Book a Demo!
CoCalc Logo Icon
StoreFeaturesDocsShareSupportNewsAboutPoliciesSign UpSign In
braverock
GitHub Repository: braverock/portfolioanalytics
 Path: blob/master/demo/demo_weight_concentration.R
1433 views
1
#' ---

2
#' title: "Weight Concentration Demo"

3
#' author: Ross Bennett

4
#' date: "7/17/2014"

5
#' ---

6


7
#' This script demonstrates how to solve a constrained portfolio optimization 

8
#' problem to minimize weight concentration.

9


10
#' Load the packages and data

11
library(PortfolioAnalytics)

12
library(ROI)

13
library(ROI.plugin.quadprog)

14
data(edhec)

15
R <- edhec[, 1:8]

16
colnames(R) <- c("CA", "CTAG", "DS", "EM", "EMN", "ED", "FIA", "GM")

17
funds <- colnames(R)

18


19
#' Set labels based on category (e.g. market capitalization)

20
cap_labels <- c(rep("Group1", 2), rep("Group2", 2), 

21
                rep("Group3", 2), rep("Group4", 2))

22


23


24
#' Create initial portfolio object with category_labels.

25
init.portf <- portfolio.spec(assets=funds, category_labels=cap_labels)

26


27
#' Add basic constraints.

28
init.portf <- add.constraint(portfolio=init.portf, type="full_investment")

29
init.portf <- add.constraint(portfolio=init.portf, type="long_only")

30


31
#' Add objective to minimize variance.

32
init.portf <- add.objective(portfolio=init.portf, type="risk", name="var")

33


34
#' Run the optimization with var as the only objective.

35
opt1 <- optimize.portfolio(R=R, portfolio=init.portf, optimize_method="ROI", trace=TRUE)

36
opt1

37


38
#' Add the weight_concentration objective.

39
#' Set the conc_aversion values to 0 so that we should get the same result as

40
#' the minimum variance portfolio.

41
conc.portf <- add.objective(portfolio=init.portf, type="weight_concentration", name="HHI", 

42
                            conc_aversion=0, conc_groups=init.portf$category_labels)

43


44
#' Run the optimization with minimum concentration objective.

45
opt2 <- optimize.portfolio(R=R, portfolio=conc.portf, optimize_method="ROI", trace=TRUE)

46
opt2

47
all.equal(opt1$weights, opt2$weights)

48
chart.Weights(opt2)

49


50
#' Now change the conc_aversion values to control the concentration aversion

51
#' parameters.

52
conc.portf$objectives[[2]]$conc_aversion <- c(0.03, 0.03, 0.06, 0.02)

53
opt3 <- optimize.portfolio(R=R, portfolio=conc.portf, optimize_method="ROI", trace=TRUE)

54
opt3

55
chart.Weights(opt3)

56


57
#' We do not have a group constraint, but we can plot the groups based on

58
#' category labels in the portfolio object.

59
chart.GroupWeights(opt3, grouping="category", plot.type="barplot", col=bluemono)

60


61
#' High conc_aversion values should result in an equal weight portfolio.

62
conc.portf$objectives[[2]]$conc_aversion <- rep(1e6, 4)

63
opt4 <- optimize.portfolio(R=R, portfolio=conc.portf, optimize_method="ROI", trace=TRUE)

64
opt4

65
chart.Weights(opt4)

66


67