getPortfolioReturns: Utility functions for creating portfolios of returns and other functions

Description

Gets portfolio returns from closing prices (configurable). This uses quantmod under the hood to retrieve prices and construct returns based on a configurable transform.

Also included is a function that returns the composition of select indexes that can be used in conjunction with getPortfolioReturns() to get the underlying returns of the given index.

Additionally, there is a utility function that ensures that symbols have been properly loaded.

Usage

getIndexComposition(ticker = '^GSPC', hint = NA, src = 'yahoo')
getPortfolioReturns(symbols, obs = NULL, start = NULL, end = Sys.Date(),
  fun = function(x) Delt(Cl(x)), reload = FALSE, na.value = NA, ...)
ensure(serie, src = 'FRED', reload = FALSE, ...)

Arguments

ticker

The ticker of the index. For best mileage, use Yahoo! compatible tickers (including the caret prefix).

hint

A hint that specifies the number of assets in the index. If omitted, a default will be used based on pre-configured data for common indexes.

src

The data vendor to use. Defaults to yahoo but could work with google

symbols

A vector (or scalar) of tickers to download

obs

The number of observations to get

start

Alternatively, a start date can be used to specify the beginning of a range to download.

end

The end date of the range. Defaults to current date.

fun

A transform applied to the downloaded data. The default is to calculate returns on the close.

reload

Whether to reload data or just download missing data

na.value

What value to use if the resulting portfolio has NAs. The default is to omit any assets containing NA values.

serie

A vector (or scalar) of tickers to ensure exist in the current environment

…

Additional parameters to pass to getSymbols

Value

getPortfolioReturns returns a TxM xts object of asset returns.

getIndexComposition returns a vector of asset symbols (i.e. tickers).

ensure returns nothing.

Details

Typically only getPortfolioReturns and getIndexComposition will be used on a regular basis.

Ensure isn't as useful as initially conceived given that naming collisions have caused numerous issues. The code now uses auto.assign=FALSE in the underlying getSymbols call.

Examples

Run this code

# NOT RUN {
# Get a portfolio
h <- getPortfolioReturns(c('A','AA','AAPL'), obs=150)

# Get an index portfolio
h <- getPortfolioReturns(getIndexComposition('^DJI'), obs=100, reload=TRUE)

# Doesn't work because of numerical symbols - need to fix
#h <- getPortfolioReturns(getIndexComposition('^HSI'), obs=100, reload=TRUE)

# Ensure that some assets exist
ensure(c('K','JNPR'), src='yahoo')
# }

Run the code above in your browser using DataLab

The Learning Leader's Guide to AI Literacy