RPEIF (version 1.0)

IF: Influence Function for Available Risk Measures

Description

IF returns the data and plots the shape of either the IF or the IF TS for a risk measure specified.

Usage

IF(risk, returns = NULL, evalShape = FALSE, retVals = NULL,
  nuisPars = NULL, k = 4, IFplot = FALSE, IFprint = TRUE,
  compile = TRUE, prewhiten = FALSE, ar.prewhiten.order = 1,
  cleanOutliers = FALSE, cleanMethod = c("locScaleRob", "Boudt")[1],
  eff = 0.99, alpha.robust = 0.05, ...)

Arguments

risk

Risk measure.

returns

Vector of the returns of the asset or portfolio.

evalShape

Evaluation of the shape of the IF risk measure if TRUE. Otherwise, a TS of the IF of the provided returns is computed.

retVals

Values used to evaluate the shape of the IF.

nuisPars

Nuisance parameters used for the evaluation of the shape of the IF (if no returns are provided).

k

Range parameter for the shape of the IF (the SD gets multiplied k times).

IFplot

If TRUE, the plot of the IF shape or IF TS of the returns is produced.

IFprint

If TRUE, the data for the IF shape or the IF TS of the returns is returned.

compile

Boolean variable to indicate if the IF TS should be computed using compiled code (C++) (TRUE) or not (FALSE).

prewhiten

Boolean variable to indicate if the IF TS is pre-whitened (TRUE) or not (FALSE).

ar.prewhiten.order

Order of AR parameter for the pre-whitening. Default is AR(1).

cleanOutliers

Boolean variable to indicate whether the pre-whitenning of the IF TS should be done through a robust filter.

cleanMethod

Robust method used to clean outliers from the TS. The choices are "Boudt" and "locScaleRob" for the function.

eff

Tuning parameter for the normal distribution efficiency for the "locScaleRob" robust data cleaning.

alpha.robust

Tuning parameter for the quantile of the "Boudt" robust data cleaning algorithm, using the minimum covariance determinant estimator (MCD).

...

Additional parameters passed on to influence function of risk measure.

Examples

# NOT RUN {
# Plot of IF using the wrapper function
outIF <- IF(risk="mean",
            returns=NULL, evalShape=TRUE, retVals=NULL, nuisPars=list(mu=0.005),
            IFplot=TRUE, IFprint=TRUE)

#' # Loading data (hedge funds returns)
data(edhec, package="PerformanceAnalytics")
colnames(edhec) = c("CA", "CTAG", "DIS", "EM","EMN", "ED", "FIA",
                    "GM", "LS", "MA", "RV", "SS", "FoF") 
                    
# Plot of IF using wrapper function and with a specified TS 
outIF <- IF(risk="mean",
            returns=edhec[,"CA"], evalShape=TRUE, 
            retVals=seq(-0.1, 0.1, by=0.001), nuisPars=NULL,
            IFplot=TRUE, IFprint=TRUE)

# Computing the IF of the returns (with outlier cleaning and prewhitening) with a plot of IF TS
outIF <- IF(risk="mean",
            returns=edhec[,"CA"], evalShape=FALSE, retVals=NULL, nuisPars =NULL,
            IFplot=TRUE, IFprint=TRUE,
            compile=TRUE, prewhiten=FALSE,
            cleanOutliers=TRUE, cleanMethod=c("locScaleRob", "Boudt")[1], eff=0.99)

# }