PerformanceAnalytics (version 0.9.5)

VaR.Beyond: calculate BVaR or loss Beyond traditional mean-VaR

Description

Beyond VaR purports to estimate average loss beyond VaR. Please note that your milage will vary; expect that values obtained from the normal distribution differs radically from the real situation.

BeyondVaR is described in theoretical detail in the paper: Gaussel, N., Legras, J., Longin, F., and Rabemananjara, R. "Beyond the VaR Horizon" 2001, Quants Review No. 37

Usage

VaR.Beyond(R, p=.95, modified=FALSE, add=FALSE, periods = 1)

Arguments

R

a vector, matrix, data frame, timeSeries or zoo object of asset returns

p

confidence level for calculation, default p=.99

modified

TRUE/FALSE default FALSE, use Cornish Fisher Expansion to take higher moments into account

add

TRUE/FALSE default FALSE, add VaR to BeyondVaR value for total predicted loss, like CVaRplus

periods

number of periods to calculate predicted loss beyond the VaR for, default 1

Value

• estimated Value of single period loss Beyond traditional mean-VaR

Details

We believe that the extension of Cornish-Fisher VaR as an option in the Beyond VaR calculation is unique to this library.

References

Gaussel, N., Legras, J., Longin, F., and Rabemananjara, R. "Beyond the VaR Horizon" 2001, Quants Review No. 37

See Also

VaR.CornishFisher VaR.traditional CVaR CVaRplus

Examples

data(edhec)

    # first do normal Beyond VaR calc
    VaR.Beyond(edhec)

    # now use modified Cornish-Fisher VaR calc to take non-normal distribution into account
    VaR.Beyond(edhec,modified=TRUE)