,2] ~ garch(11,0), data = HSI[,2], trace = F)

ajusteARCH

ajusteARCH <- garchFit(HSI[,2] ~ garch(10,0), data = HSI[,2], trace = F)

ajusteARCH

# Modelo GARCH

ajusteGARCH <- garchFit(HSI[,2] ~ garch(1,1), data = HSI[,2], trace = F)

ajusteGARCH

# Gráfico de la volatilidad estimada

volatilityGARCH = matrix(NA,nrow=2488,ncol=2)

volatilityGARCH[,1] = HSI[2:2489,1]

volatilityGARCH[,2] = ajusteGARCH@h.t

volatilityGARCH = as.data.frame(volatilityGARCH)

volatilityGARCH[,1] = as.Date(volatilityGARCH[,1])

plot(volatilityGARCH,type="l",ylab="",xlab="Time")

# Modelo IGARCH

spec1 <- ugarchspec(variance.model = list(model = "iGARCH", garchOrder = c(1, 1), 
submodel = NULL, external.regressors = NULL, variance.targeting = FALSE), 
mean.model = list(armaOrder = c(0, 0), include.mean = TRUE, garchInMean = FALSE, 
inMeanType = 1, arfima = FALSE, external.regressors = NULL), distribution.model = "norm", 
start.pars = list(), fixed.pars = list())

ajusteIGARCH <- ugarchfit(spec1,HSI[,2])

ajusteIGARCH

# Modelo EGARCH

spec1 <- ugarchspec(variance.model = list(model = "eGARCH", garchOrder = c(2, 1), 
submodel = NULL, external.regressors = NULL, variance.targeting = FALSE), 
mean.model = list(armaOrder = c(0, 0), include.mean = TRUE, garchInMean = FALSE, 
inMeanType = 1, arfima = FALSE, external.regressors = NULL), distribution.model = "norm", 
start.pars = list(), fixed.pars = list())

ajusteEGARCH <- ugarchfit(spec1,HSI[,2])

ajusteEGARCH

# Modelo TGARCH

spec1 <- ugarchspec(variance.model = list(model = "gjrGARCH", garchOrder = c(1, 1), 
submodel = NULL, external.regressors = NULL, variance.targeting = FALSE), 
mean.model = list(armaOrder = c(0, 0), include.mean = TRUE, garchInMean = FALSE, 
inMeanType = 1, arfima = FALSE, external.regressors = NULL), distribution.model = "norm", 
start.pars = list(), fixed.pars = list())

ajusteTGARCH <- ugarchfit(spec1,HSI[,2])

ajusteTGARCH