Importação dos dados
# install.packages("BatchGetSymbols")
require(BatchGetSymbols)
data_inicial <- as.Date('2015-01-01')
data_final <- as.Date('2018-01-01')
dados_ibov <- GetIbovStocks()
head(dados_ibov)
## tickers ticker.desc type.stock quantity percentage.participation ref.date
## 1 ABEV3 AMBEV S/A ON 4355174839 2.953 2021-07-29
## 2 ASAI3 ASSAI ON NM 157635935 0.554 2021-07-29
## 3 AZUL4 AZUL PN N2 327283207 0.645 2021-07-29
## 4 B3SA3 B3 ON NM 1930877944 5.167 2021-07-29
## 5 BBAS3 BRASIL ON ERJ NM 1283197221 1.760 2021-07-29
## 6 BBDC3 BRADESCO ON EJ N1 1147260246 1.125 2021-07-29
ativos <- c('ABEV3.SA','CMIG3.SA',
'PETR3.SA','VALE3.SA',
'^BVSP')
df <- BatchGetSymbols(tickers = ativos,
first.date = data_inicial,
last.date = data_final)
# Informações sobre o download
df$df.control
## # A tibble: 5 x 6
## ticker src download.status total.obs perc.benchmark.dat~ threshold.decisi~
## <chr> <chr> <chr> <int> <dbl> <chr>
## 1 ABEV3.SA yahoo OK 749 0.967 KEEP
## 2 CMIG3.SA yahoo OK 749 0.967 KEEP
## 3 PETR3.SA yahoo OK 749 0.967 KEEP
## 4 VALE3.SA yahoo OK 749 0.967 KEEP
## 5 ^BVSP yahoo OK 743 0.960 KEEP
# Dados das ações:
dados <- df$df.tickers
head(dados)
## price.open price.high price.low price.close volume price.adjusted
## 1 16.14 16.56 15.77 16.02 19396500 12.61954
## 2 15.91 15.97 15.54 15.73 13795900 12.39109
## 3 15.73 16.42 15.56 16.33 18610500 12.86373
## 4 16.34 16.72 16.14 16.56 20493500 13.04491
## 5 16.43 16.68 16.26 16.54 9054500 13.83095
## 6 16.47 16.65 16.30 16.50 11695700 13.79750
## ref.date ticker ret.adjusted.prices ret.closing.prices
## 1 2015-01-02 ABEV3.SA NA NA
## 2 2015-01-05 ABEV3.SA -0.01810265 -0.018102372
## 3 2015-01-06 ABEV3.SA 0.03814370 0.038143675
## 4 2015-01-07 ABEV3.SA 0.01408441 0.014084446
## 5 2015-01-08 ABEV3.SA 0.06025646 -0.001207609
## 6 2015-01-09 ABEV3.SA -0.00241820 -0.002418440
Organizando dados gerados
# Criar data frame com data, ativo e retorno ajustado
# install.packages('tidyverse')
require(tidyverse)
n <- length(ativos)
dados <- dados[,c(7,8,9)]
head(dados)
## ref.date ticker ret.adjusted.prices
## 1 2015-01-02 ABEV3.SA NA
## 2 2015-01-05 ABEV3.SA -0.01810265
## 3 2015-01-06 ABEV3.SA 0.03814370
## 4 2015-01-07 ABEV3.SA 0.01408441
## 5 2015-01-08 ABEV3.SA 0.06025646
## 6 2015-01-09 ABEV3.SA -0.00241820
colnames(dados) <- c("Data","Ativo","Retorno")
dados <- na.omit(dados) # Retirando os NA's
head(dados)
## Data Ativo Retorno
## 2 2015-01-05 ABEV3.SA -0.01810265
## 3 2015-01-06 ABEV3.SA 0.03814370
## 4 2015-01-07 ABEV3.SA 0.01408441
## 5 2015-01-08 ABEV3.SA 0.06025646
## 6 2015-01-09 ABEV3.SA -0.00241820
## 7 2015-01-12 ABEV3.SA 0.01272701
ABEV3 <- subset(dados,Ativo==ativos[1])[,-2]
