Optimizing portfolio composition and maximising returns with the Portfolio Analytics package

Portfolio optimization is basically composed of four main steps:

  • Definition of portfolio components and past quotations
  • Definition of portfolio constrains, for instance, in term of diversification or maximum loss
  • Definition of objective to be optimized, usually in terms of returns
  • Definition of optimal percentage composition, given constraints and objectives

In this recipe, we will employ PortfolioAnaltycs and some other packages by joining together functionalities from different packages in order to provide a convenient and straightforward way to compose a financial portfolio.

The recipe workflow will be as follows:

  • Downloading stock prices
  • Definition of portfolio constraints and objectives
  • Actual portfolio optimization

Getting ready

In this recipe, we will join together powerful functions from different packages.

First of all, we will download stock quotations from Yahoo Finance by leveraging the quantmod package.

The downloaded information will then be grouped within a portfolio object and an optimal portfolio composition will be computed, employing functions from the PortfolioAnalytics, ROI, and Deoptim packages. The ROI package will be employed in its extended version, thanks to the glpk and quadrprog plugins.

In order to add those plugins, we will first have to install the Rcmdr package.

Summing it all, you will have to run the following code to install and load the required packages:

library(DEoptim)
library(ROI)
library(ROI.plugin.glpk)
library(ROI.plugin.quadprog)
library(plugin)
library(quantmod)
library(PortfolioAnalytics)

How to do it...

  1. Download data and store it in a portfolio data frame:
    stocks <- getSymbols(c("FCA","AAPL","GOOG"), env = sp500,
  
from = as.Date("2015-01-01"), 
  to = as.Date("2015-10-31"))
FCA <- sp500$FCA
FCA <- get("FCA",envir = sp500)
FCA <- with(sp500, FCA)
AAPL <- sp500$AAPL
AAPL <- get("AAPL",envir = sp500)
AAPL <- with(sp500,AAPL)
GOOG <- sp500$GOOG
GOOG <- get("GOOG",envir = sp500)
GOOG <- with(sp500,GOOG)
FCA  <- as.data.frame(FCA)
FCA  <- FCA$FCA.Adjusted
GOOG <- as.data.frame(GOOG)
GOOG <- GOOG$GOOG.Adjusted
AAPL <- as.data.frame(AAPL)
AAPL <- AAPL$AAPL.Adjusted

portfolio <- data.frame(FCA,GOOG,AAPL)
  2. Change the portfolio row names by setting them equal to dates:
    row.names(portfolio) <- seq.Date(from = as.Date("2015-01-01"), to = as.Date("2015-01-01") + nrow(portfolio)-1,by ="days")
  3. Initialize a portfolio object by passing to it stock names:
    portfolio_obj <- portfolio.spec(assets = colnames(portfolio))
  4. Add a minimum and a maximum composition constraint:
    portfolio_obj <- add.constraint(portfolio = portfolio_obj,
  type = "box",
  min = c(0.01, 0.28, 0.1),
  max = c(0.4, 0.5, 0.25))
  5. Add a diversification constraint:
    portfolio_obj <- add.constraint(portfolio = portfolio_obj, type = "diversification", div_target = 0.7)
  6. Specify mean return maximization as an objective:
    portfolio_obj <- add.objective(portfolio = portfolio_obj, type = 'return', name = 'mean')
  7. Compute optimal portfolio composition, given the constraints and objectives:
    optimal_portfolio_obj <- optimize.portfolio( R = portfolio, portfolio = portfolio_obj, optimize_method = "ROI", trace = TRUE)
print(optimal_portfolio_obj)

> print(optimal_portfolio_obj)
***********************************
PortfolioAnalytics Optimization
***********************************

Call:
optimize.portfolio(R = portfolio, portfolio = portfolio_obj, optimize_method = "ROI", trace = TRUE)

Optimal Weights:
 FCA GOOG AAPL
0.25 0.50 0.25

Objective Measure:
 mean
322.5
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset