VaR: Calculate VaR with Various Methods

In zhangkd5/riskCalc: Risk Calculation System -- MATH G4082 Project

Description

This generic function computes VaR and creates an S3 class object that contains data and VaR results.

Usage

VaR(data, ...)

## S3 method for class 'portfolioData'
VaR(data, TimeLen = 5, conf.level = 0.99,
  method = c("parametric", "historical", "Monte-Carlo"), lookback.years = 5,
  iteration = 10000, ...)

## S3 method for class 'optionsData'
VaR(data, TimeLen = 5, conf.level = 0.99,
  method = c("parametric", "historical", "Monte-Carlo"), V0 = 10000,
  lookback.years = 5, iteration = 10000, ...)

## S3 method for class 'stocksData'
VaR(data, TimeLen = 5, conf.level = 0.99,
  method = c("parametric", "historical", "Monte-Carlo"), V0 = 10000,
  lookback.years = 5, iteration = 10000, ...)

Arguments

data	an object that contains structured data of stocks, options, or a portfolio.
...	arguments to be passed to or from methods.
TimeLen	the time horizon in days that the VaR is defined over.
conf.level	the confidence level of VaR
method	the method used to calculate VaR. Must be one of "parametric" (default), "historical", or "Monte-Carlo", can be abbreviated.
lookback.years	the time horizon in years that historical simulation is based on.
iteration	the number of iteration Monte-Carlo simulation to be performed.
V0	the total value which VaR is measured upon.

Details

This function is able to compute portfolio VaR for portfolio data, and is also able to compute VaR for each individual stocks and/or options given stocks or options data.

In parametric method and Monte-Carlo method, the portfolio components should be firstly calibrated to a GBM model. In parametric method, the portfolio is assumed normally distributed and the options are linearly approximated by underlying stocks. In historical and Monte-Carlo method, the option values are fully evaluated.

See model documentation for details about VaR calculation methods.

Value

An object of class "VaR".

See Also

backtest to perform VaR backtesting.

zhangkd5/riskCalc documentation built on May 4, 2019, 10:16 p.m.