analyticVaR: analyticVaR

In f-kailin/varmonitor: Risk Management Tools: Value-at-Risk Monitor

Description

The function calculate the contribution in different scales to the portfolio Value-at-Risk.

Usage

analyticVaR(confidence, position, sigma, horizon = 1)

Arguments

confidence	Confidence interval. (e.g. "0.95" or a vector with different confidence interval)
position	Vector of positions in monetary terms.
sigma	Variance-covariance matrix of the whole portfolio.
horizon	Time horizon. (e.g. "daily: 1", "weekly: 5", "monthly: 22")

Details

Returns the marginal, component and relative component Value-at-Risk.
It is also scalable with vectors containing multiple confidence intervals.
Due to the multidimensional nature of this metric we would not suggest using multiple time horizons.

Value

"Delta VaR", "Component VaR" in monetary terms and "Relative-Component VaR" in decimals.

Note

The function can be run for a portfolio with a single exposure, but there will be a single value for "Component VaR" and "Relative-Component VaR".
Reminder: the central assumption under this function states that underlying market variables is normally distributed. This involves assuming a model for the joint distribution of changes in the market variables and using historical data to estimate the model parameters.

Author(s)

Gatti Riccardo, Lin Francesco

References

Hull J.C. (2015): "Value-at-Risk and Expected Shortfall" in "Risk Management and Financial Institutions", Wiley.
Hull J.C. (2015): "Model-Building Approach" in "Risk Management and Financial Institutions", Wiley.
Jorion P. (2007): "Portfolio Risk: Analytical Methods" in "Value at Risk", McGraw-Hill.

See Also

parametricVaR parametricES analyticES

Examples

##
weights <- c(100000, -100000, 100000)
varcov <- matrix(c(0.05, 0.03, 0.01, 0.03, 0.04, 0.02, 0.01, 0.02, 0.03), nrow = 3)
analyticVaR(c(0.95, 0.99, 0.999), weights, varcov, 5)
##
analyticVaR(.99, weights, varcov)
## NOT RUN
analyticVaR(.99, weights, varcov, c(1, 5))

f-kailin/varmonitor documentation built on Dec. 20, 2021, 7:39 a.m.