CMIG3 <- subset(dados,Ativo==ativos[2])[,-2]
PETR3 <- subset(dados,Ativo==ativos[3])[,-2]
VALE3 <- subset(dados,Ativo==ativos[4])[,-2]
BVSP <- subset(dados,Ativo==ativos[5])[,-2]
retornos <- inner_join(ABEV3,CMIG3,by="Data")
retornos <- inner_join(retornos,PETR3,by="Data")
retornos <- inner_join(retornos,VALE3,by="Data")
retornos <- inner_join(retornos,BVSP,by="Data")
colnames(retornos) <- c("Data",ativos)
# install.packages('xts')
datas <- retornos$Data
require(xts)
retornos <- xts(retornos[,2:6],
order.by = datas)
plot(retornos, main = "Retornos dos ativos selecionados",
legend.loc = 'top')
Retorno médio e risco
(retornos_medio <- apply(retornos, 2, mean))
## ABEV3.SA CMIG3.SA PETR3.SA VALE3.SA ^BVSP
## 0.0006678302 -0.0002208844 0.0014193126 0.0016135065 0.0007185829
(desvio_p <- apply(retornos, 2, sd))
## ABEV3.SA CMIG3.SA PETR3.SA VALE3.SA ^BVSP
## 0.01292062 0.03252633 0.03392018 0.03520124 0.01458420
(variancia <- apply(retornos, 2, var))
## ABEV3.SA CMIG3.SA PETR3.SA VALE3.SA ^BVSP
## 0.0001669425 0.0010579620 0.0011505789 0.0012391271 0.0002126989
Construindo o portfólio de mínima variância
# install.packages(c('fPortfolio','PerformanceAnalytics','timeSeries'))
require(fPortfolio)
require(PerformanceAnalytics)
require(timeSeries)
serie_retornos <- as.timeSeries(retornos[,-5])
minvariancePortfolio(serie_retornos)
##
## Title:
## MV Minimum Variance Portfolio
## Estimator: covEstimator
## Solver: solveRquadprog
## Optimize: minRisk
## Constraints: LongOnly
##
## Portfolio Weights:
## ABEV3.SA CMIG3.SA PETR3.SA VALE3.SA
## 0.9364 0.0392 0.0000 0.0244
##
## Covariance Risk Budgets:
## ABEV3.SA CMIG3.SA PETR3.SA VALE3.SA
## 0.9364 0.0392 0.0000 0.0244
##
## Target Returns and Risks:
## mean Cov CVaR VaR
## 0.0007 0.0128 0.0280 0.0201
##
## Description:
## Thu Jul 29 16:18:46 2021 by user: WorkFranciscoNeto
portfolio <- minvariancePortfolio(serie_retornos)
# Pesos
pesos <- getWeights(portfolio)
pesos
## ABEV3.SA CMIG3.SA PETR3.SA VALE3.SA
## 0.93635901 0.03922184 0.00000000 0.02441914
weightsPie(portfolio)
# Simulação de $1
chart.CumReturns(serie_retornos,
wealth.index = TRUE,
legend.loc = "topleft",
main="Simulação de $1 em cada ativo")
retornos_minvar <- c()
for (i in 1:nrow(serie_retornos)) {
aux <- 0
for (j in 1:ncol(serie_retornos)) {
aux <- aux + pesos[j]*serie_retornos[i,j]
}
retornos_minvar <- c(retornos_minvar,aux)
}
retornos_minvar <- xts(retornos_minvar,
order.by = datas)
chart.CumReturns(retornos_minvar,
wealth.index = TRUE,
legend.loc = NULL,
main="Simulação de $1 no port. Min. Var")
serie_ibov <- xts(BVSP[,2], order.by = datas)
chart.CumReturns(serie_ibov,
wealth.index = TRUE,
main = "Simulação unitária no índice IBOV")
