fmmc.cov: This is an implementation of covariance matrix estimation...

In arorar/covmat: Covariance Matrix Estimation

Description

This is an implementation of covariance matrix estimation using Factor Model Monte Carlo method as described in Jiang and Martin (2013).

Usage

fmmc.cov(R, factors, robust = FALSE, parallel = TRUE, align = c("end",
  "begin"), ...)

Arguments

R	vector of asset returns in xts format
factors	matrix of factor returns in xts format
robust	boolean to indicate if robust methods must be used for fitting factor models and constucting the covariance matrix. By default this option is turned off.
parallel	boolean to indicate if all cores on the system must be used
align	string to indicate where the longer backfilled return histories must be truncated when available returns are unequal. The default value is to align them at the end truncated to align the returns
...	allows passing paramters to factorAnalytics to butild the factor model or control parameters to covrob for constructing the robust covariance matrix

Details

This method takes in returns data and factors as time series. The returns and factor time-series must be aligned. The method is applicable when we have longer factor histories and shorter asset return histories. In this case past returns can be backfilled using the FMMC methodology. We can apply FMMC to all the assets in the portfolio. Once the returns are backfilled we can construct a more accurate measure of classical and robust covariance using backfilled returns. Backfilled returns for different assets must be aligned by truncating date in the begining or end is returns have unequal available return histories.

Author(s)

Rohit Arora

arorar/covmat documentation built on May 10, 2019, 1:48 p.m.