script.buildEDHEC.R

### Construct an xts object of EDHEC composite hedge fund style indexes

# Peter Carl

# Used for updating the edhec data object in PerformanceAnalytics

require(gdata)
require(xts)

# Download the following file to the working directory:
# http://www.edhec-risk.com/indexes/pure_style/data/table/history.csv
### @TODO: Is there a way to download it directly? Maybe not, seems to require a login
x<-read.csv(file=download.file("http://www.edhec-risk.com/indexes/pure_style/data/table/history.csv"), sep=";", header=TRUE, check.names=FALSE)
#x=read.csv(file="history.csv", sep=";", header=TRUE, check.names=FALSE)
x.dates = as.Date(x[,1], format="%d/%m/%Y")
x.data = apply(x[,-1], MARGIN=2, FUN=function(x){as.numeric(sub("%","", x, fixed=TRUE))/100}) # get rid of percentage signs
edhec = xts(x.data, order.by=x.dates)
colnames(edhec)

# calculate a wealth index
edhec.idx = apply(edhec, MARGIN=2, FUN=function(x){cumprod(1 + na.omit(x))})
# identify quarters
edhec.Q.idx=edhec.idx[endpoints(edhec.idx, on="quarters"),]
# calculate quarterly returns
edhec.Q.R=ROC(edhec.Q.idx)
# trim the last data point, if needed
# dim(edhec.Q.R)
# edhec.Q.R=edhec.Q.R[-61,]
# reclass the object
edhec.Q.R=as.xts(edhec.Q.R)
# lm requires safe names
colnames(edhec.Q.R)=make.names(colnames(edhec))